The applicant was appointed harbour master at the Killybegs Fishery Harbour Centre in 1996.  A complaint arose around the applicant providing pilotage services through a company of which he was a director and a 1% shareholder. The Department of the Marine and Natural Resources (“the Department”) decided to investigate, appointing the personnel officer of the Department as the investigator in 2004.

Separately, a Minister (who was also a local TD) made unrelated complaints with reference to the applicant. The Minister then attended a meeting in 2004 with the investigator, at which the Minister outlined a wider range of complaints in relation to the applicant.

Thereafter, various matters were investigated, resulting in a final report by the investigator recommending dismissal.  This was appealed to the Civil Service Appeal Board. The Appeal Board did not accept that the applicant would not derive any benefit from his pilotage work. The Appeal Board did not disturb the recommendation of dismissal in respect of this matter. The Appeal Board also upheld the investigator’s findings concerning the other complaints, but disagreed with the investigator that these grounds were sufficiently serious to justify dismissal.

In 2009, the Cabinet met, including the Minister who had made complaints about the applicant. The Cabinet decided to dismiss the applicant.  The applicant applied to judicially review his dismissal, on the basis that the decision was affected by actual bias and objective bias. The applicant predominantly focused on two claims when presenting his case; the meeting of the Minister and the investigator in 2004 tainted the disciplinary process which followed, and the fact that the same Minister who had made the complaints in 2004, participated in a Cabinet meeting in 2009 which decided that the applicant should be dismissed from his civil service position.

The High Court refused the applicant’s application, and the Court of Appeal upheld this decision of the High Court.  Leave was given to appeal the decision to the Supreme Court. In granting leave to appeal to the Supreme Court, the Supreme Court considered that the case raised issues of law of general public importance relating to the test for bias.

Supreme Court Decision

Actual Bias

For actual bias to be established, a decision-maker must be “influenced by some existing relationship, interests, or attitude, without which the decision would be different.” In addition, the influencing factor on the decision-maker must be one that pre-dates and is external to the decision making process. In essence, the applicant must prove “the decision-maker was deliberately setting out to hold against a particular party, irrespective of the evidence.” Due to these requirements, “an allegation of actual bias is rarely likely to succeed”

 Dunne J discussed several factors as to why there was no actual bias on this occasion: the investigating officer had no pre-existing relationship/attitude towards the applicant, the decision to investigate the applicant had been made before the meeting between the Minister and the investigator in 2004, the matters complained of by the Minister did not form part of the subsequent investigation and the Minister’s involvement was included in the chronology presented to the Appeal Board (i.e. it was not concealed). Dunne J concluded that “there is simply no evidence to support the contention that [the investigating officer] was so influenced by the meeting with the Minister that he deliberately found against Mr Kelly”. All Supreme Court Judges agreed with Dunne J that actual bias was not present in this case.

Objective Bias

As stated by Denham J in Bula Mines Ltd v Tara Mines Ltd (no. 6), the test for objective bias is “whether a reasonable person in the circumstances would have a reasonable apprehension that the applicant would not have a fair hearing from an impartial judge on the issues.” 

On the question of objective bias, the Supreme Court considered both the Minister’s participation in the Cabinet meeting at which the decision was made to dismiss the applicant and the Minister’s meeting with the investigator in 2004 that preceded the dismissal.

On the Minister’s participation in the Cabinet meeting, Dunne J found:

It is, in my view, impossible to conceive of a situation in which the hypothetical, reasonable observer, aware of the relevant facts, would not have had a reasonable apprehension of bias, by reasonable involvement of the Minister in the ultimate decision to dismiss Mr Kelly…It should have been apparent to the Minister that it was inappropriate for her to participate in the Cabinet decision leading to the dismissal of Mr Kelly, given her previous interest and involvement in the matters at issue”

Dunne J stated that “a finding of objective bias against one member of a body, when that body has the responsibility of making a decision, will taint the decision of the body overall.”  Therefore, Dunne J concluded that the Cabinet’s decision to dismiss that applicant had to be quashed due to the Minister’s involvement.

While stating that both factors were relevant, O’Donnell J placed increased focus on the Minister’s meeting with the investigator in 2004 by reason of the facts: that the meeting was set up, that it is unlikely that such a meeting would have been afforded to a member of the public, that the normal function of civil servants is to accept direction from Ministers, that no official note of the meeting was kept and that shortly after the meeting the applicant was informed for the first time of the fact of the investigation and suspended from duty.

O’Donnell J criticised the application of the Caesar’s wife test when considering objective bias:

“[T]he reasonable bystander must be expected to be not just fair, but robust and aware that a standard for objective bias that is met if even a suspicion can be voiced could result in a near-impossible test which could be too easily invoked by disappointed parties who cannot point to any weakness in the individual decision.  It is easy to say that the system benefits if the most demanding standards are required, since this will exclude even unrealistic suspicions about the process, but such a test assumes that the benefit of avoiding any hint of suspicion in the mind of even the most committed cynic is costless, when in fact such a test exacts a very heavy price in decisions set aside and outcomes delayed.  Indeed, if a reasonable bystander had an interest in the classics, he or she might be aware that the test that Caesar’s wife must be above suspicion, often used to justify demanding of officials, adjudicators and judges that they not only perform their functions correctly, but do so in a way that cannot be criticised by even the most suspicious person … was first announced to allow Caesar justify divorcing his wife, Pompeia, an event that did not seem to trouble him, since it left him free in due course to marry a third wife.  Pompeia’s views are not known.  It would be more impressive if the standard of behaviour was one demanded of oneself rather than used as a vehicle to criticise and undermine the decisions of others.  This, or any other case, should not be approached on the basis that if a suspicion can be stated, particularly in a world of fevered social media commentary, a decision must inevitably be set aside.

Moving from the reign of Caesar to the reign of the smart phone, O’Donnell J brought matters back to today, warning against considering cases of objective bias “on the basis that if a suspicion can be stated, particularly in a world of fevered social media commentary, a decision must inevitably be set aside.”

O’Donnell J further acknowledged that:

the Minister did not form any part of the [investigating officer’s] investigation, that the investigation itself was meticulous, that the facts were not in dispute, and that [the investigating officer’s] recommendation as to the penalty was open to review…,an extensive applicant process before a body that had no knowledge of the Minister’s involvement, and which produced the relevant recommendation for dismissal.  Furthermore, the fact that the structure required that, before an established civil servant was dismissed, it was necessary for the Government to decide to accept the recommendation to that effect, provided another layer to the process tending to insulate it from the events of five years earlier.  If there was nothing more in the case, and a Cabinet that knew nothing of the background had accepted the recommendation of dismissal, then I think a reasonable bystander would consider that, while the events of 2004 were troubling, the suggestion that a dismissal in 2009 was a product of, or in some sense procured by, a ministerial intervention five years earlier was implausible.”

O’Donnell J went on to state that the special features of this case that would mean the reasonable bystander, whilst assumed to be more forgiving than Caesar, would reach a finding of objective bias:

“In this case, however, the Minister was also an active participant in the investigation, and made a complaint in relation to it, was present at the outset of it, was one of the first people to meet the investigator and express strong views to the investigator, and then participated in the final decision.  … I consider that the reasonable bystander would not, at a minimum, be able to confident that the procedure was not affected by the twin involvements of the Minister as a complainant/informant at the outset of the process, and as a decision-maker at its conclusion. ”

On the Minister’s meeting with the investigator, MacMenamin J warned that the investigator “whose findings were to have binding effect, embarked on the investigation with this range of quite damning criticisms in his mind”. MacMenamin J went on to explain that the focus was not on whether the investigator actually behaved in a bias manner:

[The] objective bias test does not concern whether [the investigating officer] used this information or whether it informed part of his investigation; but, rather, that a meeting had taken place between himself and an extremely important person who had told him these things about the man who he was about to investigate, and that he [the investigating officer], had carefully noted them all down.”

Despite differing views on exactly which aspect of the investigation was the most problematic in terms of objective bias, all, except Charleton J, found that the investigation process was tainted by objective bias due to the involvement of the Minister, with MacMenamin J summing up the sentiment:

It is necessary to bear in mind throughout, that the test for objective bias is, itself, an objective one … It is not what we judges think, but rather the inference which an objective observer would draw as to the process, seen in its entirety.”


This decision serves as a useful reminder for employers in relation to the following issues:

  1. Actual bias is difficult for employees to prove, as they must show that the decision-maker was deliberately setting out to find against the employee, irrespective of the evidence produced throughout the investigation.
  2. Objective bias can be established if the employee under investigation can show that there is a reasonable apprehension that they did not receive a fair investigation conducted by an impartial investigator and concluded by an impartial decision-maker.
  3. Re-affirmation of the fair procedures principle that an individual cannot be the decision-maker in their own case. Twin involvement as both the complainant and the decision-maker is a catalyst to an investigation being tainted by objective bias. This may result in the decisions from an investigation being quashed.
  4. When appointing investigators and impartial decision-makers, employers should ensure that the appointees have not previously been a party to the investigation (i.e. a witness or complainant) and do not have a history of animosity towards the employee under investigation.


The authors wish to thank Rosanna McAleese for her contribution to this article.