The effect of an academic dismissal policy on dropout, graduation rates and student satisfaction. Evidence from the Netherlands

Abstract

This paper examines the effect of the introduction of an academic dismissal (AD) policy (i.e. an intervention, which can lead to compulsory student withdrawal) on student dropout, student graduation rates and satisfaction with the study program. Using a difference-in-differences type of estimator, we compare programs that introduced an AD policy with a control group of programs which did not employ an AD policy. The robustness of the results is tested by a propensity score matching. The outcomes suggest that the implementation of an AD policy results in a higher first-year dropout rate and a higher student graduation rate. The results also indicate that, on average, student satisfaction decreases due to the introduction of an AD policy, while student satisfaction regarding program feasibility increases when an AD policy is employed.

Keywords:

• higher education
• admission policy
• compulsory student withdrawal
• student satisfaction
• school dropout
• student graduation

JEL Classification:

• I21
• I22

Acknowledgements

The authors would like to thank participants from the 2013 TIER lunch seminars, Studiekeuze123 seminar, seminar at the Dutch Ministry of Education, Wim Groot, Henriëtte Maassen van den Brink, Ivo Arnold, Don Westerheijden, Tommaso Agasisti, Kees Boele, Sofie Cabus, Joris Ghysels, and Chris Van Klaveren for their useful comments on previous versions of this article. We gratefully acknowledge the data support received by Studiekeuze123, especially by Bram Enning and Constance Dutmer.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 We use the terms ‘success’ and ‘graduation’ interchangeably throughout the paper.

2 Note that the performance requirements of the AD policy also can be qualitative, such as passing essential courses (Mennen 2013). However, we will not focus on this type of AD policy, as it is difficult to measure and interpret this kind of performance requirements.

3 Both figures make use of information extracted from our data. Since we only have information as of 2003, we are not able to provide a visual representation of the implementation process and the use of credit thresholds before 2003.

4 Note that graduation and dropout are complex concepts which are influenced by other factors than that discussed above. Because we want to keep the literature section to the point, we only discuss the factors we used in our study.

5 When the interpolation resulted in values larger than 100 percent or smaller than zero percent for the performance variables, this value is replaced by 100 percent or zero percent, respectively. Further, we excluded programs that were terminated in 2010 or 2011, since this event can influence student success and student dropout of previous years.

6 In the NSE of 2007, the central question was changed to ‘give your overall judgment about the program'. An analysis of the density functions of both questions shows that they are comparable.

7 Note that since programs which already implemented an AD policy before 2003–04 do not influence the regression specification, we exclude these programs from the analysis (it should be noted that including those observations results in very comparable outcomes, available upon request).

Table A1. Common trend assumption when AD policy is introduced in 2005–2006 for the different outcome variables.

8 Note that the total sample consists of 6012 study programs (see Section 4.3.1). For clarity, in the descriptive statistics we present only the number of unique programs. Moreover, when a study program appears every year in the data set, it is considered in the descriptive statistics as one unique program, while in the total sample it is counted as six observations (i.e. one for every year).

9 A factor analysis points out that student dropout and student graduation are no substitutes (uniqueness of the estimates is .9455) and thus should be included separately in the analyses.