ABSTRACT
Exploiting data from Ghanaian schools’ eighth grade students collected in 2011, we estimate the causal effects of school bullying on academic achievement and gender-based mitigating approaches by using propensity score matching (PSM) and doubly robust (DR) estimator approach. We find that students victimized by bullying score at least 0.22 standard deviation lower than their peers in a standardized mathematics examination. Meanwhile, we document that the effect of bullying is significantly attenuated in the presence of female teachers in the classroom. These results hold through a set of robustness checks including placebo regressions and matching quality test. We explain the results through gender difference in teaching paradigm and conclude that a feminine management approach in class is required to reduce the effect of bullying.
Acknowledgements
The authors thank seminar participants at AAEA, USAID. Xu would like to thank for the financial support from the National Natural Science Foundation of China (71203145). All errors are our own responsibility.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 Olweus pioneers in the systematical study on school bullying in the 1970s. Olweus (Citation1993) defines a student being bullied at school ‘when he or she is exposed, repeatedly and over time, to negative actions on the part of one or more other students’. These negative actions include to attack or discomfort someone physically or verbally, spreading rumours, and intentionally excluding someone from a group.
2 A detailed review of the plausible values methodology is given in Mislevy (Citation1991).
3 These home support indicators include computer possession, study desk, having their own room, internet accessibility, and number of books at home.
4 We acknowledge the suggestion from an anonymous referee. The impact of bullying may also be affected by the share of classmates also being bullied. We define large or small share by the median percentage (55%) of classmates being bullied. But we find insignificant differential impact of the share of classmates also being bullied.