After adjusting for all variables in the model, the risk factors found to be significantly associated with poor treatment by co-workers are shown in Table 6.6. The column of particular note in this table is the ‘odds ratio’: this presents the probability of a public interest whistleblower experiencing poor treatment divided by the probability of them experiencing good or similar treatment. An odds ratio greater than 1 indicates higher odds of poor treatment.

The findings highlight four key risk factors associated with the poor treatment of whistleblowers by co-workers. The odds of being treated badly by co-workers were:

Source: Employee survey: Q22, Q24, Q33, Q49 (n = 488).

This group of risk factors confirms much of what was already suspected regarding the 9 per cent of whistleblowers identified in Chapter 5 as reporting co-worker mistreatment. It is also consistent with the conclusions in the previous section, that the major factors associated with the risk of co-worker reprisal are readily apparent to most whistleblowers and are more likely to be accurately assessed by them.

In particular, it makes sense that reporters who observed wrongdoing involving more than one person had greater odds of poor treatment, given the power inequality between a group of wrongdoers and a sole reporter and the increased likelihood that the wrongdoing had become part of the workplace culture. Similarly, when the size of the reporter’s immediate workgroup is smaller, there is greater opportunity for direct or indirect reprisals (given the likely higher level of workplace interaction with the wrongdoer[s]) and there is more likely to be an environment conducive to them. These are indicators of a classically stereotypical reprisal situation.

The top two factors nevertheless have more interesting implications. Chapter 5 discussed a range of explanations for why a positive substantive outcome, such as substantiation of the report, played such a pivotal role in shaping the overall outcome of the whistleblowing incident. One explanation was that whistleblowers whose reports did not lead to positive action could be inclined to regard that result as itself constituting bad treatment—that is, they were simply dissatisfied. Given the overall pattern of results in Chapter 5, however, and the personal attitudinal variables in the previous section, this explanation would appear to fit only a limited number of cases. Moreover, this procedural dissatisfaction is not likely to be directed at their co-workers, but rather at the management processes responsible for the investigation result, as discussed further below.

A more salient explanation, therefore, is that in a range of circumstances, unless there is formal substantiation or other positive action that vindicates their report, the whistleblower could be ‘fair game’ for direct or indirect reprisals. At present, it appears that many internal whistleblowers are left in the position of having to emerge either as a ‘winner’ or ‘loser’ from the experience, rather than being left with the possibility of a more neutral outcome. In these situations, achieving substantive vindication becomes a primary factor in avoiding mistreatment, which could explain the degree of dependence on a positive substantive result described in Chapter 5.

The salience of this factor is reinforced by the role of the seriousness of the wrongdoing. It was hypothesised that more serious wrongdoing, once exposed, might not necessarily lead to a greater risk of co-worker reprisal because others were also more likely to share the whistleblowers’ view of the seriousness of the matter. Conversely, it was thought that more petty matters could lead to higher reprisal risk, because co-workers would be more likely to agree with the alleged wrongdoer that the matter should never have been reported. This analysis confirms, however, that the more serious the perceived wrongdoing, the higher is the likelihood of co-worker reprisal, presumably because of the more serious consequences for the wrongdoers. Clearly, this risk then becomes extreme when the allegation is also not substantiated.

Two further hypotheses were also not supported by this analysis. Logically, some types of alleged wrongdoing could be thought to be more likely to trigger reprisal risks, depending on the impact of the allegations on the perceived wrongdoer. In Chapter 5, it was also pointed out that certain types of wrongdoing were more likely to be substantiated than others. This analysis shows, however, that it is these other factors—seriousness and lack of substantiation—that are associated most closely with co-worker mistreatment, irrespective of what type of wrongdoing is involved. In other words, while some types of wrongdoing can be objectively considered more ‘serious’ and also more difficult to substantiate, it is not the wrongdoing type itself but these results that dictate whether mistreatment is likely to follow. Any kind of wrongdoing can therefore give rise to these risks, depending on the circumstances.

Contrary to expectations, the level of confidentiality maintained in respect of the whistleblowing incident also failed to emerge as a significant predictor of mistreatment. In total, 83 per cent of whistleblowers who said that they were treated well or the same by co-workers and by management said that either ‘no-one’ or just ‘a few’ people knew they had reported the wrongdoing—compared with 63 per cent of those who were treated badly by co-workers, management or both. The fact that confidentiality did not emerge as a clear protective factor confirms that while it could be important, in most circumstances, it is either simply unachievable or plays no protective role. For example, even if only a few people ever know the whistleblower’s identity, this usually includes the people most directly concerned. This result has significant implications for the strategies that agencies use to try to control reprisal risk and, in particular, the need for agencies to be ready with other intervention strategies at the point that confidentiality ceases to be practical.

The second analysis examined the risk factors for mistreatment for those employee survey whistleblowers who indicated poor treatment from management (as distinct from co-workers). The results are set out in Table 6.7. The findings confirmed that, for the 18 per cent of whistleblowers who reported mistreatment by management, the factors associated with this result were more complex. Six key risk factors emerged, only two of which were common to mistreatment by co-workers. The odds of being treated badly by management were:

Source: Employee survey: Q21, Q22, Q23, Q25, Q28, Q33 (n = 496).

The strongest common factor was the lack of any positive investigation result. This becomes an even stronger predictor of managerial mistreatment than of co-worker mistreatment. Its salience further confirms that, while some whistleblowers might be more likely than others to be dissatisfied by an unfavourable investigation result or push the matter in a way that generates more conflict, this is often likely to be the case because of the crucial protective value of achieving some form of vindication (fighting to emerge as a ‘winner’ rather than a ‘loser’). The second common factor—the perceived seriousness of wrongdoing—was still present but not as important. This suggests that co-workers perhaps cannot be bothered reacting negatively in less serious matters, but that adverse management reactions can also flow in response to less serious matters, even though it is still more likely when the matters are serious.

The most dominant risk factor—the fact that managerial mistreatment is strongly associated with external reporting—combines with the issue of a substantive result to highlight again that any causality can run in either direction. Chapter 4 showed that most external whistleblowers had reported internally in the first instance. Therefore, it is likely that many whistleblowers are triggered to continue to pursue a disclosure externally because they experienced mistreatment within their organisation or because their claim could not be substantiated internally, or both. While it is more than possible that some whistleblowers experience mistreatment only after they disclose externally, in most cases, it can be presumed that external reporting follows the onset of reprisals or other conflict between the whistleblower and management, with external involvement then perhaps contributing to this real or perceived mistreatment in many cases.

This result nevertheless emphasises that in current circumstances—irrespective of the cause—by the time disclosures reach outside the organisation involved, crucial damage to the wellbeing of the whistleblower has often been done. This is further emphasised by the fact that this risk factor was not present for mistreatment by co-workers, but was the strongest of all factors associated with managerial mistreatment. These results further increase the obligations on agencies to manage their investigation and employee welfare processes actively and positively in the first instance. Failure to do so simply increases the likelihood of breakdowns in trust, external complaints and increasingly entrenched conflicts—all of which can have negative impacts on the organisation as well as any chance of a return to normality for the whistleblower.

The same results also support the conclusion that additional processes are needed to circumvent or reduce this risk of conflict, wherever internal processes are more likely to be tested by particular cases. Taken together, the remainder of the factors identify these classes of cases, in which the potential for internal processes to bring the matter to a successful conclusion is clearly more limited, at least without special measures or outside assistance. For example, it stands to reason that when a whistleblower reports a more serious or systemic matter, implicating managers, but an investigation does not or is not likely to resolve the matter in the whistleblower’s ‘favour’, there is a very high risk that managerial action (or inaction) will occur that constitutes real or perceived mistreatment. This is clearly also compounded if the whistleblower feels that they are directly personally involved, with their capacity to independently self-manage the risks even further reduced, than if they are simply an observer or bystander.

Taken together, these factors emphasise why agencies need special processes for dealing with the welfare of the whistleblower and addressing the potential sources of conflict, outside the normal management chain. They also suggest the need for processes that monitor management responses to the whistleblower, before conflicts emerge or become too difficult to control. These issues, including the question of how earlier routine oversight by external integrity agencies might assist this process, are taken up in later chapters.

Again, it is salient that neither the wrongdoing type nor lack of confidentiality emerged as risk factors in this second analysis. Wrongdoing type could still remain indirectly relevant: not only are certain types of wrongdoing more likely to be substantiated than others, as shown in Chapter 5, certain types are more likely to come to the attention of whistleblowers as a result of having been directed at them personally, as shown in Chapter 3. It is, however, once again these other factors—lack of positive investigation outcome, personal involvement and, to a lesser extent, seriousness and frequency—that are most closely associated with management mistreatment, irrespective of what type of wrongdoing is involved. The absence of confidentiality as a protective factor is even less surprising than it is for co-worker mistreatment, given that, even if only a few people know, these are likely to be people in the management chain and are therefore well placed to undertake or be blamed for any reprisals.

The other very logical factor, confirming the realistic nature of these results, is that of the power differential between reporters and perceived wrongdoers. If more senior people are implicated, irrespective of their own level in the organisation, a whistleblower has double the chance of experiencing mistreatment from management, absent of any other factors. The lack of this factor in respect of co-worker mistreatment is perhaps unsurprising, but the difference tends to confirm that, overall, agencies have difficulty responding to upwards challenges to the integrity or competence of their own management. When other factors such as seriousness, frequency, personal involvement and difficulties in substantiation also combine, the risk of mistreatment is plainly extreme.

Taking both analyses together, it is clear that these very different listings of risk factors mean that the risks of mistreatment from each source need to be separated and independently assessed. The differences accord with the finding from Chapter 5 that only 5 per cent of whistleblowers experience negative treatment from both management and co-workers. The remaining 17 per cent of whistleblowers who recorded mistreatment perceived it as coming from either one or the other. The fact that the single largest group—13 per cent of all whistleblowers—perceived the mistreatment as coming from management alone, and that the risk factors for this group were more numerous, more complex and in some instances stronger in terms of significance, confirms that this is a major area for further inquiry and investment.