Discussion
The main objectives of this study were to provide cumulative estimates of the
prevalence of workplace bullying, and to investigate the influence of methodological
artifacts on reported prevalence rates. Across the studies included, an overall estimate of
Prevalence of workplace bullying 967
14.6% was established. Yet, such an overall estimate must be interpreted with caution as
the further findings confirm our hypotheses that prevalence rates of workplace bullying
are itifluenced by measurement methods and sampling procedures. With an overall rate
of about 11%, the results show that studies using the self-labelling with definition
method provide the lowest estimate of bullying. When using the behavioural
experience method, the prevalence is somewhat higher with rates about 15%, whereas
the self-labelling without definition method yields the highest estimates with overall
rates of 18%. Hence, compared to studies using self-labelling with definition, the
prevalence rates of this latter approach are almost twice as high. With regard to
sampling technique, the findings show that non-random samples at an average yield
8.7% point higher estimate of workplace buUying than do random samples. Yet, when
exatnining sampling technique within the measurement method categories, sampling
method had less impact on the prevalence estimates.
Based on the present meta-analysis, workplace bullying seems to be a substantial
problem in contemporary working life. At a worldwide average, the findings show that
at least 1 out of 10, and maybe as many as about 1 out of 5, workers are exposed to
bullying in their workplaces. As bullying and harassment have been shown to have
considerable negative effects both for organizations and those exposed (Bowling &
Beehr, 2006; Einarsen & Mikkelsen, 2003; Hoel, Einarsen, & Cooper, 2003), our findings
demonstrate that workplace buUying is a problem that should be taken into
consideration in organizational life. Hence, governments, employers, and organizations
need to take the bullying issue seriously by developing and implementing interventions
and strategies to handle the problem.
Compared to adjacent phenomena such as sexual harassment and aggression, this
study shows that workplace bullying has lower prevalence rates. For instance, in a metaanalysis
of the incidence of sexual harassment, Ilies et al (2003) found that 58%
reported exposure to sexual harassing behaviour, whereas 24% self-labelled as victims
of such harassment. In a representative survey of aggression among US workers, it was
found that 41.4% had experienced at least one instance of psychological aggression at
work during the past 12 months (Schat et al., 2006). Although based on inconsistent
reporting periods, the lower prevalence rates found in our meta-analysis support the
assumption that w^orkplace bullying is a phenomenon distinct from other forms of
interpersonal aggression. That is, whereas aggression and harassment may be described
by single and isolated behaviours, workplace bullying is per definition based on
repeated behaviours over a prolonged time period. Consequently, because more
frequent negative acts are needed to describe the bullying phenomenon, prevalence
rates should also be lower compared to aggression and harassment.