Keywords

1 Introduction

In the years 2015–2017, the immigration of refugees has increased significantly. Access to education and work is an important step for their integration. Since a large proportion of refugees are young and do not have a formal vocational qualification, there is great potential in taking up dual apprenticeship training. Unfilled apprenticeship training positions with higher access chances represent a potential for this. Whether this potential is used by companies that have problems filling the apprenticeship training places they offer is a question that had not been empirically analysed. Up to 2018, these had been limited to structural characteristics of companies that provide apprenticeship training for refugees. Another characteristic of companies is their intensity of technology use, for which both conducive and obstructive relationships to the apprenticeship training of refugees are conceivable. The multivariate panel analyses of the 2016 and 2017 waves of the BIBB-Establishment Panel on Training and Competence Development provide initial indications that companies with unfilled apprenticeship training places and more digitalised companies are more likely to train refugees than companies without unfilled apprenticeship training places or less digitalised companies. However, the use of specific technologies is associated with a lower probability of training refugees.

The article is structured as follows: Following the thematic introduction, the presentation of the empirical state of research motivates the research question, which is explained in Sect. 3. Furthermore, in Sect. 3, hypotheses are derived on the basis of the theoretical background presented. Sect. 4 describes the methodological approach, the data basis, the chosen operationalisation and the model specification, before descriptive and multivariate results are presented in Sect. 5. A discussion follows in Sect. 6.

2 Introduction and Empirical State of Research

The integration of immigrant refugees in recent years represents an economic, political and social challenge. Integration in education and work is seen as playing a central role here. Particularly for adolescents and young adults among the refugees, integration in apprenticeship training can represent a potential for sustainable integration into employment and society. However, before a apprenticeship training contract is concluded, cost-benefit considerations are made on both sides, by refugees and companies, in which potentials and obstacles are weighed up. Apprenticeship training contracts are concluded if the general willingness to train refugees also leads to the actual offer of apprenticeship training places and this matches the applications of refugees and their suitability. In the following, the empirical state of research is presented along these indicators.

2.1 Asylum Seekers in Germany

In the years 2015–2017, more than 1.3 million asylum seekers have entered Germany (Bundesministerium des Inneren 2017, 2018). However, taking up apprenticeship training with the aim of sustainable integration into employment is not the most suitable path for all of them, depending in particular on their age and level of education and qualification. 39% of those seeking protection are under 16 years old and 44% are between 16 and 35 years old (Bundesagentur für Arbeit 2018a). Especially for refugees between 18 and 25, the potential for sustainable and successful integration into the labour market lies in apprenticeship training (Brücker 2016). Furthermore, even for a group of refugees who have already exceeded the original training age, the path to the German labour market will be via a German qualification, as only 10% can demonstrate in-company training (Bundesagentur für Arbeit 2018a). However, this should not be equated with professional skills, as work experience may have been acquired in the country of origin but not formally certified (Brücker et al. 2016; pq-portal 2016; von Radetzky and Stoewe 2016).

Integration in Training: Cost-Benefit Trade-Off on Two Sides

The importance of completed apprenticeship training is reflected in labour market opportunities. Apprenticeship training reduces the risk of unemployment (Röttger et al. 2017) and increases average lifetime earnings (Schmillen and Stüber 2014). In addition, graduates’ career entry is also usually smoother and without interim unemployment (Seibert and Wydra-Sommagio 2017). In the case of refugees of the years 2015–2017, moreover, apprenticeship training contributes to the legal consolidation of their residence in Germany (Schreyer et al. 2018). From the point of view of companies, apprentices perform productive activities and represent an investment in the future security of skilled labour in the company (Franz and Soskice 1995; Jansen et al. 2015; Lindley 1975).

However, for both refugees and companies, the costs of training are a decisive factor. For refugees, dual apprenticeship training is associated with high opportunity costs, since a lower income is achieved during the period of apprenticeship training, which lasts about 3 years; than through employment, even in the case of low-qualified jobs performing helper tasks (Ebbinghaus 2017a; Geis et al. 2016; Schreyer et al. 2018). On the side of the companies, bureaucratic hurdles and the special need for support and supervision in preparation as well as during apprenticeship training in particular represent an increased input of personnel, time and financial resources (Ebbinghaus 2016; Ebbinghaus and Gei 2017; Flake et al. 2017a). Added to this, is the often unclear legal situation and thus uncertain duration of stay (Flake et al. 2017a; Schreyer et al. 2018).

Refugees as Applicants for Apprenticeship Training Positions and Apprentices

In February 2018, 19,142 people in the context of refugee migration were registered with the Federal Employment Agency as applicants for apprenticeship training positions (Bundesagentur für Arbeit 2018b). In addition, according to assumption-based model calculations, the demand for apprenticeship training by young refugees will continue to rise: in 2019, up to 41,500 young refugees under the age of 30 will demand apprenticeship training (Winnige et al. 2017). However, apprenticeship training contracts are not concluded with all registered applicants for apprenticeship training places. In the apprenticeship training year 2015/2016, around 3500 of the approximately 10,300 youths and young adults in the context of refugee migration registered with the Federal Employment Agency entered into a apprenticeship training relationship (Granato and Neises 2017). For the following training year 2016/2017, the rate of successfully integrated applicants increased by 2.2 percentage points; just under 9500 of the approximately 26,500 applicants with a refugee background registered with the Federal Employment Agency were able to conclude an apprenticeship training contract (Matthes et al. 2018b).

2.2 Companies That Offer Apprenticeship Training Positions and Train Refugees

Recording the companies that are already training refugees is difficult due to the fact that the definition of refugee differs in some cases between surveys, e.g. through the citizenship of one of the main countries of origin of the asylum seekers or the status of an asylum application that has been submitted. The proportion of small and medium-sized companies surveyed in one study that had concluded apprenticeship training contracts with refugees was just under 3% in 2016 (Ebbinghaus 2016). At just under 6%, the proportion was higher among the craft companies surveyed than among the companies surveyed in industry and commerce and the liberal professions. According to the results of a company survey by the IW Personnel Panel (December 2016), 15% of companies qualified refugees through internships, training or regular employment (Flake et al. 2017a, b). In March 2017, training was the least common form of employment for refugees, at 7.2%, compared to internships (17%) and regular employment (10.2%) (Flake et al. 2017a).

Previous empirical work in the field of apprenticeship training of refugees from the company perspective analyses the assessment of apprenticeship training as an integration pathway, the willingness to train refugees, the actual supply of internships and apprenticeship training places, and the awareness and use of funding instruments (including Bellmann et al. 2017; Ebbinghaus 2016, 2017a, b; Flake et al. 2017a, b; Gerhards 2018). Accordingly, apprenticeship training engagement seems to differ by company size and industry (Baic et al. 2017; Deutscher Industrie- und Handelskammertag 2016; Ebbinghaus 2017a; Ebbinghaus and Gei 2017; Geis et al. 2016). Some obstacles are seen as more problematic in companies with 250 or more employees, whereas the effort with which small and medium-sized companies integrate refugees seems to be lower (Flake et al. 2017a). This could possibly be due to more standardised processes in larger companies and flatter and more flexible structures in smaller companies. Nevertheless in 2016, larger companies are more likely to offer training places for migrants than smaller ones (Bellmann et al. 2017) and small and medium-sized companies are significantly more likely than micro companies to offer internships/apprenticeship training places for refugees (Ebbinghaus 2017a). On the other hand, small companies in particular consider the bureaucratic effort required for clarifying immigration and labour law to be too great (Juretzka 2014; Söhn et al. 2017). Thus, the additional time and personnel resources required are a particular obstacle for small and medium-sized companies (Ebbinghaus 2016).

Unfilled apprenticeship training places in occupations with staffing problems could lead to higher access opportunities and represent a potential for integrating refugees interested in apprenticeship training into in-company training. In regional terms, filling problems are particularly pronounced in eastern Germany and, beyond that, in the craftwork occupations (Matthes et al. 2018b), which recorded the strongest absolute increase in the number of new contracts, regardless of refugee status (Matthes et al. 2018b). At the same time, a large proportion of refugees have already gained work experience in the craftwork occupations (Neske and Rich 2016). According to multivariate analyses, refugees are more likely to be employed in the craftwork than in other sectors (Flake et al. 2017b). In addition, multivariate results are available on the offer of internships/apprenticeship training places for refugees by companies: Companies with unfilled apprenticeship training places are more likely to offer internships/apprenticeship training places for refugees on their own initiative than companies without apprenticeship training vacancies (Ebbinghaus 2017a). However, this is only the case if the difficulties in filling apprenticeship training places are due to quantitative reasons, i.e. too few applicants, and not if applicants do not meet the requirements, for example (Ebbinghaus 2017a).

Nevertheless, both aspects cannot be equated with the filling of unfilled apprenticeship training places and the use of their potential for the integration of refugees in apprenticeship training. As a result, in 2018 it was not yet possible to make an empirically proven statement on the extent to which the supply of apprenticeship training places actually leads to the creation of apprenticeship training contracts. The result that there is no significant correlation between the situation on the apprenticeship training market and the probability of refugees making the transition to apprenticeship training (Matthes et al. 2018a) provides an initial indication.

3 Research Question, Theoretical Background and Hypotheses

This paper examines the probability that apprenticeship training contracts under the Vocational Training Act (BBiG)/Crafts and Trade Regulation Code (HwO) will be established between companies and refugees and the extent to which filling problems with apprenticeship training places on offer and the companies’ status of digitalisation influence this probability. Unfilled apprenticeship training places represent a potential for integrating refugees into apprenticeship training just as refugees represent a new target group for companies with recruitment problems. But are companies also aware of this potential and make use of it? With regard to the extent of the companies’ digitalisation, it is conceivable that there are both, contexts that hinder and conducive to the apprenticeship training of refugees.

Since unfilled apprenticeship training places as well as digitalisation are related to the apprenticeship training decision of companies via the coverage of qualification requirements, they are considered under control of the companies’ qualification structure, which can influence both aspects as well as the companies’ training commitment.

3.1 Problems in Filling Apprenticeship Training Places on Offer

One apprenticeship training motive for companies is the investment motive, which places the focus of a companies’ training decision on investing in its own skilled labour supply. If vacancies remain unfilled, the development of new target groups can represent a strategy for action in addition to the external recruitment of skilled workers (Becker and Lübbers 2013; Ebbinghaus 2017a). With regard to the (expected) increased resource expenditure for companies associated with the apprenticeship training of refugees, it is assumed that companies are more likely to train refugees if they can also benefit from increased returns. If there are recruitment difficulties due to a lack of or shortage of external training place applicants, this increases search and recruitment costs for apprenticeship training positions and skilled worker positions. These can be avoided through in-house apprenticeship training; the opportunity returns from apprenticeship training increase (Beckmann 2002; Kempf 1985; Lehne 1991; Niederalt 2004).

The expected-value-theory goes beyond this investment approach: according to it, a decision takes into account both the value of the outcome of the action and the probability that the action will actually lead to the desired outcome (Beckmann and Heckhausen 2006; Granato et al. 2016). Companies that have a high probability of finding apprentices in their existing target group and thus have no staffing difficulties will compare the cost-benefit ratio of training non-refugees with the cost-benefit ratio of training refugees. Due to the higher costs and the constant or even lower return of training refugees, they are less likely to decide in favour of training refugees than companies with staffing difficulties. Companies with staffing difficulties compare “training refugees” with “not training refugees”. In the latter case, however, they have a low probability of finding apprentices, so that the cost-benefit ratio deteriorates due to lost training returns and costs of external skilled labour recruitment and the relative cost-benefit ratio of the “training of refugees” case improves. If costs and returns increase in a similar proportion or if returns increase more than costs, companies with staffing problems are more likely to train refugees than companies without staffing problems. However, if the costs associated with the action are higher than the hoped-for return, a threshold value of the cost-return expectation is undercut (Esser 1999; Granato et al. 2016) and the apprenticeship training relationship between the company and the refugee would not come about. The reason for this could be, for example, that the applicant does not meet the requirements.

However, an apprenticehsip training relationship can only come about if there is demand from young people for the apprenticeship training places on offer. Applicants’ chances of gaining access to apprenticeship training vary with the occupational and regional problems of filling and supplying apprenticeship training places. However, a high access opportunity in an occupation with filling problems can only have a motivating effect if, on the one hand, it is known and subjectively assessed as being higher and, on the other hand, the incentive of a target does not fall below a certain threshold value. Thus there are also occupations which, from the point of view of young people, are fundamentally outside the “zone of acceptable alternatives” (Gottfredson 2004). The decision thus takes into account varying access opportunities and the value of apprenticeship training in the desired occupation as well as in the alternative occupation.

Hypothesis 1

The probability of training refugees is higher in companies with unfilled apprenticeship training places than in companies that do not have filling difficulties.

3.2 The Companies’ Status of Digitalisation

Another strategy for meeting qualification needs can be the introduction or expansion of digital infrastructure. If technologies are subtitutively related to employees in the middle skills segment, a decrease in apprenticeship training commitment can be assumed. If, on the other hand, they have a complementary relationship, an increase in company investment in apprenticeship training can be expected. The question of the extent to which the increasing digitalisation of companies determines the apprenticeship training of refugees can be plausibly linked to both facilitating and hindering relationships.

On the one hand, greater digitalisation could make processes and tasks more complex and the requirements overall, as well as linguistic and technical requirements in particular, higher and thus the probability of integrating refugees into an apprenticeship training programme could be lower compared to less digitalised companies. The use of information and communication technologies could be an indication of the communication intensity of the tasks, in which the language barriers of non-native speakers are particularly significant and the probability of training refugees could be lower. The use of digital production and control infrastructure could be associated with a lower probability of training refugees due to more complex processes and high demands on refugees with language barriers or (still) incomplete previous training.

Hypothesis 2a

The probability of training refugees is lower in companies that introduce or expand technologies and digital infrastructure than in companies that are less digitalised.

On the other hand, in more digitalised companies the obstacle of insufficient language skills could be lower due to technological support and thus increase the probability of integrating refugees into apprenticeship training compared to less digitalised companies. The use of information and communication technologies could support internal and external communication. The use of digital production and control infrastructure could be an indication of lower communication intensity and be positively related to the apprenticeship training of refugees.

Hypothesis 2b

The probability of training refugees is higher in companies that introduce or expand technologies and digital infrastructure than in companies that are less digitalised.

3.3 Taking into Account the Evolution of the Skill Structure

The development of the companies’ qualification structure can be related to apprenticeship training and unfilled apprenticeship training positions as well as to investments in digitalisation. The companies’ qualification structure, in particular the development of the share of employees in the middle qualification level in total employment, can be positively or negatively related to apprenticeship training. If the share of employees in the middle qualification level decreases due to shortages of skilled workers, this can be accompanied by an intensification of apprenticeship training investments or investments in digital technologies that automate processes and substitute tasks in the middle qualification level. In this context, shortages of skilled workers are often also linked to problems in filling apprenticeship training places. If a polarised development in the qualification structure of the workforce is assumed as a result of increasing digitalisation, this can be accompanied by a decline in the demand for employees with middle qualifications and in-company training. If, on the other hand, increasing digitalisation leads to an upgrading/upskilling process in which the demand for employees in the middle qualification level increases, an increase in in-company training investment can be assumed. This can lead to additional or even intensified problems in filling apprenticeship training places. However, it should be noted that increasing technisation is not the sole driver of developments in companies’ qualification structures (Helmrich et al. 2016).

These considerations are just as relevant for the question of whether or not refugees are trained in companies as they are for non-refugees. A declining share of qualified employees in total employment could be due to a declining demand for apprenticeship-trained skilled workers and thus be associated with a lower probability of apprenticeship training. If, on the other hand, a declining share of qualified employees is due to a shortage of skilled workers, this is likely to be associated with an increased commitment to apprenticeship training.

4 Methodological Approach

Data Basis

The 2016 and 2017 survey waves of the BIBB-Establishment Panel on Training and Competence Development are used as the data basis for the analyses conducted here. This is a representative panel survey of 3500 companies in Germany that has been conducted annually since 2011. The survey focuses on the activities of companies in the areas of apprenticeship training and continuing vocational training. In addition, varying thematic focuses on current, scientifically and socio-politically relevant issues are surveyed. The survey on the apprenticeship training of refugees refers to persons who have been recognised as refugees or have the status of asylum seekers or tolerated persons.

Operationalization

The probability that a company trains at least one refugee in accordance with the BBiG/HwO is operationalised via the binary variable “refugee(s) undergoing training in accordance with BBiG/HwO”, which takes the value 1 if the company employs at least one refugee as an apprentice in accordance with BBiG/HwO and the value 0 if this is not the case. The filling problems regarding apprenticeship training places offered under the BBiG/HwO as one of the central independent variables are taken into account in the model via the indicator variable “Unfilled apprenticeship training places”, which takes on the value 0 if the company has no unfilled apprenticeship training places and the value 1 if the company has unfilled apprenticeship training places but not all of the apprenticeship training places offered are unfilled. Companies for which all apprenticeship training places on offer are unfilled and thus by definition do not provide apprenticeship training are thus not taken into account. The companies’ digitalisation or use of technology is modelled using two indicator variables that each take the value 1 if information and communications technologies or digital production and control infrastructure have been newly introduced and/or expanded. In addition, five indicator variables are used, each taking the value 1 if the company uses a corresponding technology, otherwise 0. Technologies are included that support internal and external communication as well as internal organisation, such as “Digital technologies related to networking with suppliers and between companies”, “Information and communication-related digital work equipment or applications” and “Personnel or work organisation-related technologies”. Furthermore, the use of “Digital network technologies to support business and work activities” and “Computer-controlled equipment and digital work devices or tools to create products and services” are included. In addition, the development of the qualification structure is taken into account via the indicator variable “Decreasing share of middle qualification level”, which takes the value 1 if the share of employees in the middle qualification level in total employment has decreased compared to the previous year, otherwise 0.

In addition, the following companies’ structure characteristics are included as control variables: The size of the company is included as a categorical four-level variable. Larger companies are more likely to have the necessary organisational, financial and personnel requirements than smaller companies and therefore train refugees more frequently. In addition, financial resources are positively related to the status of the digital infrastructure and more professional training structures are negatively related to problems filling apprenticeship training places. The location in western or eastern Germany is taken into account as an indicator variable because staffing problems are more pronounced in the east, but the proportion of refugees among applicants may be lower there. In addition, in line with empirical analyses of the apprenticeship training decision, the economic sector and the validity of a sectoral or company collective bargaining agreement are included in the models via a categorical eight-stage variable and an indicator variable.

Model Specification

Due to the binary dependent variable “refugee(s) in training according to BBiG/HwO”, a logistic model specification is chosen which is estimated as a panel model with randomised effects. In the basic model (model 1), the above-mentioned variables (unfilled apprenticeship training places, information and communication technologies, digital production and control infrastructure as well as the decreasing share of the middle qualification level, economic sector, company size, West German location and collective bargaining agreement coverage) are included. In model 2, the five indicator variables of the specific technologies are included instead of the indicator variables of information and communication technologies and production and control infrastructure in order to obtain indications of existing heterogeneity. As there are no major deviations in other model specifications in which additional interaction effects between the decreasing share of the middle qualification level and the introduction or expansion of digital technologies and infrastructure were included, in the coefficients of the central independent variable “unfilled apprenticeship training places” as well as the coefficients of information and communication technology and digital production and control infrastructure, they are not presented here.

5 Empirical Results

5.1 Results of Descriptive Analyses

Tables 1 and 2 contain descriptive statistics. Approximately 7% of companies employ refugees as apprentices under the BBiG/HwO. Overall, some 18% of companies have unfilled training places pursuant to the BBiG/HwO, although this does not account for 100% of the training places on offer. The share of companies that train refugees is 6.3% for companies that have no unfilled apprenticeship training places and 10.1% for companies that have unfilled apprenticeship training places.

Table 1 Share of companies with apprenticeship training of refugees pursuant to BBiG/HwO according to selected characteristics
Full size table
Table 2 Descriptive statistics, incl. differentiation by companies with and without refugees as apprentices according to BBiG/HwO
Full size table

The share of companies that introduce and/or expand new information and communication technologies (ICT) and digital production and control infrastructure (PST) is significantly higher in the group of companies that train refugees than in the group of companies that do not train refugees. The average share of companies that introduce and/or expand ICT is just under 60%, whereas the share with regard to PST is significantly lower at just under 40%. The shares of companies using “Technologies related to networking with suppliers and between companies” and “Computer-controlled systems and digital work equipment or tools for the creation of products and services” are comparatively low at around 55% and just under 60% respectively, whereas “Digital network technologies to support business and work activities” and “Information and communication-related work equipment” have a penetration rate of over 90%. Only in the case of the share of use of “Personnel and work organisation-related technologies” there is a significant difference between companies with and without refugees as apprentices in accordance with the BBiG/HwO. This technology is used more frequently in companies that train refugees (85%) than in companies that do not train refugees (79%).

5.2 Results of Multivariate Analyses

According to the estimation results, the probability that a company trains refugees in occupations in accordance with the BBiG/HwO is significantly influenced by the companies’ situation of unfilled apprenticeship training places, the status of the companies’ digitisation and the development of the qualification structure.

The basic model (Model 1 in Fig. 1) shows that companies with unfilled apprenticeship training places are more likely to train refugees than companies without unfilled apprenticeship training places. This can be taken as an indication that both, companies with problems filling their apprenticeship training places perceive refugees as a new target group and that refugees take advantage of better access opportunities for themselves. This positive coefficient also exists in Model 2 (in Fig. 1). Contrary to the hypothesis that increased use of technology, including more complex processes, should increase requirements and inhibit rather than promote the likelihood of hiring refugees as apprentices, the coefficients of the indicator variables “information and communication technologies (ICT)” and “digital production and control infrastructure (PST)” are both significant and positive. The introduction or expansion of information and communication technologies and digital production and control infrastructure thus does not appear to hinder the apprenticeship training of refugees. On the contrary, they seem to be related in a supportive way.

Fig. 1
A table has 3 columns labeled refugees in apprenticeship training according to B B i G or H w O, M 1, and M 2. The significance level is illustrated at the bottom. The last row reads, controlling for firm size, industry, West German firm location, and collective bargaining agreement coverage.

Regression results on the apprenticeship training of refugees according to BBiG/HwO. (Source: BIBB-Establishment Panel on Training and Competence Development 2016–2017, logit models with randomised effects, average marginal effects, N = 2376, own calculations)

Full size image

By including the indicator variables for the use of specific technologies, these findings can be supplemented: “Digital network technologies to support business and work activities” and “digital technologies related to networking with suppliers and between companies” are associated with a significantly lower probability of training refugees and thus correspond to the initial hypothesis of the negative influence of technology use on the probability of training refugees. The coefficients of the use of “computer-controlled equipment and digital work tools or means to create products and services” and the use of “information and communication-related digital work tools or applications” have positive signs, as do PST and ICT in Model 1 (in Fig. 1), but are not significantly different from zero. In the case of “information and communication-related digital work devices or applications”, the high penetration rate could ensure that there is no significant difference in terms of their use.

The positive and significant coefficients of ICT and PST show that companies that invest in technologies in these areas are more likely to invest in the apprenticeship training of refugees. The introduction and/or expansion of digital technologies or infrastructure in these areas does not appear to be an obstacle to the apprenticeship training of refugees or even to promote this. The inclusion of the indicator variables for individual technologies shows heterogeneous correlations here. “Digital network technologies and technologies for networking with suppliers” appear to be more of an obstacle to the apprenticeship training of refugees, possibly due to increased (language) requirements. When interpreting the coefficients for the technologies, however, it should be noted that the introduction and/or expansion of ICT or PST, as well as the use of the specific technologies, relate to the respective company as a whole and do not necessarily influence the tasks of the apprentices.

6 Discussion

According to the results, the potential of unfilled apprenticeship training places for the integration of refugees in apprenticeship training seems to be perceived and used by companies and refugees. The results of this article strengthen the empirical evidence, that has emerged up to 2018, that companies with unfilled apprenticeship training places are more likely to offer internships/apprenticeship training places for refugees on their own initiative than companies without unfilled apprenticeship training places (Ebbinghaus 2017a). At least if the staffing problems are quantitative. The present analyses extend these findings to the extent that apprenticeship training offers by companies also lead to apprenticeship training contracts. With regard to the transition probability of refugees into apprenticeship training, Matthes et al. (2018a) identified no significant correlation with the situation on the appreticeship training market, which is operationalised on the one hand via apprenticeship training places offered per 10 persons interested in apprenticeship training, and on the other hand via unfilled apprenticeship training place offers in the Federal Employment Agency District of the applicant. The result that companies with unfilled apprenticeship training places are more likely to provide apprenticeship training for refugees than companies without unfilled apprenticeship training places implies that companies with staffing problems are more willing to incur the increased expense of providing apprenticeship training for refugees and thus to mitigate the already perceived shortage of applicants and future shortages of skilled workers in the long-term.

Despite the positive situation that the existing potential of unfilled apprenticeship training places is being used for the integration of refugees into apprenticeship training, additional apprenticeship training offers will be necessary to meet the increasing demand. After all, in 2017, the total number of young people in demand (603,500) exceeds the number of training places on offer (572,200) (Matthes et al. 2017). At the same time, the number of unsuccessful applicants is almost twice as high as the number of unfilled apprenticeship training places (Matthes et al. 2017). In order to make apprenticeship training possible for all interested parties, additional apprenticeship training offers must be created by companies, also due to the expected increase in demand. To this end, the level of awareness of existing funding and support measures must be increased and their fit with the needs of companies improved (Flake et al. 2017b). In addition, a narrow focus on occupations with staffing problems and shortages of skilled workers should be avoided. Within the framework of vocational orientation programmes, the entire spectrum of occupations should be presented for this purpose, as well as for non-refugees. In the long-term, the framework conditions and the quality of apprenticeship training in the single companies should also be improved in order to reduce filling problems due to a lack of attractiveness from the perspective of young people.

In addition, the observation of this issue will remain interesting in the future as a changed composition according to countries of origin and thus the (school) education and qualification structure can be observed among immigrants of the years 2015–2017. The utilisation of the potential of unfilled apprenticeship training places could possibly require even more intensive preparation for apprenticeship training and thus slow down or fundamentally differ from the situation observed to date. In addition, the successful completion of the apprenticeship training taken up will be of particular importance for the assessment of the long-term situation, since this enables access to the qualified labour market.

Analyses of sectors of the economy show heterogeneous influence structures within the sectors. Economic sectors differ greatly in some cases with regard to the proportion of companies that train refugees in occupations in accordance with the BBiG/HwO, the proportion of unfilled apprenticeship training places compared to those offered and the level of digitalisation and technology use in the company. In the construction industry and the other, predominantly personal services (e.g. accommodation and catering as well as transport and warehousing), the proportion of companies with refugees in apprenticeship training as well as the proportion of unfilled apprenticeship training places are both above average, while investment in PST is below average. In agriculture and mining, too, the share of unfilled apprenticeships on offer and the share of companies introducing and/or expanding digital production and control infrastructure are above average. Nonetheless, agriculture and mining have the lowest share of companies that train at least one refugee, at around 3% (Table 1). Further detailed analyses of the economic sectors could provide information on barriers, but are not the subject of this article.

In addition, further analyses could be devoted to apprenticeship training in social care occupations in connection with the integration of refugees into apprenticeship training, as there are also increasing shortages of skilled workers and recruitment problems in this occupational field. This topic would complement the limited consideration of apprenticeship training in accordance with the BBiG/HwO that results from the data situation in this article. In addition, the evaluation of official statistics at occupational level according to the status of a persons’ refugee context could provide further, supplementary insights into the topic.