Dropout prevention measures in the Netherlands, an explorative evaluation

Abstract

In line with the Lisbon Agenda, set by the European Council in the year 2000, European governments formulated ambitious plans to halve the level of early school-leavers by 2012. This paper outlines the dropout prevention measures in the Netherlands and analyzes their influence at both the individual and school level. While most policy measures correlate negatively with the individual dropout decision, only “mentoring and coaching” (i.e., matching of students with a coach from public or private organizations), “optimal track or profession” (e.g., work placement) and “dual track” (i.e., re-entering education for dropout students) have a significant negative impact on the individual dropout decision. By means of quantile regressions, we observe that schools with a relatively high dropout rate benefit the most from dropout prevention measures.

Keywords:

• school dropout prevention
• secondary education
• probit
• quantile regression

Acknowledgements

We would like to thank Wim Groot, Henriette Maassen van den Brink, Chris van Klaveren, Marton Csillag, participants of the TIER seminar at the University of Groningen, two anonymous referees, and members of the feedback committee at the Dutch Ministry of Education, Culture and Sciences and The Scientific Review Commission at NICIS Institute for insightful comments. We are grateful to the Dutch Ministry of Education for providing the data. The authors acknowledge financial support of NICIS. The usual caveat applies.

Notes

1. On 1 August 2008, a total of 20 schools were selected for a government experiment. If students were going to school in a selected school, they had the possibility to attend classes in their pre-vocational school. As such, they did not have to switch from a pre-vocational to a senior vocational school. Most schools were located in Amsterdam and Rotterdam (rijksoverheid.nl).

2. The decentralized policy is accommodated with a large accountability. Well performing schools and regions are “named”, while poorly performing regions and schools are “shamed”. The naming and shaming incentive is implemented by the website http://www.aanvalopschooluitval.nl.

3. In this paper, truancy and school absenteeism are used interchangeably.

4. The authors mainly use descriptive statistics and multiple regressions to point out that average class size and ethnic minority groups are responsible for about 56% of the variance in the observed truancy rate.

5. Attwood and Croll (2006) have used the British Household Panel Survey and in-depth interviews to ask persistent truants about the extent, consequences and explanations for truancy at secondary schools. Poor relationships with teachers, bullying and a more general dislike of the school’s atmosphere are considered as triggers for the dropout decision. Attwood and Croll (2006) suggest a distinction between socio-economic and attitudinal factors. Davis and Lee (2004) also adhere to above findings. They went into discussion with truants, as well as attendees and some parents. Davis and Lee (2004) add to the discussion that, in contrast to professionals, the study curriculum is not considered as a dropout trigger. This finding has been weakened by Beekhoven and Dekkers (2005) who put emphasis on learning problems, lack of motivation and problems arising from choosing the wrong vocational track.

6. In line with the European definition, senior vocational education is denoted by ISCED (International Standard Classification of Education) 3.

7. On 1 August 2008, a total of 20 schools were selected for a government experiment. If students were going to school in a selected school, they had the possibility to attend classes in their pre-vocational school. As such, they did not have to switch from a pre-vocational to a senior vocational school. Most schools were located in Amsterdam and Rotterdam (rijksoverheid.nl).

8. We realize a general accepted definition of dropout (in all its nuances of, for example, event and status dropouts) does not exist. It is beyond the scope of this paper to discuss the various definitions, however, we follow the “official” European and Dutch definition.

9. Some of these variables (e.g., school type) are endogenous with respect to the risk of dropping out. An instrumental variable (IV) approach would be appropriate to deal with this endogeneity. However, due to data restrictions, the appropriate instruments are as yet unavailable. We consider the endogeneity issue as scope for further research.

10. Extensive additional robustness checks were performed. First, some measures may have been implemented in very few regions. Their effects may be confounded with autonomous time trends in these regions. Given that dropout rates have been falling substantially in recent years, we included the interaction effect of the region dummy with the time trend into the regression model: if the autonomous time trend in the fall in dropout rate differs between regions, between region differences in time trends may be confounded with effects of certain measures that were implemented in these regions. The analysis delivered robust results. Second, some measures are implemented by almost all regions. Both removing these measures, and clustering of the measures at a higher level, delivered robust outcomes.

11. Using an ordinary least squares regression model could yield probabilities outside the (0.1) boundary. As an alternative econometric technique multi-level modeling could be employed. We consider this as scope for further research.

12. Careful exploration of the data does not show a significant difference in measures used by early implementers and the other regions.

13. Note, however, that there might also be unobserved heterogeneity at the school level. This unobserved heterogeneity might arise from a different source.