The willingness to pay to reduce school bullying
Highlights
► We estimate the societal willingness (WTP) to pay to reduce school bullying. ► We estimate mean marginal WTP to 5.95–8.48 Swedish kronor (€0.66–0.95). ► We also calculate the aggregate societal WTP per reduced victim of bullying (VSBV). ► The VSBV is estimated to 585,090–835,280 Swedish kronor (€65,446–93,431). ► Results can be used for e.g. economic evaluations of antibullying programs.
Introduction
A substantial share of children and adolescents attend a school where they are victims of direct and/or indirect bullying (Beaty & Alexeyev, 2008). Direct bulling includes teasing, taunting, threatening, hitting or name-calling, while examples of indirect bullying are leaving others out on purpose and spreading vicious rumors. Using a national representative sample of Swedish adolescents aged 15 in 2005/06, approximately 5% of the boys and 3% of the girls stated that they had been bullied repeatedly during the previous few months. These estimates are similar to earlier national studies dating back to 1997/98 (6% of boys and 4% of girls bullied) and 1993/1994 (5% of boys and 5% of girls bullied). The share of adolescents stating that they had been bullied during the last school year was higher, 11–14% in 2005/2006 (Danielson, 2006, Danielson and Sundbaum, 2003). There is more substantial variance in the international prevalence estimates of bullying, which tend to be between 5 and 15% (Beaty and Alexeyev, 2008, Christie, 2005). Most of these studies are based on a typical definition of bullying as outlined in an early work by Olweus; a pupil is bullied when (1) he/she is exposed, repeatedly and over time, to negative actions on the part of one or more of the other pupils, (2) the relationship between the bully and the bullied individual is characterized by an imbalance of power, and (3) the bully has the intention of doing harm (Limber, 2004, Olweus, 1978, Olweus, 1993).
Being a victim of bullying is associated with low self-esteem, self-harm, suicidal intention, depression, loneliness and physical ill-health (Barker et al., 2008, Fekkes et al., 2006, Ferguson et al., 2003, Hawker and Boulton, 2000, Nansel et al., 2004, Nishina et al., 2005, Rigby, 2003). It has also been shown that pupils who are bullied at a young age invest less in higher education compared to non-bullied control pupils (Brown & Taylor, 2008). On average, bullies, have poorer academic skills, perform below average in school, often lack empathy, are more likely to have substance abuse problems and face an increased risk of becoming criminals in adult life (Juvonen et al., 2003, Kaltiala-Heino et al., 2000, Le et al., 2005, Merrell and Isava, 2008, Nansel et al., 2001, Sourander et al., 2007).
Programs to reduce school bullying are on the increase, but there is often a lack of understanding of their effectiveness and monetary benefits. In 2007, the Swedish Ministry of Education and Research financed a project to summarize published data on the effectiveness of antibullying programs currently being used in Swedish schools. The preliminary results indicated that only 1 of the 21 antibullying programs in Swedish schools could be considered as evidence-based with proven efficiency in a proper evaluation (SNAE, 2007). In a similar context, the Swedish Council on Health Technology Assessment evaluation of 33 programs, used in Swedish schools to prevent mental ill-health among children, found that only seven programs had any evidence-based meaningful effects (SBU, 2010).
Antibullying programs are often described as being based on a “whole-school” approach, focusing on general interventions across all individuals, or an “individual-based” approach, targeting a small number of pupils that are considered to be at risk of being bullied or becoming bullies (e.g. disciplinary methods, parent training). A recent meta-analysis of 16 studies of school-based bullying-intervention (“whole-school” and “individual-level” programs) found that, for 10 out of 28 outcome variables, studies identified a significant reducing effect on bullying (Merrell & Isava, 2008). Another meta-analysis, only including “whole-school” anti-bullying programs, concluded that the majority of programs evaluated did not produce any significant beneficial effects (Smith, Schneider, Smith, & Ananiadou, 2004). A report on 59 studies documenting 30 different antibullying programs shows an average reduction rate in victimization (being bullied) of 17–23% (Ttofi, Farrington, & Baldry, 2008). The most important program elements for reducing bullying were disciplinary methods, videos, work with peers, parent training, cooperative group work and school yard supervision. Further, the number of elements and intensity of the programs were related to the size of the effect.
However, we have not been able to identify any studies that evaluate the total welfare effects of the intervention programs, i.e. none of the studies relate the costs of the interventions to the benefits.1 If a region, municipality, school et cetera, plans to invest in an antibullying program, it is relevant to compare the benefits and the costs of the program in order to ascertain if the investment is worthwhile. A cost–benefit analysis requires both the costs and the benefits of the intervention to be monetized. If the present value of monetized benefits of the program is larger than the costs, according to the Hicks–Kaldor criteria, the program can be said to increase welfare. Costs of the antibullying programs are mostly personnel costs (training and implementation), and some programs may also include material costs (educational material et cetera). It is more difficult to value the benefits of an antibullying program, i.e. the economic value of reduced bullying, and market data cannot (at least directly) be used for this purpose.
In this paper we show how the benefits of an antibullying program can be valued by estimating the societal willingness to pay (WTP) to reduce school bullying. We apply a discrete choice experiment (DCE) using a stated preference method. As stated, this can be used as a measure to compare the implementation costs with the benefits and thus obtain an economic and welfare evaluation of antibullying programs. It may also provide policymakers with useful information on taxpayers’ preferred allocations to antibullying programs. The rest of the paper is structured as follows: Section 2 describes the data collection and the DCE. Section 3 contains the descriptive statistics of the data, and Section 4 presents the econometric approach to estimating WTP. The results are shown in Section 5, and in Section 6 we show how the results can be used in economic evaluations of antibullying programs illustrated with a simple cost–benefit analysis of a common antibullying program. The paper is concluded with a discussion in Section 7.
Section snippets
Survey design
The data in this paper come from a stated preference (SP) survey conducted by mail in February 2010, with a reminder sent out three weeks later, in the municipality of Örebro in Sweden. The survey was sent to a random sample of 2001 individuals between the ages of 18 and 70 based on the Swedish Governments Personal- and Address-Register (SPAR), which includes all individuals with an address and currently living in Sweden. Örebro, situated approximately 200 km west of the capital Stockholm, has a
Response rates
Of the 2001 questionnaires, 29 (1.5%) were returned because they had not been sent to valid addresses. Of the remaining 1972 questionnaires we received 789 responses, i.e. a response rate of 39.4%. Of the returned questionnaires, 108 were returned blank and 19 had to be excluded due to missing information. The final sample available for use in the empirical estimation is based on answers from 662 respondents.
The willingness to pay/referendum question
Fig. 2 summarizes responses to the willingness to pay/referendum questions. It shows
Econometric approach
The willingness to pay to reduce school bullying is estimated using both non-parametric and parametric approaches. In Section 4.1 we describe the distribution-free and less restrictive non-parametric approach, which we consider as a starting point providing a lower bound WTP estimate. In Section 4.2 we proceed with describing the parametric approach and its foundation in the random utility model.
Results
As outlined in the previous section, WTP is estimated using both non-parametric and parametric approaches. In the final set of respondents WTP is estimated by (Model A) excluding all the respondents who gave protest answers as indicated by the follow-up questions, and (in Model B) all respondents who gave protest answers and respondents who gave the same answer (yes or no) in all five choice sets (see Section 3.2).
An application: cost–benefit analysis of an antibullying program
In order to demonstrate the application of the VSBV as calculated in the previous section, we perform a very simple cost–benefit analysis using the perhaps most well-known antibullying program: the “Olweus program”. The Swedish National Agency for Education has estimated that the average cost per pupil for bullying prevention programs for a typical school with 300 pupils in Sweden ranges from 97,673 kronor per school and year (“Friends” antibullying program) to 416,787 kronor per school and
Discussion
This study estimates the WTP for an antibullying program to reduce bullying in Swedish schools. Estimates of mean marginal WTP range from 5.95 to 8.48 Swedish kronor (€0.66–0.95) using non-parametric and parametric approaches. Estimates using the non-parametric approach are lower, which is expected considering that, as outlined in Section 4, the non-parametric estimation implies placing a lower-bound on the WTP. The aggregate mean WTP for the local public good, which we define as the value of a
References (57)
- et al.
Comparative benefit–cost analysis of the Abecedarian program and its policy implications
Economics of Education Review
(2007) - et al.
Bullying, education and earnings: Evidence from the National Child Development Study
Economics of Education Review
(2008) A new paradigm for valuing non-market goods using referendum data: Maximum likelihood estimation by censored logistic regression
Journal of Environmental Economics and Management
(1988)- et al.
Referendum models and negative willingness to pay: Alternative solutions
Journal of Environmental Economics and Management
(1997) - et al.
Bullying at school – An indicator of adolescents at risk for mental disorders
Journal of Adolescence
(2000) - et al.
Early childhood behaviours, schooling and labour market outcomes: Estimates from a sample of twins
Economics of Education Review
(2005) - et al.
Benefits and costs of investments in preschool education: Evidence from the Child–Parent Centers and related programs
Economics of Education Review
(2007) - ASEK (2012). ASEK (Arbetsgruppen samhällsekonomiska kalkyler). The Swedish Transport Administration,...
- et al.
Predictive validity and early predictors of peer-victimization trajectories in preschool
Archieves of General Psychiatry
(2008) - et al.
The problem of school Bullies: What the research tells us
Adolescence
(2008)
Discrete choice analysis
A computationally efficient quadrature procedure for the one-factor multinomial probit model
Econometrica
Microeconometrics using Stata
Value of statistical life and cause of accident: A choice experiment
Risk Analysis
Chasing the bullies away
Phi Delta Kappan
Bullied children get ill, or do ill children get bullied? A prospective cohort study on the relationship between bullying and health-related symptoms
Pediatrics
Vulnerability and resiliency to suicidal behaviours in young people
Psychological Medicine
Teacher support of bullying prevention
Valuing environmental and natural resources: The econometrics of non-market valuation
Willingness to pay for health protection: Inadequate sensitivity to probability?
Journal of Risk and Uncertainty
Welfare evaluations in contingent valuation experiments with discrete responses
American Journal of Agricultural Economics
Twenty years’ research on peer victimization and psychosocial maladjustment: A metaanalytic review of cross-sectional studies
The Journal of Child Psychology and Psychiatry and Allied Disciplines
Stated choice methods: Analysis & application
The value of changes in the probability of death or injury
Journal of Political Economy
Bullying among young adolescents: The strong, the weak, and the troubled
Pediatrics
Cited by (13)
The negative impact of bullying victimization on academic literacy and social integration: Evidence from 51 countries in PISA
2021, Social Sciences and Humanities OpenCommentary: Ways of preventing cyberbullying and evidence-based practice
2018, Reducing Cyberbullying in Schools: International Evidence-Based Best PracticesCommentary: Ways of preventing cyberbullying and evidence-based practice
2017, Reducing Cyberbullying in Schools: International Evidence-Based Best PracticesThe cost-effectiveness of the Olweus Bullying Prevention Program: Results from a modelling study
2015, Journal of AdolescenceCitation Excerpt :The first approach is almost never practically feasible, given the lack of data on all potential investments a decision-maker could choose instead. However, there is a recent study (Persson & Svensson, 2013) that estimated a policy-relevant threshold value in terms of the willingness to pay to spare one statistical victim of bullying. The authors used a discrete-choice experiment in which respondents (random sample of Swedish tax-payers in a municipality) in a survey experiment were asked to indicate preferential trade-offs between different types of policy investments of bullying intervention programs, i.e. the level of tax increases that respondents are willing to accept in order to implement bullying intervention programs in schools in their own municipality.
Disciplinary measures in schools: Are enough against bullying?
2014, Perfiles Educativos