Whistleblowing can be about a wide spectrum of illegal, immoral or illegitimate practices. In this research, the spectrum is captured by the general term ‘wrongdoing’, with data collected in terms of a large number of different types of behaviour that could be categorised in a number of different ways—including whether there could be primarily private or public interests involved. A wide definitional approach was necessitated not only by the imprecision of existing categories, but by the inherent dangers of taking too restrictive an approach, whether in research or in legislation. For example, restricting whistleblowing to only ‘serious’ examples of public interest wrongdoing begs the question of in whose judgment the wrongdoing meets that threshold. An agency wishing to minimise the amount of whistleblowing to which it need respond can simply increase the threshold. In reality, the benefits of whistleblowing are likely to be encouraged only when agencies promote an ‘if in doubt, report’ approach (Brown et al. 2004). For this reason, the present research sought to establish a comprehensive picture of the breadth of wrongdoing types that public officials observed and might potentially disclose.

Appendix 2 lists the 38 different behaviours that were offered to survey respondents in the present research as examples of wrongdoing. These were then grouped into seven major categories for the purpose of analysis:

The first six of these major categories align substantially with the subjects about which statutory public interest disclosures may be made in several jurisdictions (see Brown 2006:16–19). For the same reason, they also align with the jurisdictions of several of the key integrity agencies that were partner organisations to the project. Consistent with the definition discussed in the previous section, when this report refers to ‘whistleblowing’, it is referring to reports made about wrongdoing in one of these six categories—that is, ‘public interest’ whistleblowing.

Data about wrongdoing types in the last category—personnel or workplace grievances—were also collected in order to maintain the comprehensive approach. In this report, however, they are presumed to constitute mostly private or personal complaints and they are not included in the results for whistleblowing. At times, where relevant, these grievances are nevertheless included in analyses of reporting behaviour and outcomes as a whole, but when this occurs it is made clear in the relevant sections of the report (primarily in Chapters 2 and 3).

As can already be seen, all attempts to measure and categorise perceptions of wrongdoing are subject to their own limitations and are unlikely to ever be perfect. For example, it can be argued that many of the wrongdoing types included in some of the public interest categories are chiefly issues of organisational integrity and will become matters of public interest only if exceptionally serious. This possibility alone, however, combined with the fact that the under-counting and under-reporting of public interest disclosures is already known to be a problem in many jurisdictions, makes it important to include such items. Similarly, some forms of public interest wrongdoing are clearly more likely than others to be triggered first and foremost by a private interest or personal grievance (for example, several of those wrongdoing types grouped as ‘improper behaviour’). As outlined earlier, however, the fact that such disclosures might also involve personal grievances does not detract from their public significance. Indeed, the extent to which these wrongdoing types could also be re-categorised as personal grievances is offset by the likelihood that some ‘personnel or workplace grievances’ excluded from the analyses of whistleblowing in this report involve entrenched or systemic wrongdoing of sufficient seriousness that they could be reclassified objectively as matters of public interest.

For all these reasons, the results given in this research are best understood as simply an approximation of the numbers of cases falling into the different categories used. While the project team is confident that this is a more comprehensive and useful approximation than previously achieved, it should not be seen as definitive.