Abstract
Career guidance assists students with the school-to-work transition. Based on a survey conducted in secondary schools in Germany, we analyze career guidance activities and how these affect career plans. The take-up of career guidance depends upon the school track attended, and the school and the class setting, while personal characteristics are barely relevant. The effects of counseling depend upon the type of counseling provider. Counseling by the employment agency reduces plans for educational upgrading and increases the probability of applying for an apprenticeship, while the effects of counseling by school counselors works in the opposite direction for lower track students.
1 Introduction
Adolescents make far reaching decisions under considerable uncertainty regarding the continuation of education and the transition into the labor market after graduation from secondary schools (McNally, 2016; Harmon et al., 2003; Heckman et al., 2018). Career guidance assists students in making these decisions by providing career related information, counseling, mentoring, or first hand work experience. Whereas counseling and mentoring mostly aim to provide information and advice, work experience placements allow students to learn about work organization in general as well as specific occupations and employers offering apprenticeships. Lower track students[1] are the most intensively targeted group for career guidance in Germany because of the increasing difficulties they face during school-to-work transitions, despite the very good general labor market conditions (BMBF, 2018; Bonin et al., 2016). Furthermore, career guidance has been expanded in the middle and upper track of secondary schools in Germany.[2] To our knowledge, there is only a scarce literature estimating the effects of counseling on the path choice below tertiary education, which amounts to a choice between a continuation of general schooling and a vocational track.
This paper analyzes the take-up of career counseling and work experience placements as well as their effects on career planning, based on a survey we conducted in lower and middle track secondary schools in Germany. Focusing on the differences across school tracks, we analyze different types of career guidance activities and their effect on different career planning outcomes. This way we account for the different potential pathways of the school-to-work transitions in Germany below tertiary-level education.
Career guidance in U.S. high schools focuses on encouraging students to attend college and assisting them with applications. In Europe, career guidance is mostly concerned with the adolescents’ choice of different tracks. In most European countries, career guidance assists in making the choice between a general, more academic track and a vocational track within the school-based education system (e. g. Goux et al., 2015; Bernardi et al., 2014). Because of the crucial role of the vocational training system in Germany, career guidance in lower and middle track secondary schools targets the choice between a direct transition into vocational training, mostly by starting an apprenticeship, and continuing general schooling, often with the goal to obtain a higher school qualification,[3] and starting vocationaltraining later (Bonin et al., 2016). Germany requires adolescents to make this choice regarding the school-to-work transition at a young age, 15 to 16, upon graduating from lower or middle track secondary schools (Biewen and Tapalaga, 2017). Because of the importance of this choice, career guidance has been expanded over time. Hence, research on career guidance is highly relevant for Germany as well as for other European countries which have a similar system of vocational training as Germany or which view the German vocational training system as a role model.
The decision to start an apprenticeship in Germany involves the choice among more than 300 apprenticeship occupations and the timing of the entry into the labor market has lasting consequences on later life outcomes (Bonin et al., 2016; Hanushek et al., 2017). This complex decision is made under imperfect information and uncertainty. Behavioral economics suggests that at this young age it is unlikely that individuals make rational decisions regarding human capital investments (DellaVigna, 2009; Golsteyn et al., 2014; Koch et al., 2015; Goux et al., 2015; Lavecchia et al., 2016) – and the same argument applies with regard to career choices (Bonin et al., 2016). While career guidance in Germany has the long-term goal of a smooth and successful school-to-work transition, short- and medium-term goals are more complex, having to accommodate the students’ decision process. The short-term measures assist students in career planning, forming realistic expectations and aspirations, and making timely career-related choices. The medium-term goals are ambiguous as to whether a direct transition into an apprenticeship or a continuation of general schooling is the preferable choice.
Evidence on the effectiveness of career guidance based on randomized controlled experiments is mixed. Various studies find that providing students with information on returns to tertiary education, support with applying for financial aid, and individual mentoring show positive short-run effects on graduations from high school, applications for college, and college attendance (Bettinger et al., 2012; Rodríguez-Planas, 2012; Carell and Sacerdote, 2017; Ehlert et al., 2017; Peter and Zambre, 2017). Long-run effects are much smaller and/or non-significant (Bettinger et al., 2012; Rodríguez-Planas, 2012). Evidence for Finland shows no effect on applications or enrollment in post-secondary education when providing additional information on the returns of different degree options and on the associated occupations to be working in (Kerr et al., 2015). Randomized experiments typically measure the causal effect of small scale treatments designed by researchers, where typically all participants in the treatment arm receive the treatment. Such studies provide limited insights on the take-up of existing large scale non-mandatory career guidance activities and the effects of take-up on different outcomes. Our study uses nonexperimental regression analyses to estimates the impact of such counseling practices as applied in the field, thus complementing the experimental evidence.
Existing non-experimental studies tend to find positive effects of career guidance. Neumark and Rothstein (2007) and Boockmann and Nielen (2016) show that counseling programs that assist low-performing students with educational decisions and applications improve labor entry. Bernardi et al. (2014) and Fitzenberger and Licklederer (2017) find that additional career assistance in secondary school results in a revision of education plans, possibly through a growing awareness of opportunities and risks. Hoest et al. (2013) and Saniter et al. (2019) find that professional, standardized career guidance increases educational attainment. For tertiary education decisions, Borghans et al. (2015) study the take-up and effectiveness of career guidance in Dutch secondary schools, which is found to increase the enrollment rate in the preferred field in university. Both, Neumark and Rothstein (2006) and Borghans et al. (2015) find that individual characteristics do little to explain take-up of career guidance.
Solga and Kohlrausch (2013) and Fitzenberger and Licklederer (2015) investigate the effectiveness of work experience placements in Germany, which are a key component of career guidance activities in lower and middle secondary schools. Work experience placements show a positive effect on apprenticeship take-up. For the UK, work experience placements result in some positive, but weak effects on career planning, employability, and wages of students (Hillage et al., 2001; Mann and Percy, 2014; Messer, 2018). Internships during secondary school in the U.S. both increase college attendance and employment after high school (Neumark and Rothstein, 2007).
Our study addresses two research questions. First, we provide evidence on the supply of different types of career guidance and on the determinants of the individual take-up and intensity of use. In Germany, career guidance is provided by schools, local initiatives, and the employment agencies, taking the form of counseling and work experience placements. Second, we estimate the effect of the take-up of career guidance activities on career plans. We conducted a school survey in two cities in Southwest Germany focusing on students in lower and middle track secondary schools at the point in time at which they chose between beginning an apprenticeship or continuing general schooling. We find that the take-up of counseling and work experience placements is barely associated with individual characteristics. Rather, differences in take-up are strongly driven by class-level effects. There is only limited evidence that students who are expected to face greater difficulties in career planning engage more intensively in career guidance activities, thus adding credibility to our nonexperimental estimates.
Career plans are measured by reporting a desired occupation, apprenticeship applications, and plans to continue schooling. For lower track students, frequent counseling by school counselors increases the probability of reporting a desired occupation, but school counseling does not affect the other measures. Further, counseling by the employment agency shows a positive effect on reporting a desired occupation and apprenticeship applications but it negatively affects plans for the continuation of schooling. Among middle track students, counseling by the employment agency has a positive effect on reporting a desired occupation, and frequent meetings increase apprenticeship applications. A higher number of work experience placements increases (reduces) apprenticeship applications (the continuation of schooling) in the middle track. A key finding is that the employment agency appears to be more effective in supporting career planning towards starting an apprenticeship.
The paper is organized as follows. Section 2 describes our survey. Section 3 provides evidence on the take-up and type of career guidance counseling and work experience placements. Section 4 investigates the relationship between career guidance and career planning. Section 5 concludes.
2 Data
The data was collected through use of a survey among secondary school students in 9th and 10th grade in spring 2014, in the two cities of Mannheim and Freiburg. Both cities are in the state of Baden-Württemberg and have the same education system. Our empirical analysis is restricted to lower track and middle track students, because after graduation these students face the choice between an apprenticeship (or school-based vocational training) and the continuation of general schooling.[4] Middle track students were surveyed in 10th grade, while lower track students were surveyed both in 9th and 10th grade.[5]
Using a paper and pencil questionnaire, we surveyed students in the classroom, provided parents had given their consent. The use of financial incentives for participation in the classroom survey was not allowed. Under these circumstances, we had a satisfactory response rate in contacted classes of 29 %. In addition, we surveyed parents and teachers. Parents were asked about their level of education, migratory background, and educational aspirations for their children.
Table 1 presents descriptive statistics of the students in the sample and in the overall student population. The share of students with a migratory background and the share of females are comparable to the overall population. We oversampled lower and middle track students for whom career guidance traditionally is more important than for upper track students. Table A.1 in the Appendix shows descriptive statistics for lower and middle track students separately.
Representativeness of the Sample
| Mannheim | Freiburg | |||
| Population | Sample | Population | Sample | |
| Lower Track | 19 % | 29 % | 13 % | 29 % |
| Middle Track | 24 % | 16 % | 21 % | 27 % |
| Upper Track | 47 % | 32 % | 58 % | 31 % |
| Share with Migratory Background | 47 %a | 42 %b | 21 %c | 22 %c |
| Female | 50 % | 53 % | 50 % | 52 % |
Notes: aEducation Report Mannheim school year 2012–2013: Population share below the age of 27 with migratory background. bShare of surveyed students growing up in bilingual families. cOnline Statistics Freiburg school year 2012–2013: Population share below the age of 27 with migratory background.
3 Take-up of career guidance
For the purpose of our analysis we focus on counseling and work experience placements as career guidance activities. Counseling is provided by teachers, school-based counselors, and the local employment agency. Work experience placements are common in Germany to familiarize students with work environments, with the option to gain contacts for an apprenticeship later on.
What are the determinants of take-up of career guidance activities? One hypothesis is that students receiving less support from their parents due to weaker labor market knowledge and networks as well as low-performing students are more likely to take up career guidance, because they need more support (henceforth, need-hypothesis [NH]). A second hypothesis is that schools and teachers affect the amount of career guidance that students actually use (henceforth, supply-hypothesis [SH]), because they affect students’ behavior by communicating the benefit of career guidance and the importance of career planning. This section describes the observed career guidance activities and provides evidence on the determinants of take-up.
3.1 Counseling
Career guidance through individual counseling and coaching of secondary school students has expanded over the last few decades in Germany, especially in the lower track where is has become a major part of the school curriculum (Kohlrausch and Solga, 2012; Saniter et al., 2019). Career guidance counseling is provided by local employment agencies and within schools. Local employment agencies offer counseling at their own job information centers. In some cases, and in particular for lower track students, counselors of the employment agency offer counseling hours at schools.
In contrast, school-based career guidance counseling is typically managed by schools or local school authorities in cooperation with municipalities without being standardized across Germany. Thus, there is a lot of regional variation in the type and quantity of school-based counseling. Often, one teacher or the head teacher is in charge of career guidance for students, providing job information and some assistance with applications for work experience placements or apprenticeships. Further, there exists a large number of local programs providing additional intensive career guidance mostly targeted at the lower track. In Mannheim, the local career guidance counseling project (“Ausbildungslotsen”) was extended in 2013 with the aim of providing individual counseling to all lower track students. In Freiburg, the program “Successful into Apprenticeship” (“Erfolgreich in Ausbildung”) for the lower track has been running since the late 2000’s. It involves additional classroom-based career guidance as well as group and individual counseling both provided by local educational providers (Fitzenberger and Licklederer, 2015).
The effect of counseling may differ by type of provider (here: employment agency vs. school counselors of local initiatives). They have different training backgrounds and slightly different perspectives on the school-to-work transition. Counselors of the employment agency are case workers with expertise on youth labor markets. School counselors are usually hired by local educational providers and allocated to schools, with typically one counselor per school. Most counselors are trained social workers. The employment agency focuses on the immediate transition into the labor market while school-based counseling may put a greater emphasis on the continuation of schooling as it might improve the students’ future chances on the labor market.
Table 2 shows first descriptive evidence on the take-up of different types of career guidance by students in the middle and lower tracks of secondary school. Career guidance by school-based counselors is used more intensively by students of the lower track.[6] Whereas 85 % of the students in the lower secondary school track have taken up the support of counselors at school, only 37 % of students in the middle track speak with a school counselor about career guidance, reflecting that the school-based counseling programs focus on the lower track. In addition, in the lower track students have more meetings (7.7 on average) with school counselors than in the middle track (2.4 on average). Hence, individual counseling of lower track students not only involves almost all students but is also quite intensive.
Take-up of Career Guidance Counseling Services by School Track
| Type of Secondary School | sig | |||
| lower track | middle track | |||
| meeting school counselor | 0.85 | 0.37 | ∗∗∗ | |
| Av. number of counseling meetings | 7.72 | 2.35 | ∗∗∗ | |
| meeting employment agency | 0.50 | 0.71 | ∗∗∗ | |
| Av. number of counseling meetings | 1.99 | 1.60 | ∗∗ | |
| meeting teacher | 0.34 | 0.21 | ∗∗∗ | |
| Av. number of teacher meetings | 4.42 | 1.92 | ∗∗∗ | |
| counseling outside school | 0.12 | 0.09 | ||
| multiple take-up of difference services | ||||
| meeting 1 counselor | 0.28 | 0.48 | ∗∗∗ | |
| meeting 2 counselors | 0.38 | 0.24 | ∗∗∗ | |
| meeting 3 counselors | 0.22 | 0.14 | ∗ | |
| meeting 4 counselors | 0.01 | – | ||
Notes: Stat. significant difference: ∗
Type of Support Provided by Career Guidance Counselors by School Track
| lower track | middle track | sig | ||
| School counselor | ||||
| Type of Support provided | ||||
| Discussion of career/educational options | 0.84 | 0.93 | ∗ | |
| Support with applications | 0.74 | 0.37 | ∗∗∗ | |
| Information about vacant apprenticeships | 0.54 | 0.44 | ||
| Matching of apprenticeships | 0.40 | 0.31 | ||
| Support was helpful | 0.80 | 0.80 | ||
| Employment agency | ||||
| Type of Support provided | ||||
| Discussion of career/educational possibilities | 0.68 | 0.86 | ∗∗∗ | |
| Support with applications | 0.28 | 0.20 | ||
| Information about vacant apprenticeships | 0.54 | 0.32 | ∗∗∗ | |
| Matching of apprenticeships | 0.39 | 0.32 | ||
| Support was helpful | 0.70 | 0.77 | ||
| Teacher | ||||
| Type of Support provided | ||||
| Discussion of career/educational possibilities | 0.79 | 0.79 | ||
| Support with applications | 0.48 | 0.45 | ||
| Information about vacant apprenticeships | 0.29 | 0.21 | ||
| Matching of apprenticeships | 0.29 | 0.15 | ||
| Support was helpful | 0.79 | 0.65 | ||
Notes: Conditional on take up. Stat. significant difference: ∗
Counseling offered by the employment agency is the most commonly used type of career guidance for middle track students. 71 % of the students in the middle track and 50 % of those in the lower track have at least one meeting with a counselor of the employment agency. However, in the lower track this type of counseling is less intensive than the counseling by school counselors. For the middle track, our subsequent analysis focuses on counseling by the employment agency.
Teachers play only a minor role as advisers for career guidance as only 34 % of the lower track students and 21 % of the middle track students make use of such support. Students in the lower track on average meet 4.4 times with teachers, whereas students from the middle track have 2 meetings. The majority of the lower track students meets with two or more different counselors (school counselors, teachers, the employment agency etc.) while middle track students on average meet a counselor only once. Overall, students in the lower track thus receive significantly more career guidance than students in the middle track.
Table 3 shows the different types of support provided by teachers, school counselors, and the employment agency as well as evidence on students’ satisfaction with the support both conditional upon meeting one of the counselors. The most important type of support is a discussion of career and education options. Lower track students also receive support by school counselors regarding applications (73 %) and information about vacant apprenticeships (54 %). The employment agency mostly offers information on career and education options for middle track students and on vacant apprenticeships for lower track students. Teachers also discuss career and education options with the majority of students (80 %) and they provide application support for about half of the students in both tracks.
The majority of students considers counseling to be helpful. With 80 % satisfied students school counselors seem to be most helpful, but the employment agency is deemed helpful by 70 % of the students in the lower track and 78 % in the middle track. Support by teachers is considered somewhat less helpful by middle track students as well.
In what follows, we focus on counseling by the employment agency and by school counselors, because the take-up of career guidance by teachers is less common (Table 2) and difficult to separate from regular schooling. Our analysis of the take-up of counseling by school counselors and by the employment agency distinguishes between the incidence of take-up and the intensity of counseling (for school counselors/employment agency intensive use means at least three/two meetings).[7]
Determinants of Take-Up of Counseling by an Employment Agency or School Counselor – Lower Track (Marginal effects)
| School counselors | Employment agency | |||||||
| Take up | at least 3 meetings | Take up | at least 2 meetings | |||||
| (1) | (2) | (1) | (2) | (1) | (2) | (1) | (2) | |
| Female | 0.109∗∗ | 0.035 | −0.048 | −0.202∗∗ | −0.051 | −0.118 | −0.023 | −0.082 |
| (0.055) | (0.032) | (0.094) | (0.096) | (0.099) | (0.096) | (0.082) | (0.081) | |
| 9th Grade | −0.015 | 0.006 | −0.190 | −0.345∗∗∗ | −0.137 | −0.145 | −0.262∗∗∗ | −0.316∗∗∗ |
| (0.075) | (0.038) | (0.127) | (0.123) | (0.113) | (0.121) | (0.075) | (0.061) | |
| German spoken in family | −0.074 | −0.057 | −0.342∗∗ | −0.402∗∗∗ | −0.046 | −0.015 | 0.066 | 0.099 |
| (0.078) | (0.035) | (0.140) | (0.122) | (0.116) | (0.127) | (0.086) | (0.099) | |
| Parents college | 0.007 | −0.011 | 0.067 | 0.068 | 0.062 | 0.083 | −0.155 | −0.154 |
| (0.066) | (0.034) | (0.094) | (0.139) | (0.125) | (0.128) | (0.146) | (0.134) | |
| Parents encourage effort in school | 0.012 | −0.003 | −0.117 | −0.126 | −0.063 | −0.059 | −0.072 | −0.079 |
| (0.054) | (0.035) | (0.086) | (0.082) | (0.097) | (0.099) | (0.068) | (0.068) | |
| Parents proud of educ. achievement | 0.049 | 0.018 | 0.266∗∗ | 0.242∗∗ | 0.058 | 0.049 | 0.160∗ | 0.127 |
| (0.054) | (0.030) | (0.109) | (0.106) | (0.135) | (0.141) | (0.088) | (0.090) | |
| Ambitious friends | −0.029 | −0.035 | 0.093 | −0.031 | −0.035 | −0.069 | −0.023 | −0.048 |
| (0.055) | (0.030) | (0.116) | (0.109) | (0.107) | (0.106) | (0.093) | (0.114) | |
| Good Math grade | 0.011 | −0.009 | −0.117 | −0.246∗∗ | −0.086 | −0.065 | −0.010 | −0.036 |
| (0.052) | (0.024) | (0.102) | (0.111) | (0.123) | (0.121) | (0.102) | (0.099) | |
| Good German grade | 0.015 | 0.020 | −0.115 | −0.088 | −0.005 | −0.047 | −0.045 | −0.045 |
| (0.069) | (0.039) | (0.092) | (0.101) | (0.128) | (0.126) | (0.061) | (0.064) | |
| Openness | −0.053∗∗ | −0.029∗∗ | −0.012 | 0.002 | 0.031 | 0.033 | 0.004 | 0.002 |
| (0.022) | (0.012) | (0.036) | (0.043) | (0.040) | (0.039) | (0.027) | (0.024) | |
| Extraversion | −0.007 | 0.001 | 0.079∗∗ | 0.101∗∗ | −0.026 | −0.016 | 0.004 | 0.008 |
| (0.024) | (0.016) | (0.038) | (0.043) | (0.041) | (0.038) | (0.033) | (0.035) | |
| Conscientiousness | −0.012 | −0.003 | −0.013 | −0.020 | 0.052 | 0.044 | 0.044 | 0.039 |
| (0.022) | (0.014) | (0.049) | (0.058) | (0.034) | (0.037) | (0.036) | (0.042) | |
| Neuroticism | −0.006 | −0.000 | 0.053 | 0.071 | −0.047 | −0.026 | −0.050 | −0.027 |
| (0.024) | (0.016) | (0.047) | (0.051) | (0.045) | (0.047) | (0.033) | (0.035) | |
| Agreeableness | 0.038 | 0.022 | 0.044 | 0.040 | 0.046 | 0.039 | 0.095∗∗ | 0.077∗ |
| (0.028) | (0.015) | (0.037) | (0.038) | (0.039) | (0.040) | (0.040) | (0.042) | |
| external locus of control | 0.029∗∗ | 0.019∗∗∗ | 0.022 | 0.009 | 0.066∗ | 0.047 | 0.105∗∗∗ | 0.077∗∗ |
| (0.014) | (0.006) | (0.043) | (0.054) | (0.036) | (0.038) | (0.034) | (0.031) | |
| internal locus of control | 0.034 | 0.023 | 0.020 | 0.006 | −0.022 | −0.059 | −0.068∗ | −0.105∗∗ |
| (0.034) | (0.020) | (0.041) | (0.041) | (0.045) | (0.049) | (0.037) | (0.041) | |
| Risk loving | 0.011 | 0.006 | 0.005 | −0.000 | 0.013 | 0.014 | 0.016 | 0.014 |
| (0.009) | (0.004) | (0.015) | (0.019) | (0.013) | (0.014) | (0.012) | (0.012) | |
| School dummies | yes | yes | yes | yes | ||||
| pseudo | 0.126 | 0.247 | 0.151 | 0.282 | 0.076 | 0.131 | 0.185 | 0.277 |
| Observations | 154 | 154 | 154 | 154 | 153 | 153 | 153 | 153 |
Notes: Marginal effects of probit estimations, controls for city and missing grades included. Standard errors clustered by class in parentheses, ∗
Determinants of Take-Up of Counseling from the Employment Agency – Middle Track (Marginal effects)
| Employment agency | ||||
| Take up | at least 2 meetings | |||
| (1) | (2) | (1) | (2) | |
| Female | −0.096 | −0.090 | −0.007 | 0.024 |
| (0.066) | (0.062) | (0.062) | (0.043) | |
| German spoken in family | 0.129 | 0.055 | −0.010 | 0.051 |
| (0.125) | (0.133) | (0.132) | (0.100) | |
| Parents college | −0.058 | −0.081 | −0.102 | −0.050 |
| (0.080) | (0.092) | (0.084) | (0.065) | |
| Parents encourage effort in school | −0.114 | −0.121 | 0.024 | −0.007 |
| (0.088) | (0.086) | (0.052) | (0.045) | |
| Parents proud of educ. achievement | 0.115∗∗ | 0.166∗∗∗ | 0.061 | 0.074 |
| (0.057) | (0.060) | (0.078) | (0.058) | |
| Ambitious friends | 0.232∗∗ | 0.298∗∗∗ | 0.115∗ | 0.135∗∗ |
| (0.091) | (0.065) | (0.068) | (0.054) | |
| Good Math grade | 0.063 | 0.044 | −0.113∗∗ | −0.060∗ |
| (0.079) | (0.074) | (0.047) | (0.035) | |
| Good German grade | 0.123∗ | 0.082 | −0.069 | −0.029 |
| (0.075) | (0.077) | (0.098) | (0.076) | |
| Openness | 0.034 | 0.077∗ | 0.031 | 0.037 |
| (0.037) | (0.040) | (0.027) | (0.024) | |
| Extraversion | −0.097∗∗∗ | −0.119∗∗∗ | −0.052 | −0.049∗ |
| (0.034) | (0.043) | (0.033) | (0.027) | |
| Conscientiousness | −0.038 | −0.030 | 0.011 | 0.001 |
| (0.029) | (0.035) | (0.032) | (0.029) | |
| Neuroticism | 0.007 | −0.008 | −0.025 | −0.032 |
| (0.037) | (0.038) | (0.034) | (0.029) | |
| Agreeableness | 0.062 | 0.027 | −0.061∗∗ | −0.039∗ |
| (0.038) | (0.042) | (0.024) | (0.020) | |
| external locus of control | 0.006 | −0.002 | 0.018 | 0.023 |
| (0.043) | (0.046) | (0.045) | (0.040) | |
| internal locus of control | 0.155∗∗∗ | 0.180∗∗∗ | −0.004 | 0.003 |
| (0.054) | (0.049) | (0.051) | (0.043) | |
| Risk loving | 0.018 | 0.004 | 0.019 | 0.007 |
| (0.015) | (0.015) | (0.014) | (0.010) | |
| School dummies | yes | yes | ||
| pseudo | 0.147 | 0.282 | 0.097 | 0.258 |
| Observations | 160 | 160 | 160 | 160 |
Notes: Marginal effects of probit estimations, controls for city included. Standard errors clustered by class in parentheses, ∗
First, we consider the determinants of take-up in the lower school track. Table 4 reports the average marginal effects of probit regressions on the take-up probability. There are almost no significant individual determinants of lower track students’ take-up of counseling at school or at the employment agency. We find no evidence for the need-hypothesis with regard to the incidence of take-up. However, looking at the intensity of take-up we find some evidence for the need-hypothesis: Lower track students from non-German speaking families are more likely to meet with school-based counselors more frequently. We do not find a comparable relationship for counseling by the employment agency.
As very few middle track students meet with school counselors, Table 5 focuses on meeting a counselor from the employment agency. Contrary to the lower track, middle track students meeting employment agency counselors are slightly positively selected with regards to peers and non-cognitive skills. Hence, there is once again no support for the need-hypothesis in the incidence of counseling take-up. Only good math grades are negatively correlated with the intensity of take-up of counseling provided by the employment agency, and thus provides only very weak evidence for the need-hypothesis for the middle tracks students.
Our findings on the take-up of counseling are robust when accounting for class and school fixed effects. The OLS regressions reported in Tables A.2 and A.3 in the Appendix provide very similar findings to the Probit regressions discussed above, even after accounting for class fixed effects. Further, the OLS regressions show that including school fixed effects or class fixed effects increases the explanatory power considerably, in particular for the intensity of counseling in the lower track. Thus, the school and the class setting are important determinants of the take-up and intensity, being even more relevant than personal characteristics. This is in line with the supply-hypothesis, while our findings provide only weak evidence for the need-hypothesis. The differences regarding the relevance of the supply-hypothesis between school tracks fit the observed setting of strongly institutionalized career guidance at lower track schools and weaker institutions at middle track schools.
3.2 Work experience placements
As a second type of career guidance activity, we consider work experience placements in local firms. Most placements last about a week and workplaces are not predetermined by the school. They are key opportunities for secondary school students to acquire practical job experience in different occupations, and to present themselves to potential apprenticeship employers. In addition, there are also job visit days in firms (“Praxistage”), sometimes organized by sponsors and firms who partner with the school.
While job visit days are not used intensively in our sample (on average less than 2 days), work experience placements are much more relevant in career guidance (Table 6). On average, lower track students complete 3.5 placements with an average total duration of about 23 days (exceeding the state target of at least 20 days for the lower track (Schröder, 2015)). Middle track students complete, on average, 2.1 placements with a total duration of 12 days. The differences are highly significant and sizeable, particularly in light of the fact that about two thirds of the lower track students are in 9th grade (see Table A.1) while all middle track students are in 10th grade.
75 % of students find work experience placements by themselves, while the second most frequent channel involves family and relatives. However, with a share of 36.8 %, lower track students use this search channel significantly less than those in the middle track. This difference probably reflects social selection by track (see Table A.1). Lower track students receive additional support from counselors and teachers when searching for work experience placements, while this is not the case for middle track students.
Descriptive Statistics on Work Experience Placements by School Track
| Secondary School Track | sig | |||
| lower track | middle track | |||
| Number of “Job Visit Days” | 1.83 | 1.75 | ||
| Number of work experience placements | 3.52 | 2.08 | ∗∗∗ | |
| Av. duration of work experience placement (days) | 7.96 | 6.00 | ∗∗∗ | |
| Total duration of work experience placements (days) | 22.52 | 11.99 | ∗∗∗ | |
| Search channels for work experience placements | ||||
| Student by him/herself | 0.72 | 0.75 | ||
| School counselor | 0.15 | 0.01 | ∗∗∗ | |
| Teacher | 0.10 | 0.03 | ∗∗∗ | |
| Family/relatives | 0.37 | 0.51 | ∗∗∗ | |
| Work experience placement Quality | ||||
| Quality of supervision at work experience placement (scale 0–3) | 1.56 | 1.75 | ∗∗∗ | |
| Enjoyed work experience placement (scale 0–3) | 1.43 | 1.57 | ∗∗ | |
| work experience placement in desired occupation | 0.43 | 0.37 | ∗∗ | |
| Most enjoyed work experience placement in desired occupation | 0.47 | 0.40 | ||
Notes: Stat. significant difference ∗
Middle track students, on average, rated their work experience placements better than lower track students, both in regard to the quality of supervision during the placement and by how much they enjoyed it. Only a third of the students in the sample completed a work experience placement in their desired future occupation. The fit of the placements to the students’ interests might be an important channel for successful career planning, as students can adjust their expectations and preferences. Table 7 contrasts the sector shares among all actual placements, among placements rated best by each student, and among the desired occupations. Manufacturing and health have the highest share of desired occupations and many students have work experience placements in these sectors. However, some sectors (like trade and sales, social/care work, education) show a lot of placements, even though student interest is much lower. At the same time, there are other sectors (like public service/administration, information technology) that often fit the desired future occupation but only a few student complete placements in these sectors. The evidence in Table 7 reveals a mismatch between desired occupations and actual placements, suggesting that there is a need to inform students about the actual labor market opportunities and to help students form more realistic expectations (Goux et al., 2015). It could also point to the need to offer more diversified placements.
Sector of Work Experience Placements, Best Work Experience Placement and Desired Occupation
| Sector | Work experience pl. | Best Work experience pl. | Desired occupation |
| Health | 20.6 % | 21.4 % | 17.0 % |
| Trade and sales | 18.3 % | 17.9 % | 14.8 % |
| Social/care work, education | 17.0 % | 12.4 % | 11.2 % |
| manufacturing/engineering | 15.8 % | 16.9 % | 17.3 % |
| Humanities | 1.1 % | 1.4 % | 1.1 % |
| Information technology | 1.7 % | 1.7 % | 5.1 % |
| Natural Sciences | 1.3 % | 1.4 % | 2.9 % |
| Skilled crafts and trades | 4.1 % | 2.4 % | 2.2 % |
| Construction | 2.8 % | 3.4 % | 3.2 % |
| Creative/Entertainment | 5.0 % | 6.6 % | 5.8 % |
| Food production/gastronomy | 4.5 % | 5.2 % | 5.1 % |
| Public service/administration | 2.1 % | 3.8 % | 7.2 % |
| Other Services | 5.6 % | 5.5 % | 7.2 % |
| Observations | 753 | 290 | 277 |
Determinants of Quantity and Quality of Work Experience Placement (Lower Track)
| 3 or more Work experience pl. | Work experience pl. in desired occup. | |||
| Female | −0.187∗∗ | −0.121 | 0.309∗∗∗ | 0.371∗∗∗ |
| (0.075) | (0.086) | (0.086) | (0.097) | |
| 9th Grade | −0.228∗∗ | −0.328∗∗∗ | 0.096 | 0.058 |
| (0.108) | (0.086) | (0.093) | (0.091) | |
| German spoken in family | −0.235∗∗ | −0.320∗∗∗ | −0.115 | −0.142 |
| (0.098) | (0.102) | (0.091) | (0.096) | |
| Parents college | −0.293∗ | −0.312∗∗ | 0.244 | 0.278 |
| (0.152) | (0.154) | (0.180) | (0.186) | |
| Parents encourage effort in school | 0.087 | 0.084 | 0.121 | 0.120 |
| (0.110) | (0.118) | (0.091) | (0.086) | |
| Parents proud of educ. achievement | 0.093 | 0.113 | 0.000 | 0.011 |
| (0.097) | (0.099) | (0.096) | (0.096) | |
| Ambitious friends | −0.162 | −0.119 | −0.022 | 0.036 |
| (0.099) | (0.110) | (0.113) | (0.115) | |
| Good Math grade | 0.077 | 0.062 | −0.012 | −0.024 |
| (0.120) | (0.120) | (0.139) | (0.140) | |
| Good German grade | 0.064 | 0.122 | 0.115 | 0.132 |
| (0.086) | (0.083) | (0.110) | (0.111) | |
| Openness | 0.007 | 0.004 | −0.053 | −0.057 |
| (0.033) | (0.033) | (0.040) | (0.042) | |
| Extraversion | −0.013 | −0.012 | 0.013 | 0.019 |
| (0.039) | (0.045) | (0.032) | (0.032) | |
| Conscientiousness | 0.091∗∗∗ | 0.093∗∗ | 0.026 | 0.026 |
| (0.034) | (0.040) | (0.047) | (0.051) | |
| Neuroticism | 0.025 | 0.004 | −0.037 | −0.047 |
| (0.030) | (0.033) | (0.035) | (0.037) | |
| Agreeableness | −0.046 | −0.058 | 0.004 | 0.001 |
| (0.038) | (0.042) | (0.040) | (0.045) | |
| external locus of control | −0.039 | −0.048 | −0.042 | −0.043 |
| (0.039) | (0.047) | (0.051) | (0.052) | |
| internal locus of control | −0.066 | −0.022 | −0.039 | −0.022 |
| (0.056) | (0.059) | (0.053) | (0.056) | |
| Risk loving | −0.007 | −0.017 | 0.023 | 0.019 |
| (0.018) | (0.019) | (0.016) | (0.018) | |
| Take-up employment agency | 0.086 | 0.158∗∗ | −0.066 | −0.064 |
| (0.098) | (0.077) | (0.133) | (0.146) | |
| 2 or more meetings employment agency | 0.084 | 0.145 | 0.218∗ | 0.250∗∗ |
| (0.130) | (0.142) | (0.113) | (0.108) | |
| Take-up school counselor | −0.020 | −0.022 | 0.102 | 0.135 |
| (0.120) | (0.135) | (0.165) | (0.174) | |
| 3 or more meetings school counselor | 0.044 | 0.047 | 0.063 | 0.061 |
| (0.081) | (0.091) | (0.115) | (0.118) | |
| Own placement search | 0.069 | 0.039 | ||
| (0.124) | (0.127) | |||
| Placement search family | 0.221∗∗ | 0.195∗∗ | ||
| (0.105) | (0.097) | |||
| Placement search counselor | −0.039 | −0.035 | ||
| (0.164) | (0.178) | |||
| School dummies | yes | yes | ||
| pseudo | 0.143 | 0.233 | 0.178 | 0.205 |
| Observations | 159 | 159 | 159 | 159 |
Notes: Marginal effects of probit estimations, controls for city and missing grades included. Standard errors clustered by class in parentheses, ∗
Next, we analyze the determinants of both the quantity and the quality of work experience placements (Table 8 for lower track and 9 for middle track). We measure quantity here by a dummy indicating the completion of three or more work experience placements. The quality of work experience placements is measured by the dummy variable for a match between sector of placement and desired occupation.
Lower track students from non-German speaking families and of parents without college degrees are more likely to have completed three or more work experience placements. Thus, they seem to use these to compensate for missing labor market networks, which supports the need-hypothesis. If lower track students have met at least once with the employment agency they also have more placements, a higher frequency of counseling meetings however is not relevant. In contrast, a higher frequency of counseling meetings with the employment agency increases the probability of completing a work experience placement in the desired occupation. This might be the result of adjusted expectations or search assistance for a matching placement by the employment agency. There is no support for the need-hypothesis in the quality of the work experience placements. Rather, to the contrary, students with assistance from their family and relatives are more likely to find a matching placement. Female students are more likely to complete a matching placement. We cannot disentangle whether female students have adjusted their expectations earlier to the available options, or whether they got lucky because more placements are offered in female-dominated sectors (see Table 7).
Determinants of Quantity and Quality of Work Experience Placement (Middle Track)
| 3 or more Work experience pl. | Work experience pl. in desired occup. | |||
| Female | 0.010 | −0.027 | 0.111 | 0.114 |
| (0.079) | (0.082) | (0.091) | (0.091) | |
| German spoken in family | 0.069 | 0.051 | −0.035 | −0.066 |
| (0.131) | (0.126) | (0.156) | (0.177) | |
| Parents college | 0.058 | 0.022 | −0.189∗∗ | −0.233∗∗ |
| (0.060) | (0.059) | (0.096) | (0.091) | |
| Parents encourage effort in school | −0.092 | −0.059 | −0.025 | −0.051 |
| (0.065) | (0.068) | (0.092) | (0.096) | |
| Parents proud of educ. achievement | −0.082 | −0.084 | 0.148 | 0.170∗ |
| (0.066) | (0.069) | (0.092) | (0.097) | |
| Ambitious friends | −0.000 | −0.022 | 0.055 | 0.086 |
| (0.104) | (0.101) | (0.083) | (0.089) | |
| Good Math grade | 0.004 | 0.007 | −0.102 | −0.099 |
| (0.064) | (0.063) | (0.095) | (0.096) | |
| Good German grade | −0.017 | 0.003 | −0.142 | −0.202∗ |
| (0.049) | (0.053) | (0.103) | (0.109) | |
| Openness | −0.021 | −0.020 | −0.022 | −0.025 |
| (0.020) | (0.020) | (0.037) | (0.038) | |
| Extraversion | 0.012 | 0.008 | 0.025 | 0.019 |
| (0.026) | (0.030) | (0.039) | (0.044) | |
| Conscientiousness | 0.065∗∗ | 0.068∗∗ | −0.012 | −0.017 |
| (0.028) | (0.028) | (0.048) | (0.050) | |
| Neuroticism | 0.032 | 0.049∗∗ | −0.003 | −0.009 |
| (0.023) | (0.023) | (0.041) | (0.047) | |
| Agreeableness | 0.099∗∗∗ | 0.097∗∗∗ | −0.005 | −0.011 |
| (0.032) | (0.033) | (0.044) | (0.040) | |
| external locus of control | 0.038 | 0.022 | −0.126∗∗∗ | −0.146∗∗∗ |
| (0.040) | (0.043) | (0.042) | (0.039) | |
| internal locus of control | −0.060 | −0.060 | 0.041 | 0.032 |
| (0.047) | (0.047) | (0.073) | (0.072) | |
| Risk loving | 0.026∗ | 0.026∗ | 0.011 | 0.013 |
| (0.014) | (0.014) | (0.018) | (0.020) | |
| Take-up employment agency | −0.122∗ | −0.148∗ | −0.011 | −0.069 |
| (0.068) | (0.089) | (0.099) | (0.104) | |
| 2 or more meetings employment agency | 0.117 | 0.155∗ | −0.072 | −0.026 |
| (0.073) | (0.092) | (0.085) | (0.085) | |
| Own placement search | 0.199∗ | 0.227∗∗ | ||
| (0.104) | (0.106) | |||
| Placement search family | 0.098 | 0.151 | ||
| (0.101) | (0.111) | |||
| School dummies | yes | yes | ||
| pseudo | 0.112 | 0.149 | 0.123 | 0.159 |
| Observations | 161 | 161 | 161 | 161 |
Notes: Marginal effects of probit estimations, controls for city included. Standard errors clustered by class in parentheses, ∗
For middle tracks students, there is no evidence supporting the need-hypothesis for the quantity of work experience placements regarding family background and grades. On the contrary, more conscientious and agreeable students complete more placements. In terms of quality of the work experience placement, we again find no support for the need-hypothesis. Middle track students with parents with college degree are less likely to complete a placement in their desired occupation. This might possibly be due to higher occupational aspirations accordant to their parents’ background. They are more likely to complete a matching placement when they searched by themselves, contrary to lower track students who needed assistance from their family.
Columns (2) and (4) in Tables 8 and 9 include school fixed effects. Schools might differ in the default number of placements students are expected to complete, in their network of cooperating firms offering placements, and the effort made to help students complete adequate placements.[8] The marginal effects of personal and family characteristics do not change much compared to columns (1) and (3).
The OLS regressions for work experience placements reported in Tables A.4 and A.5 in the Appendix provide very similar findings to the Probit regressions discussed above, even after accounting for class fixed effects. The quality of placements is affected by personal characteristics, and to some extent by counseling from the employment agency, whilst school and class fixed effects as well as school-based counseling do not matter. In accordance with the supply hypothesis, the OLS regressions for the quantity of placements also show that including school fixed effects or class fixed effects increases the explanatory power considerably, which is similar to the results for counseling. The increase is again particularly strong for lower track schools.
3.3 Class-level variation of take-up
The results thus far indicate that take-up of career guidance is strongly supply-driven, and individual level variables play only a minor role. When estimating the effect of the take-up of career guidance on career plans in the next section, one may be concerned about potential endogeneity due to unobserved student differences both affecting the take-up of career guidance and career planning. Therefore, we further investigate the variation in take-up of career guidance between classes. This variation may be driven by supply differences (Borghans et al., 2015), which are unrelated to career planning conditional on the covariates controlled for, or by learning based on the behavior of other students in the class. Factors driving supply differences may involve constraints in the work schedule of counselors, teacher attitudes towards career guidance or randomness in scheduling, time conflicts, and cancellations of career guidance activities. If one is willing to assume an exclusion restriction for career plans conditional on school-fixed effects, these differences between classes could be used as instrumental variables.
Our first approach follows Borghans et al. (2015) and uses average participation in career guidance at the class level as a driver of take-up. This variable is computed as the leave-one-out average of the share of students participating in the respective measure within the class of the student.[9] There is scope for this variable to be a good predictor of actual take-up because there is a lot of variation in take-up across classes which is not explained by our rich set of personal characteristics (similar to the variation across schools reported in Borghans et al. (2015)), see Tables A.2 to A.5 in the appendix. The same supply of career guidance may affect students within a class in the same direction, such that certain options may appear more salient than others, or there may be peer effects.
The leave-one-out class-level averages are good predictors of the individual take-up of counseling in all cases, except that significance is low for the quantity of counseling in the middle track (Table A.12).[10] For the quantity of work experience placements the effect is only significant in the lower track – and no significant effect is found for the quality of placements.
Our second approach uses the within school variation across classes in take-up of career guidance activities. Recall that Tables A.2 to A.5 in the appendix involve stepwise first stage OLS regressions, where columns (1) involve personal characteristics, columns (2) add school fixed effects, and columns (3) add class fixed effects. As discussed above, adding class fixed effects strongly increases the explanatory power (measured by R2), except for the quality of work experience placements. The partial increase in explanatory power due to the school fixed effects and the class fixed effects is stronger for the lower track, while still being sizeable for the middle track. Table A.13 shows that the partial effect of class fixed effects (contrasting columns (2) and (3) in Tables A.2 to A.5) is highly significant in all cases.
Altogether, our further analysis confirms that there is a strong class-level component in all career guidance measures, except for the quality of work experience placements, conditional on school fixed effects. If one is willing to assume an exclusion restriction for career plans, these differences between classes conditional on school-fixed effects could be used as instruments. We will return to this point at the end of the next section.
4 The effect of career guidance on career planning
In this section, we investigate whether career guidance activities improve the state of career planning among students. Our first measure of the advancement of career planning is the probability of reporting a desired occupation. For students who intend to apply for an apprenticeship, being able to state a desired occupation is a signal of improved career planning. Note that students in our sample do not report unrealistic “dream jobs” as their desired occupation. 75 % of the lower track and 58 % of the middle track students report a desired occupation that requires an apprenticeship. The students were asked separately which level of school qualification they think they can achieve and in the vast majority of cases the students’ educational aspirations fit their desired occupations’ required qualification (83 % of lower track and 85 % of middle track students). Thus, even though the high educational aspirations seem unrealistic on average, students appear to have a realistic view about the level of education needed to work in their desired occupation, indicating some realism in career planning.
Our second measure of career planning is the probability of having applied for an apprenticeship. A successful application typically requires a sufficient level of career planning. Additionally, applying for an apprenticeship shows that the students do not avoid making choices but actively make decisions for their future.
Our third measure of career planning is whether students plan to continue general secondary education in the next school year. This usually implies reaching a higher secondary school qualification. A higher secondary school qualification might increase chances to find a more advanced apprenticeship position or even enter tertiary education. Thus planning an upgrading can serve as measure of advanced career planning because it implies knowledge of the apprenticeship labor market. However, it could also imply a lower level of career planning as students might opt to continue schooling to avoid the occupational choice and rather stick to something that they already know, i. e. school (Lavecchia et al., 2016).
Tables 10 and 11 report the average marginal effects of the probit regressions for the three measures of career planning: reporting a desired occupation, applying for apprenticeships and planning to continue school.
Probit Regression: Career Planning for Lower Track Students (Marginal Effects)
| reporting desired occupation | application apprenticeship | continue schooling | |
| Take-up employment agency | 0.165∗∗ | 0.375∗∗∗ | −0.250∗ |
| (0.077) | (0.088) | (0.130) | |
| 2 or more meetings employment agency | −0.157 | −0.143 | 0.134 |
| (0.132) | (0.093) | (0.171) | |
| Take-up school counselor | 0.070 | 0.065 | −0.169 |
| (0.154) | (0.092) | (0.208) | |
| 3 or more meetings school counselor | 0.173∗ | 0.095 | 0.135 |
| (0.092) | (0.098) | (0.158) | |
| 3 or more Work experience pl. | 0.043 | −0.144∗ | 0.040 |
| (0.099) | (0.081) | (0.123) | |
| Work experience pl. in desired occupation | 0.251∗∗∗ | −0.086 | |
| (0.072) | (0.094) | ||
| Female | 0.148∗ | −0.046 | 0.003 |
| (0.080) | (0.091) | (0.094) | |
| 9th grade | 0.193∗∗∗ | −0.320∗∗∗ | 0.227∗∗ |
| (0.058) | (0.074) | (0.114) | |
| German spoken in Family | 0.019 | 0.209 | 0.030 |
| (0.104) | (0.141) | (0.108) | |
| Parents college | 0.085 | −0.007 | −0.002 |
| (0.144) | (0.109) | (0.101) | |
| Parents encourage effort in school | 0.149∗ | −0.110 | 0.330∗∗∗ |
| (0.086) | (0.092) | (0.086) | |
| Parents proud of educ. achievement | −0.001 | −0.032 | 0.095 |
| (0.097) | (0.103) | (0.145) | |
| Ambitious friends | −0.076 | −0.120 | −0.068 |
| (0.059) | (0.084) | (0.098) | |
| Good Math grade | −0.125 | −0.133 | 0.368∗∗∗ |
| (0.090) | (0.108) | (0.102) | |
| Good German grade | −0.069 | −0.077 | 0.208 |
| (0.105) | (0.089) | (0.147) | |
| Openness | −0.092∗∗∗ | −0.014 | 0.036 |
| (0.033) | (0.038) | (0.054) | |
| Extraversion | −0.017 | 0.027 | −0.019 |
| (0.040) | (0.027) | (0.041) | |
| Conscientiousness | 0.035 | 0.051 | 0.051 |
| (0.031) | (0.038) | (0.040) | |
| Neuroticism | −0.015 | −0.037 | 0.076∗ |
| (0.040) | (0.030) | (0.045) | |
| Agreeableness | 0.025 | −0.070∗ | −0.062 |
| (0.037) | (0.036) | (0.057) | |
| external locus of control | −0.007 | 0.070∗ | −0.002 |
| (0.032) | (0.039) | (0.037) | |
| internal locus of control | −0.033 | −0.042 | 0.104 |
| (0.038) | (0.050) | (0.084) | |
| Risk loving | 0.011 | −0.022 | 0.020 |
| (0.012) | (0.018) | (0.018) | |
| pseudo | 0.192 | 0.373 | 0.257 |
| Observations | 159 | 159 | 147 |
Notes: Marginal effects of Probit estimations, controls for city and missing grades included. Standard errors clustered by class in parentheses, ∗
Lower track students meeting with a counselor from the employment agency are more likely to report a desired occupation and to have applied for an apprenticeship, and are less likely to continue schooling. The frequency of the meetings with the employment agency does not show significant effects. Students that met more often with school counselors are also more likely to report a desired occupation. There are no significant effects of school counselors on other career planning measures. Students with at least one work experience placement in their desired occupation are more likely to apply for apprenticeships, while a high number of placements is negatively associated with applying for an apprenticeship, possibly confusing the student’s career planning. He/She might not be ready to apply for an apprenticeship, or perhaps finds it difficult to match his/her preferences with the apprenticeship positions available. Thus, advice by school counselors and adequate placements significantly affect career planning, though not in the same way at the extensive and the intensive margin. Specifically, the employment agency shapes students’ career plan towards the labor market and away from continuing general schooling.
Probit Regression: Career Planning for Middle Track Students (Marginal Effects)
| reporting desired occupation | application apprenticeship | continue schooling | |
| Take-up employment agency | 0.272∗∗∗ | −0.022 | −0.008 |
| (0.075) | (0.106) | (0.136) | |
| 2 or more meetings employment agency | 0.050 | 0.254∗∗ | −0.005 |
| (0.089) | (0.111) | (0.150) | |
| 3 or more Work experience pl. | 0.006 | 0.181∗∗ | −0.199∗∗∗ |
| (0.089) | (0.082) | (0.076) | |
| Work experience pl. in desired occupation | 0.182∗∗ | −0.310∗∗∗ | |
| (0.081) | (0.094) | ||
| Female | 0.176 | −0.155∗ | 0.054 |
| (0.122) | (0.080) | (0.122) | |
| German spoken in Family | −0.021 | −0.312∗∗∗ | 0.248∗ |
| (0.147) | (0.087) | (0.136) | |
| Parents college | −0.000 | −0.096 | 0.270∗∗ |
| (0.058) | (0.113) | (0.128) | |
| Parents encourage effort in school | 0.059 | 0.114 | −0.060 |
| (0.089) | (0.092) | (0.116) | |
| Parents proud of educ. achievement | 0.233∗∗ | 0.067 | −0.137 |
| (0.105) | (0.081) | (0.100) | |
| Ambitious friends | −0.049 | −0.122∗ | 0.181∗∗ |
| (0.058) | (0.071) | (0.088) | |
| Good Math grade | −0.093 | −0.126 | 0.236∗∗ |
| (0.089) | (0.087) | (0.096) | |
| Good German grade | −0.144 | −0.235∗∗ | 0.250∗∗ |
| (0.103) | (0.092) | (0.119) | |
| Openness | −0.053∗ | −0.000 | −0.038 |
| (0.030) | (0.039) | (0.046) | |
| Extraversion | 0.084∗∗ | −0.030 | 0.019 |
| (0.039) | (0.028) | (0.034) | |
| Conscientiousness | −0.051∗ | 0.021 | 0.004 |
| (0.028) | (0.040) | (0.050) | |
| Neuroticism | −0.107∗∗∗ | −0.021 | 0.049 |
| (0.029) | (0.053) | (0.062) | |
| Agreeableness | −0.077 | 0.179∗∗∗ | −0.090 |
| (0.050) | (0.047) | (0.063) | |
| external locus of control | −0.095∗∗∗ | 0.029 | −0.020 |
| (0.033) | (0.035) | (0.045) | |
| internal locus of control | 0.106∗ | −0.040 | 0.029 |
| (0.056) | (0.055) | (0.080) | |
| Risk loving | −0.017 | 0.046∗∗∗ | 0.000 |
| (0.014) | (0.017) | (0.023) | |
| pseudo | 0.201 | 0.316 | 0.287 |
| Observations | 159 | 161 | 153 |
Notes: Marginal effects of Probit estimations, controls for city included. Standard errors clustered by class in parentheses, ∗
Table 11 presents the results for middle track students. Again, we find a positive relationship between career guidance measures and career planning for middle track students, however there are differences between school tracks. Middle track students that met with the employment agency are more likely to report a desired occupation. Students that had more than two meetings with the employment agency are more likely to have applied for apprenticeships. There is no effect of counseling from the employment agency on the plan to continue general schooling. Quantity and quality of work experience placements influences career planning in a very similar way. More than three placements results in a higher probability of having applied for an apprenticeship and a lower probability of planning to continue school. A placement in the desired occupation is highly relevant for career planning because such students are more likely to apply for apprenticeships and are less likely to plan to continue schooling.
There are some noteworthy covariate effects on plans to continue schooling, indicating that middle track students are positively selected. There is a positive effect of counseling on career planning regarding the probability of reporting a desired occupation, and of applying for apprenticeship, in both school tracks. However, we do not find stronger effects of school counselors than of employment agency counselors in the lower track. Counseling shows a slightly negative effect on plans to continue school. A work experience placement in the desired occupation increases the probability of applying for an apprenticeship.
The effect differences by school track can possibly be explained by the difference in the stage of career planning. Middle track students are confronted by the decision to pursue a higher education entrance qualification (if they continue schooling) or the traditional path of starting an apprenticeship. These two options might be clear-cut, and thus career guidance can be more effective by focusing on these two choices. Lower track students are realistically expected to aspire to an apprenticeship, but need to decide whether upgrading their lower track secondary school qualification to a middle track secondary school qualification is beneficial (in order to improve their chances in the apprenticeship labor market). While this is likely to increase the need for career guidance, counseling also has to address various individual barriers to investing in learning and to decision making. Thus, the challenge for school counselors in lower secondary schools is very high, which may explain the lack of significantly positive effects on career planning.
Our results are robust to different specifications of the estimation models as the step-wise addition of control variables and school dummies in Tables A.6 to A.11 shows.
Finally, we provide a short discussion regarding concerns about possible selection in the take-up of career guidance. Based on a priori reasoning, there may be positive or negative selection. On the one hand, students meeting with the employment agency or school counselors may be more motivated, or they may have concrete plans to enter the labor market. In that case, career planning may be more advanced independently of career guidance. On the other hand, students whose state of career planning is less advanced may seek more career guidance or are advised to do so. Hence, the direction of possible bias is not clear.
For two reasons, however, we think that selection in the take-up of counseling and of the quantity of work experience placements is not a major issue for our results. First, our findings in Section 3 provide strong evidence for the supply hypothesis, and for peer effects. At the same time, personal characteristics as drivers of selection on observables play only a minor role. Thus, we consider it unlikely that selection on unobservables is strong. Second, if one is willing to assume an exclusion restriction for class-level drivers of take-up conditional on school-fixed effects, and implements an instrumental variable (IV) approach, the first stage is mostly satisfactory, however the point estimates are very noisy and are not satisfactory. The point estimates typically show the same sign, are mostly larger, and most likely overestimate the effect size.[11] Still, they do not differ significantly from the OLS estimates.
5 Conclusions
This paper analyzes the take-up of career counseling and work experience placements as well as their effects on career planning, based on a survey we conducted in lower and middle track secondary schools in Germany. We distinguish between incidence and quantity of counseling and between quantity and quality of work experience placements.
Career guidance is offered more intensively to students in the lower track than in the middle track, reflecting that career planning is a more pressing issue for lower track students. A key finding is that the incidence of take-up of counseling provided by the employment agency or the school counselor is barely related to individual characteristics, including parental background or grades. Noteworthy exceptions are: lower track students from non-German speaking families are more likely to meet with school counselors frequently and complete three or more work experience placements, and in the middle track low-performing students are more likely to use intensive counseling from the employment agency. Overall, there is only limited evidence that students facing greater difficulties in career planning are more engaged in career guidance. Rather, there are strong differences in take-up of career guidance across schools and classes which are unrelated to the individual characteristics of the students. As a quality measure of work experience placements, we use whether or not students complete a work experience placement in their desired occupation. In contrast to the other activities, this quality indicator is much less affected by school and class effects, and also depends very little on individual characteristics. One noteworthy exception: in the middle track, frequent counseling by the employment agency and own search effort shows a positive effect on the quality of placements, which suggests a positive selection of students with high quality placements.
The second part of our study estimates the effect of career guidance on the state of career planning, measured by whether students report a desired occupation, have applied for apprenticeships, and plan to continue schooling. Our findings show that a higher number of work experience placements improve career planning only in the middle track, where students with at least three work experience placements are more likely to have applied for an apprenticeship. For lower track students, there is an opposite effect. Placements in the preferred occupation are associated with better career planning in both school tracks, a finding which we do not interpret as causal. Furthermore, a higher number of placements show a negative effect on the probability of continuing schooling for middle track students, which is consistent with placements making an apprenticeship more attractive relative to the continuation of schooling. However, the number of placements does not show such an effect for lower track students, i. e. the policy implications of our findings are ambiguous in light of the focus of career guidance on the number of placements. Possibly, lower track students are less ready to apply for an apprenticeship and more placements cannot change that. Schools and counselors are not successful in improving the quality of placements, which rather depends on the students’ own search activities or their family’s support.
For lower track students, frequent counseling provided by school counselors increases the probability of reporting a desired occupation, while counseling provided by the employment agency increases the probability of applying for apprenticeships and of reporting a desired occupation but reduces the probability of planning to continue schooling. Frequent school counseling does not affect the other types of career planning. Middle track students meeting with an employment agency counselor have a higher probability of reporting a desired occupation and frequent meetings increase the probability of applying for an apprenticeship. In sum, the employment agency seems more effective than school counseling in supporting career planning aimed at entering the labor market through an apprenticeship. The employment agency seems to attenuate high educational aspirations, similar to the treatment considered in Goux et al. (2015) for the case of France.
Altogether, our findings suggest that career guidance can improve secondary school students’ career planning. However, the impact differs by school track and type of counseling provider. Clearly, further research on the effects of career guidance is needed, especially in light of the current policy initiatives to expand career guidance in upper secondary schools in Germany (see Bundesagentur für Arbeit, 2018).
Acknowledgment
We gratefully acknowledge financial support by the Baden-Württemberg Stiftung for the research project “Übergänge am Ende der Sekundarstufe I in weiterführende Schulen und die berufliche Bildung” as part of the program “Netzwerk Bildungsforschung (Network Educational Research)”. We thank Holger Bonin and the researchers in the Netzwerk Bildungsforschung for very helpful comments. We are grateful to many school representatives in the cities of Freiburg and Mannheim for their support in undertaking the school survey, which provides the data for our analysis. All errors are our sole responsibility.
Appendix
Descriptive Statistics of the Sample by School Track
| Secondary School Track | sig | |||
| lower track | middle track | |||
| Female | 0.54 | 0.45 | ∗ | |
| City (= Mannheim) | 0.56 | 0.43 | ∗∗ | |
| 9th grade | 0.68 | – | ||
| German spoken in family | 0.81 | 0.94 | ∗∗∗ | |
| At least one parent with college degree | 0.11 | 0.34 | ∗∗∗ | |
| Parents encourage effort in school | 0.65 | 0.63 | ||
| Parents are proud of educational achievement | 0.69 | 0.65 | ||
| Ambitious friends: Many friends strive for upgrading | 0.26 | 0.69 | ∗∗∗ | |
| Good or excellent grade in Math | 0.19 | 0.39 | ∗∗∗ | |
| Good or excellent grade in German | 0.31 | 0.32 | ||
| Grades variable missing | 0.08 | 0.02 | ∗∗ | |
| College degree is achievable | 0.25 | 0.46 | ∗∗∗ | |
| Higher education entrance qualification is achievable | 0.22 | 0.42 | ∗∗∗ | |
| Personality Traits (Big Five, scale 1–7) | ||||
| Conscientiousness | 4.8 | 4.85 | ||
| Extraversion | 4.66 | 4.88 | ||
| Agreeableness | 5.11 | 5.39 | ∗∗ | |
| Neuroticism | 4.18 | 4.06 | ||
| Openness to new experiences | 4.6 | 4.9 | ∗∗ | |
| Locus of Control (scale 1–7) | ||||
| External LOC | 3.28 | 3.17 | ||
| Internal LOC | 5.92 | 5.83 | ||
| Risk aversion (risk averse 0–10 risk loving) | 6.31 | 6.37 | ||
| Application for apprenticeship | 0.3 | 0.34 | ||
| Continue schooling (upgrading qualification) | 0.49 | 0.59 | ∗ | |
| Reporting desired occupation | 0.7 | 0.67 | ||
| Observations | 159 | 161 | ||
Notes: Stat. significant difference ∗
Robustness Check: OLS Regression Take-up of Counseling Including Class/School Dummies – Lower Track
| School counselors | Employment agency | ||||||||
| at least 3 meetings | Take up | at least 2 meetings | |||||||
| (1) | (2) | (3) | (1) | (2) | (3) | (1) | (2) | (3) | |
| Female | −0.037 | −0.116 | −0.028 | −0.042 | −0.099 | −0.089 | −0.025 | −0.086 | −0.080 |
| (0.089) | (0.081) | (0.063) | (0.098) | (0.088) | (0.098) | (0.081) | (0.073) | (0.084) | |
| 9th Grade | −0.154 | −0.260∗∗∗ | −0.006 | −0.128 | −0.114 | −0.032 | −0.241∗∗ | −0.277∗∗∗ | −0.221∗∗ |
| (0.114) | (0.091) | (0.205) | (0.113) | (0.118) | (0.107) | (0.099) | (0.074) | (0.084) | |
| German spoken in family | −0.271∗∗ | −0.257∗∗ | −0.285∗∗∗ | −0.050 | −0.007 | −0.046 | 0.044 | 0.090 | −0.013 |
| (0.117) | (0.093) | (0.090) | (0.115) | (0.126) | (0.133) | (0.081) | (0.099) | (0.096) | |
| Parents college | 0.062 | 0.048 | 0.075 | 0.059 | 0.074 | 0.048 | −0.121 | −0.116 | −0.138 |
| (0.073) | (0.092) | (0.089) | (0.128) | (0.132) | (0.126) | (0.121) | (0.115) | (0.128) | |
| Parents encourage effort in school | −0.088 | −0.099 | −0.080 | −0.062 | −0.039 | −0.006 | −0.071 | −0.061 | −0.056 |
| (0.082) | (0.072) | (0.078) | (0.096) | (0.094) | (0.117) | (0.075) | (0.075) | (0.084) | |
| Parents proud of educ. achievement | 0.211∗∗ | 0.137 | 0.180∗∗ | 0.054 | 0.053 | 0.080 | 0.133 | 0.111 | 0.103 |
| (0.099) | (0.092) | (0.077) | (0.138) | (0.139) | (0.150) | (0.091) | (0.090) | (0.095) | |
| Ambitious friends | 0.076 | −0.010 | 0.081 | −0.023 | −0.042 | 0.006 | −0.011 | −0.051 | −0.066 |
| (0.103) | (0.079) | (0.076) | (0.101) | (0.097) | (0.097) | (0.104) | (0.116) | (0.114) | |
| Good Math grade | −0.087 | −0.135 | −0.113 | −0.087 | −0.076 | −0.137 | −0.029 | −0.042 | −0.056 |
| (0.099) | (0.096) | (0.093) | (0.124) | (0.120) | (0.146) | (0.085) | (0.072) | (0.075) | |
| Good German grade | −0.094 | −0.078 | −0.105 | −0.005 | −0.033 | −0.032 | −0.043 | −0.051 | −0.105 |
| (0.085) | (0.083) | (0.092) | (0.128) | (0.120) | (0.136) | (0.068) | (0.066) | (0.074) | |
| School dummies | yes | yes | yes | ||||||
| Class dummies | yes | yes | yes | ||||||
| Constant | −0.026 | 0.486 | 0.648 | 0.289 | 0.271 | 0.564 | 0.040 | 0.196 | 0.585 |
| (0.305) | (0.427) | (0.407) | (0.469) | (0.408) | (0.429) | (0.446) | (0.340) | (0.347) | |
| 0.184 | 0.340 | 0.464 | 0.100 | 0.184 | 0.325 | 0.171 | 0.264 | 0.425 | |
| Adjusted | 0.069 | 0.217 | 0.305 | −0.029 | 0.031 | 0.115 | 0.052 | 0.126 | 0.247 |
| Observations | 154 | 154 | 154 | 153 | 153 | 153 | 153 | 153 | 153 |
Notes: Same model specification with covariates as in Table 4. Standard errors clustered by class in parentheses, ∗
Robustness Check: OLS Regression Take-up of Counseling Including Class/School Dummies – Middle Track
| Employment agency | ||||||
| Take up | at least 2 meetings | |||||
| (1) | (2) | (3) | (1) | (2) | (3) | |
| Female | −0.086 | −0.103 | −0.067 | −0.017 | −0.005 | −0.013 |
| (0.065) | (0.062) | (0.069) | (0.064) | (0.063) | (0.069) | |
| German spoken in family | 0.124 | 0.107 | 0.010 | −0.028 | 0.066 | 0.011 |
| (0.144) | (0.174) | (0.166) | (0.147) | (0.147) | (0.153) | |
| Parents college | −0.052 | −0.059 | −0.045 | −0.078 | −0.061 | −0.061 |
| (0.079) | (0.080) | (0.088) | (0.083) | (0.080) | (0.089) | |
| Parents encourage effort in school | −0.109 | −0.132 | −0.121 | 0.030 | 0.004 | 0.002 |
| (0.083) | (0.087) | (0.096) | (0.055) | (0.059) | (0.071) | |
| Parents proud of educ. achievement | 0.099 | 0.104∗ | 0.072 | 0.064 | 0.083 | 0.056 |
| (0.059) | (0.060) | (0.051) | (0.082) | (0.071) | (0.080) | |
| Ambitious friends | 0.217∗∗ | 0.214∗∗ | 0.209∗∗ | 0.115 | 0.147∗ | 0.165∗ |
| (0.088) | (0.081) | (0.085) | (0.072) | (0.071) | (0.088) | |
| Good Math grade | 0.068 | 0.060 | 0.050 | −0.101∗∗ | −0.057 | −0.073 |
| (0.075) | (0.064) | (0.066) | (0.048) | (0.046) | (0.054) | |
| Good German grade | 0.093 | 0.087 | 0.108 | −0.045 | −0.008 | 0.036 |
| (0.074) | (0.077) | (0.081) | (0.109) | (0.085) | (0.089) | |
| School dummies | yes | yes | ||||
| Class dummies | yes | yes | ||||
| Constant | −0.402 | −0.461 | −0.573 | 0.544 | 0.187 | 0.063 |
| (0.497) | (0.549) | (0.571) | (0.480) | (0.455) | (0.488) | |
| 0.164 | 0.196 | 0.275 | 0.096 | 0.235 | 0.278 | |
| Adjusted | 0.064 | 0.060 | 0.085 | −0.012 | 0.106 | 0.088 |
| Observations | 160 | 160 | 160 | 160 | 160 | 160 |
Notes: Same model specification with covariates as in Table 5. Standard errors clustered by class in parentheses, ∗
Robustness Check: OLS Regression Participation in Work Experience Placements Including Class/School Dummies – Lower Track
| 3 or more Work experience pl. | Work exp. pl. in desired occup. | |||||
| (1) | (2) | (3) | (1) | (2) | (3) | |
| Female | −0.162∗∗ | −0.081 | −0.072 | 0.248∗∗∗ | 0.285∗∗∗ | 0.279∗∗∗ |
| (0.067) | (0.076) | (0.089) | (0.075) | (0.080) | (0.098) | |
| 9th Grade | −0.200∗ | −0.255∗∗∗ | −0.173 | 0.079 | 0.042 | 0.216 |
| (0.099) | (0.073) | (0.134) | (0.082) | (0.081) | (0.151) | |
| German spoken in family | −0.167∗ | −0.201∗∗ | −0.161∗ | −0.100 | −0.115 | −0.053 |
| (0.085) | (0.091) | (0.092) | (0.092) | (0.094) | (0.101) | |
| Parents college | −0.273∗ | −0.268∗ | −0.288∗ | 0.216 | 0.227 | 0.285 |
| (0.155) | (0.153) | (0.156) | (0.171) | (0.180) | (0.204) | |
| Parents encourage effort in school | 0.073 | 0.066 | 0.047 | 0.096 | 0.096 | 0.110 |
| (0.105) | (0.108) | (0.114) | (0.084) | (0.080) | (0.084) | |
| Parents proud of educ. achievement | 0.070 | 0.075 | 0.033 | −0.017 | −0.007 | 0.005 |
| (0.094) | (0.089) | (0.092) | (0.083) | (0.083) | (0.093) | |
| Ambitious friends | −0.131 | −0.079 | −0.159 | −0.028 | 0.012 | −0.059 |
| (0.098) | (0.101) | (0.118) | (0.107) | (0.111) | (0.131) | |
| Good Math grade | 0.059 | 0.042 | −0.013 | −0.005 | −0.008 | −0.045 |
| (0.111) | (0.103) | (0.125) | (0.123) | (0.121) | (0.166) | |
| Good German grade | 0.059 | 0.089 | 0.138 | 0.092 | 0.100 | 0.163 |
| (0.085) | (0.081) | (0.088) | (0.094) | (0.095) | (0.100) | |
| Take-up employment agency | 0.083 | 0.129∗ | 0.205∗∗ | −0.069 | −0.060 | −0.092 |
| (0.094) | (0.067) | (0.079) | (0.122) | (0.137) | (0.150) | |
| 2 or more meetings employment agency | 0.064 | 0.108 | 0.008 | 0.195∗ | 0.210∗ | 0.314∗∗ |
| (0.124) | (0.123) | (0.146) | (0.108) | (0.108) | (0.148) | |
| Take-up school counselor | −0.021 | −0.029 | −0.009 | 0.089 | 0.111 | 0.076 |
| (0.125) | (0.135) | (0.153) | (0.148) | (0.158) | (0.178) | |
| 3 or more meetings school counselor | 0.031 | 0.034 | −0.017 | 0.054 | 0.058 | 0.092 |
| (0.073) | (0.083) | (0.092) | (0.106) | (0.110) | (0.169) | |
| Own placement search | 0.065 | 0.039 | 0.115 | |||
| (0.116) | (0.122) | (0.158) | ||||
| Placement search family | 0.188∗ | 0.161∗ | 0.105 | |||
| (0.092) | (0.086) | (0.101) | ||||
| Placement search counselor | −0.030 | −0.030 | 0.021 | |||
| (0.151) | (0.165) | (0.186) | ||||
| School dummies | yes | yes | ||||
| Class dummies | yes | yes | ||||
| Constant | 1.201∗∗∗ | 1.102∗∗∗ | 1.179∗∗ | 0.233 | 0.212 | −0.239 |
| (0.362) | (0.360) | (0.424) | (0.382) | (0.383) | (0.426) | |
| 0.169 | 0.262 | 0.340 | 0.216 | 0.242 | 0.295 | |
| Adjusted | 0.028 | 0.109 | 0.116 | 0.061 | 0.065 | 0.032 |
| Observations | 159 | 159 | 159 | 159 | 159 | 159 |
Notes: Same model specification with covariates as in Table 8. Standard errors clustered by class in parentheses, ∗
Robustness Check: OLS Regression Participation in Work Experience Placements Including Class/School Dummies – Middle Track
| 3 or more Work experience pl. | Work exp. pl. in desired occup. | |||||
| (1) | (2) | (3) | (1) | (2) | (3) | |
| Female | 0.027 | −0.003 | 0.012 | 0.107 | 0.104 | 0.088 |
| (0.073) | (0.075) | (0.077) | (0.085) | (0.083) | (0.082) | |
| German spoken in family | 0.060 | 0.041 | −0.013 | −0.038 | −0.076 | −0.136 |
| (0.129) | (0.119) | (0.127) | (0.154) | (0.174) | (0.211) | |
| Parents college | 0.045 | 0.015 | 0.023 | −0.168∗ | −0.196∗∗ | −0.204∗∗ |
| (0.068) | (0.069) | (0.064) | (0.094) | (0.090) | (0.088) | |
| Parents encourage effort in school | −0.092 | −0.062 | −0.019 | −0.009 | −0.025 | −0.080 |
| (0.067) | (0.073) | (0.054) | (0.086) | (0.093) | (0.096) | |
| Parents proud of educ. achievement | −0.080 | −0.079 | −0.090 | 0.125 | 0.133 | 0.098 |
| (0.070) | (0.070) | (0.069) | (0.091) | (0.089) | (0.105) | |
| Ambitious friends | −0.023 | −0.056 | −0.027 | 0.053 | 0.071 | 0.109 |
| (0.106) | (0.105) | (0.108) | (0.085) | (0.092) | (0.107) | |
| Good Math grade | −0.006 | −0.009 | −0.041 | −0.086 | −0.080 | −0.066 |
| (0.066) | (0.064) | (0.080) | (0.088) | (0.091) | (0.093) | |
| Good German grade | −0.017 | −0.002 | −0.027 | −0.143 | −0.187∗ | −0.185 |
| (0.049) | (0.055) | (0.069) | (0.099) | (0.103) | (0.108) | |
| Take-up employment agency | −0.094 | −0.106 | −0.142∗ | −0.014 | −0.060 | −0.061 |
| (0.074) | (0.084) | (0.077) | (0.097) | (0.095) | (0.099) | |
| 2 or more meetings employment agency | 0.101 | 0.123 | 0.193∗ | −0.075 | −0.032 | −0.044 |
| (0.079) | (0.086) | (0.107) | (0.085) | (0.082) | (0.089) | |
| Own placement search | 0.178∗ | 0.197∗ | 0.199∗ | |||
| (0.098) | (0.102) | (0.107) | ||||
| Placement search family | 0.086 | 0.122 | 0.158 | |||
| (0.098) | (0.106) | (0.106) | ||||
| School dummies | yes | yes | ||||
| Class dummies | yes | yes | ||||
| Constant | −0.377 | −0.278 | −0.196 | 0.426 | 0.695 | 0.389 |
| (0.418) | (0.490) | (0.499) | (0.394) | (0.423) | (0.479) | |
| 0.108 | 0.148 | 0.310 | 0.150 | 0.191 | 0.259 | |
| Adjusted | −0.012 | −0.010 | 0.116 | 0.022 | 0.026 | 0.036 |
| Observations | 161 | 161 | 161 | 161 | 161 | 161 |
Notes: Same model specification with covariates as in Table 9. Standard errors clustered by class in parentheses, ∗
Robustness Check: Stepwise Probit Regression Effect of Career Guidance on Reporting a Desired Occupation – Lower Track (Marginal Effects)
| (1) | (2) | (3) | (4) | (5) | (6) | |
| Take-up employment agency | 0.107∗ | 0.107∗ | 0.138∗ | 0.147∗ | 0.166∗∗ | 0.192∗∗∗ |
| (0.065) | (0.065) | (0.072) | (0.080) | (0.077) | (0.075) | |
| 2 or more meetings employment agency | −0.125 | −0.124 | −0.122 | −0.125 | −0.162 | −0.119 |
| (0.128) | (0.128) | (0.146) | (0.141) | (0.136) | (0.127) | |
| Take-up school counselor | 0.121 | 0.120 | 0.032 | 0.073 | 0.058 | 0.001 |
| (0.133) | (0.131) | (0.144) | (0.142) | (0.157) | (0.153) | |
| 3 or more meetings school counselor | 0.095 | 0.095 | 0.172∗∗ | 0.143∗ | 0.185∗∗ | 0.188∗ |
| (0.082) | (0.081) | (0.082) | (0.085) | (0.092) | (0.107) | |
| 3 or more Work experience pl. | −0.007 | 0.021 | 0.035 | 0.037 | 0.028 | |
| (0.081) | (0.096) | (0.099) | (0.100) | (0.113) | ||
| Gender, family background, peers | yes | yes | yes | yes | ||
| Grades | yes | yes | yes | |||
| Personality traits | yes | yes | ||||
| School dummies | yes | |||||
| Pseudo | 0.034 | 0.034 | 0.108 | 0.134 | 0.190 | 0.227 |
| Observations | 159 | 159 | 159 | 159 | 159 | 159 |
Notes: Marginal effects of Probit estimations, same model specification with covariates as in Table 10. Standard errors clustered by class in parentheses, ∗
Robustness Check: Stepwise Probit Regression Effect of Career Guidance on Applying for Apprenticeship – Lower Track (Marginal Effects)
| (1) | (2) | (3) | (4) | (5) | (6) | |
| Take-up employment agency | 0.284∗∗∗ | 0.316∗∗∗ | 0.366∗∗∗ | 0.376∗∗∗ | 0.398∗∗∗ | 0.403∗∗∗ |
| (0.084) | (0.082) | (0.066) | (0.066) | (0.084) | (0.090) | |
| 2 or more meetings employment agency | −0.004 | −0.030 | −0.135 | −0.141 | −0.183∗ | −0.145 |
| (0.132) | (0.131) | (0.114) | (0.112) | (0.106) | (0.100) | |
| Take-up school counselor | −0.051 | −0.067 | −0.026 | −0.000 | 0.044 | 0.045 |
| (0.075) | (0.083) | (0.092) | (0.090) | (0.106) | (0.102) | |
| 3 or more meetings school counselor | 0.158 | 0.166 | 0.161 | 0.131 | 0.144 | 0.088 |
| (0.108) | (0.113) | (0.104) | (0.102) | (0.096) | (0.106) | |
| 3 or more Work experience pl. | −0.121 | −0.149∗∗ | −0.127∗ | −0.176∗∗ | −0.142∗ | |
| (0.080) | (0.069) | (0.073) | (0.074) | (0.077) | ||
| Work experience pl. in desired occupation | 0.189∗∗∗ | 0.247∗∗∗ | 0.233∗∗∗ | 0.278∗∗∗ | 0.245∗∗∗ | |
| (0.071) | (0.065) | (0.063) | (0.070) | (0.076) | ||
| Gender, family background, peers | yes | yes | yes | yes | ||
| Grades | yes | yes | yes | |||
| Personality traits | yes | yes | ||||
| School dummies | yes | |||||
| Pseudo | 0.106 | 0.144 | 0.254 | 0.280 | 0.347 | 0.372 |
| Observations | 159 | 159 | 159 | 159 | 159 | 159 |
Notes: Marginal effects of Probit estimations, same model specification with covariates as in Table 10. Standard errors clustered by class in parentheses, ∗
Robustness Check: Stepwise Probit Regression Effect of Career Guidance on Continue Schooling – Lower Track (Marginal Effects)
| (1) | (2) | (3) | (4) | (5) | (6) | |
| Take-up employment agency | −0.181∗∗ | −0.197∗∗ | −0.214∗∗ | −0.210∗ | −0.250∗ | −0.286∗ |
| (0.088) | (0.091) | (0.096) | (0.121) | (0.131) | (0.153) | |
| 2 or more meetings employment agency | 0.006 | 0.008 | 0.050 | 0.074 | 0.134 | 0.150 |
| (0.153) | (0.161) | (0.163) | (0.164) | (0.172) | (0.190) | |
| Take-up school counselor | −0.014 | −0.005 | −0.048 | −0.125 | −0.167 | −0.153 |
| (0.129) | (0.142) | (0.159) | (0.174) | (0.212) | (0.190) | |
| 3 or more meetings school counselor | 0.036 | 0.031 | 0.052 | 0.150 | 0.132 | 0.132 |
| (0.129) | (0.132) | (0.145) | (0.155) | (0.159) | (0.142) | |
| 3 or more work experience pl. | 0.108 | 0.086 | 0.045 | 0.041 | 0.022 | |
| (0.109) | (0.117) | (0.126) | (0.123) | (0.136) | ||
| work experience pl. in desired occupation | −0.061 | −0.097 | −0.102 | −0.089 | −0.105 | |
| (0.073) | (0.087) | (0.087) | (0.097) | (0.120) | ||
| Gender, family background, peers | yes | yes | yes | yes | ||
| Grades | yes | yes | yes | |||
| Personality traits | yes | yes | ||||
| School dummies | yes | |||||
| Pseudo | 0.022 | 0.031 | 0.125 | 0.199 | 0.256 | 0.310 |
| Observations | 147 | 147 | 147 | 147 | 147 | 147 |
Notes: Marginal effects of Probit estimations, same model specification with covariates as in Table 10. Standard errors clustered by class in parentheses, ∗
Robustness Check: Stepwise Probit Regression Effect of Career Guidance on Reporting a Desired Occupation – Middle Track (Marginal Effects)
| (1) | (2) | (3) | (4) | (5) | (6) | |
| Take-up employment agency | 0.156∗∗ | 0.152∗∗ | 0.145∗ | 0.153∗∗ | 0.273∗∗∗ | 0.262∗∗∗ |
| (0.074) | (0.075) | (0.075) | (0.075) | (0.079) | (0.073) | |
| 2 or more meetings employment agency | 0.041 | 0.044 | 0.043 | 0.034 | 0.052 | 0.058 |
| (0.091) | (0.092) | (0.091) | (0.099) | (0.090) | (0.096) | |
| 3 or more work experience pl. | −0.049 | −0.032 | −0.037 | 0.007 | 0.044 | |
| (0.068) | (0.072) | (0.076) | (0.089) | (0.100) | ||
| Gender, family background, peers | yes | yes | yes | yes | ||
| Grades | yes | yes | yes | |||
| Personality traits | yes | yes | ||||
| School dummies | yes | |||||
| Pseudo | 0.022 | 0.024 | 0.058 | 0.060 | 0.201 | 0.242 |
| Observations | 159 | 159 | 159 | 159 | 159 | 159 |
Notes: Marginal effects of Probit estimations, same model specification with covariates as in Table 11. Standard errors clustered by class in parentheses, ∗
Robustness Check: Stepwise Probit Regression Effect of Career Guidance on Applying for Apprenticeship – Middle Track (Marginal Effects)
| (1) | (2) | (3) | (4) | (5) | (6) | |
| Take-up employment agency | −0.088 | −0.069 | −0.023 | 0.009 | −0.035 | −0.012 |
| (0.107) | (0.113) | (0.103) | (0.107) | (0.113) | (0.083) | |
| 2 or more meetings employment agency | 0.237∗∗ | 0.256∗∗ | 0.233∗∗ | 0.210∗ | 0.261∗∗ | 0.252∗∗ |
| (0.100) | (0.102) | (0.105) | (0.113) | (0.116) | (0.113) | |
| 3 or more work experience pl. | 0.164∗∗∗ | 0.228∗∗∗ | 0.228∗∗∗ | 0.166∗∗ | 0.229∗∗∗ | |
| (0.061) | (0.064) | (0.065) | (0.081) | (0.069) | ||
| work experience pl. in desired occupation | 0.217∗∗∗ | 0.188∗∗∗ | 0.167∗∗ | 0.160∗∗ | 0.146∗ | |
| (0.071) | (0.065) | (0.070) | (0.080) | (0.075) | ||
| Gender, family background, peers | yes | yes | yes | yes | ||
| Grades | yes | yes | yes | |||
| Personality traits | yes | yes | ||||
| School dummies | yes | |||||
| Pseudo | 0.031 | 0.090 | 0.167 | 0.220 | 0.303 | 0.399 |
| Observations | 161 | 161 | 161 | 161 | 161 | 161 |
Notes: Marginal effects of Probit estimations, same model specification with covariates as in Table 11. Standard errors clustered by class in parentheses, ∗
Robustness Check: Stepwise Probit Regression Effect of Career Guidance on Continue Schooling – Middle Track (Marginal Effects)
| (1) | (2) | (3) | (4) | (5) | (6) | |
| Take-up employment agency | 0.090 | 0.059 | 0.020 | −0.021 | −0.002 | 0.073 |
| (0.117) | (0.125) | (0.115) | (0.126) | (0.138) | (0.130) | |
| 2 or more meetings employment agency | −0.064 | −0.083 | −0.061 | −0.002 | −0.012 | −0.182 |
| (0.114) | (0.124) | (0.141) | (0.158) | (0.149) | (0.161) | |
| 3 or more work experience pl. | −0.174∗ | −0.224∗∗ | −0.212∗∗ | −0.194∗∗∗ | −0.178∗∗ | |
| (0.098) | (0.092) | (0.090) | (0.073) | (0.088) | ||
| work experience pl. in desired occupation | −0.349∗∗∗ | −0.322∗∗∗ | −0.295∗∗∗ | −0.299∗∗∗ | −0.311∗∗∗ | |
| (0.080) | (0.081) | (0.089) | (0.096) | (0.088) | ||
| Gender, family background, peers | yes | yes | yes | yes | ||
| Grades | yes | yes | yes | |||
| Personality traits | yes | yes | ||||
| School dummies | yes | |||||
| Pseudo | 0.005 | 0.108 | 0.205 | 0.267 | 0.284 | 0.372 |
| Observations | 153 | 153 | 153 | 153 | 153 | 153 |
Notes: Marginal effects of Probit estimations, same model specification with covariates as in Table 11. Standard errors clustered by class in parentheses, ∗
Robustness Check: First Stage Probit Regression – Counseling Instrumented by Class Averages of Participation (Marginal Effects)
| Employment Agency | school counselors | 3 or more work exp. pl. | work. exp. pl. desired occ. | |||
| Take up | 2 or more times | Take up | 3 or more times | |||
| Lower Track | ||||||
| IV: class average in participation | 0.576∗∗∗ | 0.626∗∗∗ | 0.375∗∗∗ | 0.938∗∗∗ | 0.468∗∗∗ | −0.475 |
| (0.216) | (0.194) | (0.071) | (0.191) | (0.147) | (0.439) | |
| Observations | 154 | 154 | 154 | 154 | 154 | 154 |
| Middle Track | ||||||
| IV: class average in participation | 0.538∗∗∗ | 0.332∗ | 0.171 | −0.283 | ||
| (0.179) | (0.177) | (0.258) | (0.332) | |||
| Observations | 161 | 161 | 161 | 161 | ||
Notes: Marginal effects of Probit estimations. Controlled for gender, city, 9th grade, parents’ background and support, friends, grades, grades missing, personality traits. Standard errors clustered by class in parentheses, ∗
Robustness Check: Joint Significance of Class Dummies in Estimations of Career Guidance Participation
| Employment agency | School counselor | 3 or more work exp. pl. | work exp. pl. desired occ. | |||
| Take up | 2 or more times | Take up | 3 or more times | |||
| Lower Track | ||||||
| p-value | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 |
| Observations | 159 | 159 | 159 | 159 | 159 | 159 |
| Middle Track | ||||||
| p-value | 0.00 | 0.00 | 0.00 | 0.00 | ||
| Observations | 161 | 161 | 161 | 161 | 161 | 161 |
Notes:
References
Bernardi, M., M. Bratti, and G. De Simone. 2014. “I Wish I Knew ..; – Misperceived Ability, School Track Counseling Services and Performances in Upper Secondary Education.” IZA Discussion Papers 7940, Institute for the Study of Labor (IZA). 10.2139/ssrn.2396435Search in Google Scholar
Bettinger, E. P., B. T. Long, P. Oreopoulos, and L. Sanbonmatsu. 2012. “The Role of Application Assistance and Information in College Decisions: Results from the H&R Block Fafsa Experiment.” The Quarterly Journal of Economics 127:1205–1242. 10.1093/qje/qjs017Search in Google Scholar
Biewen, M., and M. Tapalaga. 2017. “Life-cycle Educational Choices in a System with Early Tracking and ‘Second Chance’ options.” Economics of Education Review 56:80–94. 10.1016/j.econedurev.2016.11.008Search in Google Scholar
BMBF. 2018. “Berufsbildungsbericht 2018.” Tech. Rep., Bundesministerium für Bildung und Forschung. Search in Google Scholar
Bonin, H., B. Fitzenberger, and A. Hillerich. 2016. “Schule–Berufsausbildung–Arbeitsmarkt: Herausforderungen und Potenziale der ökonomischen Berufsbildungsforschung.” Perspektiven der Wirtschaftspolitik 17:208–231. 10.1515/pwp-2016-0019Search in Google Scholar
Boockmann, B., and S. Nielen. 2016. “Mentoring Disadvantaged Youth During School-to-Work Transition: Evidence from Germany.” IAW Discussion Paper 123, Institut für Angewandte Wirtschaftsforschung. Search in Google Scholar
Borghans, L., B. H. H. Golsteyn, and A. Stenberg. 2015. “Does Expert Advice Improve Educational Choice?” PLoS ONE 10, e0145378. Search in Google Scholar
Bundesagentur für Arbeit. 2018. “Lebensbegleitende Berufsberatung – Flächendeckende Einführung der “Berufsberatung vor dem Erwerbsleben.” Weisung 201810017, Bundesagentur für Arbeit. Download on 23 July 2019. https://con.arbeitsagentur.de/prod/apok/ct/dam/download/documents/Weisung-201810017_ba021962.pdf. Search in Google Scholar
Carell, S., and B. Sacerdote. 2017. “Why Do College-Going Interventions Work?” American Economic Journal: Applied Economics 9:124–151. 10.3386/w19031Search in Google Scholar
DellaVigna, S. 2009. “Psychology and Economics: Evidence from the Field.” Journal of Economic Literature 47:315–372. 10.3386/w13420Search in Google Scholar
Ehlert, M., C. Finger, A. Rusconi, and H. Solga. 2017. “Applying to College: Do Information Deficits Lower the Likelihood of College-eligible Students from Less-privileged Families to Pursue Their College Intentions?: Evidence from a Field Experiment.” Social Science Research 67:193–212. 10.1016/j.ssresearch.2017.04.005Search in Google Scholar
Fitzenberger, B., and S. Licklederer. 2015. “Career Planning, School Grades, and Transitions: The Last Two Years in a German Lower Track Secondary School.” Journal of Economics and Statistics (Jahrbuecher fuer Nationaloekonomie und Statistik) 235:433–458. 10.1515/9783110510805-006Search in Google Scholar
Fitzenberger, B., and S. Licklederer. 2017. “Additional Career Assistance and Educational Outcomes for Students in Lower Track Secondary Schools.” ZEW Discussion Paper 17-024, Centre for European Economic Research, Mannheim. 10.2139/ssrn.3010660Search in Google Scholar
Golsteyn, B. H., H. Grönqvist, and L. Lindahl. 2014. “Adolescent Time Preferences Predict Lifetime Outcomes.” The Economic Journal 124:F739–F761. 10.1111/ecoj.12095Search in Google Scholar
Goux, D., M. Gurgand, and E. Maurin. 2015. “Adjusting Your Dreams? High School Plans and Dropout Behaviour.” The Economic Journal 127:1025–1046. 10.1111/ecoj.12317Search in Google Scholar
Hanushek, E., G. Schwerdt, L. Woessmann, and L. Zhang. 2017. “General Education, Vocational Education, and Labor-Market Outcomes over the Life-Cycle.” Journal of Human Resources 52:48–87. 10.3386/w17504Search in Google Scholar
Harmon, C., V. Hogan, and I. Walker. 2003. “Dispersion in the Economic Return to Schooling.” Labour Economics 10:205–214. 10.1016/S0927-5371(03)00003-4Search in Google Scholar
Heckman, J. J., J. E. Humphries, and G. Veramendi. 2018. “Returns to Education: The Causal Effects of Education on Earnings, Health, and Smoking.” Journal of Political Economy 126:197–246. 10.3386/w22291Search in Google Scholar
Hillage, J., J. Kodz, and G. Pike. 2001. “Pre-16 Work Experience Practice in England: An Evaluation.” Research Report 263, UK Department for Education and Employment. Search in Google Scholar
Hoest, A., V. M. Jensen, and L. P. Nielsen. 2013. “Increasing the Admission Rate to Upper Secondary School: the Case of Lower Secondary School Student Career Guidance.” Education Economics 21:213–229. 10.1080/09645292.2013.789825Search in Google Scholar
Kerr, S., T. Pekkarinen, M. Sarvimäki, and R. Uusitalo. 2015. “Post-Secondary Education and Information on Labor Market Prospects: A Randomized Field Experiment.” IZA Discussion Paper, 9372. Search in Google Scholar
Koch, A., J. Nafziger, and H. S. Nielsen. 2015. “Behavioral Economics of Education.” Journal of Economic Behavior & Organization 115:3–17. 10.1016/j.jebo.2014.09.005Search in Google Scholar
Kohlrausch, B., and H. Solga. 2012. “Übergänge in die Ausbildung: Welche Rolle Spielt die Ausbildungsreife?” Zeitschrift für Erziehungswissenschaft 15:753–773. 10.1007/s11618-012-0332-6Search in Google Scholar
Lavecchia, A., H. Liu, and P. Oreopoulos. 2016. “Behavioral Economics of Education: Progress and Possibilities.” In Handbook of the Economics of Education, edited by S. M. Eric A. Hanushek and L. Woessmann. Vol. 5 of Handbook of the Economics of Education, 1–74. Elsevier. 10.1016/B978-0-444-63459-7.00001-4Search in Google Scholar
Mann, A., and C. Percy. 2014. “Employer Engagement in British Secondary Education: Wage Earning Outcomes Experienced by Young Adults.” Journal of Education and Work 27:496–523. 10.1080/13639080.2013.769671Search in Google Scholar
McNally, S. 2016. “How Important is Career Information and Advice?” IZA World of Labor 317:1–10. 10.15185/izawol.317Search in Google Scholar
Messer, D. 2018. “Work Placements at 14–15 Years and Employability Skills.” Education + Training 60:16–26. 10.1108/ET-11-2016-0163Search in Google Scholar
Neumark, D., and D. Rothstein. 2006. “School-to-career Programs and Transitions to Employment and Higher Education.” Economics of Education Review 25:374–393. school-to-Work and Educational Reform Symposium. 10.3386/w10060Search in Google Scholar
Neumark, D., and D. Rothstein. 2007. “Do School-to-Work Programs Help the “Forgotten Half”?” In Improving School-to-Work Transitions, edited by D. Neumark, 87–133. Russell Sage Foundation. Search in Google Scholar
Peter, F. H., and V. Zambre. 2017. “Intended College Enrollment and Educational Inequality: Do Students Lack Information?” Economics of Education Review 60:125–141. 10.1016/j.econedurev.2017.08.002Search in Google Scholar
Rodríguez-Planas, N. 2012. “Longer-Term Impacts of Mentoring, Educational Services, and Learning Incentives: Evidence from a Randomized Trial in the United States.” American Economic Journal: Applied Economics 4:121–139. 10.1257/app.4.4.121Search in Google Scholar
Saniter, N., D. D. Schnitzlein, and T. Siedler. 2019. “Occupational Knowledge and Educational Mobility: Evidence from the Introduction of Job Information Centers.” Economics of Education Review 69:108–124. 10.1016/j.econedurev.2018.12.009Search in Google Scholar
Schröder, R. 2015. “Reformen zur Berufsorientierung auf Bundes- und Landesebene im Zeitraum 2004–2015.” Tech. Rep., Bertelsmann Stiftung. Search in Google Scholar
Solga, H., and B. Kohlrausch. 2013. “How Low-achieving German Youth Beat the Odds and Gain Access to Vocational Training—Insights from Within-Group Variation.” European Sociological Review 29:1068–1082. 10.1093/esr/jcs083Search in Google Scholar
Code and Datasets
The author(s) published code and data associated with this article in the ZBW Journal Data Archive, a storage platform for datasets. See: http://doi.org/10.15456/ger.2019303.132443.
© 2020 Walter de Gruyter GmbH, Berlin/Boston
