Oyugi v Teachers Service Commission & 4 others

(Ruling on the Petitioner/Applicant's Notice of Motion dated 15th June 2026, seeking stay of delivery of judgment and the recusal of the presiding Judge) RULING A. Introduction 1. Few applications place a judge in as delicate a position as one that invites him to sit in judgment upon an allegation that he is not t to judge. The temptation, on the one hand, is to react defensively; on the other, to recuse too readily simply to avoid controversy, however unmeritorious the allegation. Neither temptation serves the administration of justice. What is required, instead, is the same dispassionate, reasoned analysis this Court would bring to any other application — tested against settled principle, and against the record, rather than against personal sentiment on either side. 2. This Ruling determines the Petitioner/Applicant's Notice of Motion dated 15th June 2026, led under a Certicate of Urgency of even date, by which he seeks, principally, an order that I recuse myself from any further conduct, hearing, determination and/or delivery of judgment in this Petition, together with an order staying the delivery of judgment that had been xed for 2nd July 2026 pending the hearing and determination of the Application, and an order that the le be placed before the Principal Judge of this Court for directions and reallocation to another Judge. 3. The Application arises out of, and must be understood against, a Petition that has already occasioned two substantive interlocutory rulings of this Court: a Ruling delivered on 29th January 2026 dismissing the Respondents' preliminary objections to jurisdiction and, upon hearing the parties inter partes, vacating the ex parte conservatory orders earlier obtained by the Petitioner; and a Ruling delivered on 4th June 2026 declining the Petitioner's application for leave to withdraw the Petition, in which this Court made ndings, adverse to the Petitioner, that his intended withdrawal amounted to forum shopping and an abuse of the process of this Court. It is those two rulings, and in particular the second, that the Petitioner says give rise to a reasonable apprehension that this Court has already formed, and cannot now discard, a concluded view adverse to him on matters that remain for determination in the substantive Petition. 4. Given the gravity that necessarily attaches to any suggestion, however framed, that a sitting judge is not to be trusted with the impartial determination of a cause properly before him, I have thought it right to address this Application with the fullest measure of care: to set out, at some length, the nature and limits of the judicial function in an adversarial system such as ours; the reasons why applications of this kind, though a legal entitlement of every litigant, ought as a matter of sound principle to be brought sparingly; the settled law, drawn from this jurisdiction and from other common law jurisdictions, on when a court may and may not properly recuse itself; and, nally, a critical and dispassionate assessment of the specic grounds advanced in this particular case. For completeness, and so that neither party may afterwards suggest that its position was not fully or fairly captured, I have set out the rival submissions in a distinct section of this Ruling before turning to my own analysis. B. Background 5. The Petitioner instituted this Petition on 25th May 2025 in the public interest, challenging, inter alia, the constitutionality of Section 16(2) of the Teachers Service Commission Act, and the constitutionality of the ongoing recruitment process for the position of Secretary/Chief Executive Ocer of the 1st Respondent, advertised on 6th May 2025. Contemporaneously with the Petition, the Petitioner https://new.kenyalaw.org/akn/ke/judgment/keelrc/2026/2228/eng@2026-07-31 2

page 2 of 16

led a Notice of Motion application seeking, among other orders, a conservatory order staying the recruitment process pending the hearing and determination of the Petition, which orders he obtained ex parte on 27th May 2025 and enjoyed for a period of approximately eight months. 6. The 1st Respondent, together with the 2nd, 3rd and 4th Respondents, opposed the Petition and the interlocutory application, and additionally raised preliminary objections contending that this Court lacked jurisdiction to entertain either, principally on the ground that the dispute did not fall within the classes of matter and person contemplated by Section 12 of the Employment and Labour Relations Court Act, and that jurisdiction to pronounce upon the constitutionality of a statute lay exclusively with the High Court under Article 165(3)(d)(i) of the Constitution. 7. By a Ruling delivered on 29th January 2026, this Court dismissed the Respondents' preliminary objections, holding that it possessed the requisite jurisdiction to entertain the Petition and the constitutional question raised within it, that jurisdiction being neither ousted by the absence of an employer-employee relationship at the recruitment stage, nor conned, as the Respondents had urged, to matters falling squarely within Section 12 of the Act. In the same Ruling, however, and upon a full inter partes hearing of the merits of the conservatory application, the Court vacated the ex parte orders earlier obtained by the Petitioner, and directed that the Petition proceed to hearing and determination on its merits. 8. It bears emphasis, at the very outset, that the Ruling of 29th January 2026 was not, in substance, a ruling wholly or even predominantly adverse to the Petitioner. It vindicated the very forum the Petitioner had chosen to litigate his grievance, over the determined objection of four Respondents who sought to have the Petition struck out in limine for want of jurisdiction; it was only on the distinct and narrower question whether the ex parte conservatory orders should be continued that the Court ruled against him. I return to the signicance of this fact in the critical analysis that follows. 9. Shortly after the delivery of that Ruling, the Petitioner led a Notice of Withdrawal of the Petition, which withdrawal the Respondents did not oppose save as to the issue of costs. Pursuant to directions issued by this Court, the Petitioner thereafter led a formal Application dated 10th April 2026 seeking leave to withdraw the Petition under the Constitution of Kenya (Protection of Rights and Fundamental Freedoms) Practice and Procedure Rules, 2013. 10. The 1st Respondent opposed the withdrawal application, ling a Replying Adavit sworn by its Chairperson, Dr. Jamleck Muturi John, on 18th May 2026, and written submissions of even date, in which it contended, inter alia, that the intended withdrawal formed part of a broader scheme connected to related proceedings pending before a court of concurrent jurisdiction — specically, Kiambu Petition No. 006 of 2026 — and amounted to forum shopping and an abuse of the process of this Court. The 1st Respondent expressly framed, as one of the issues for the Court's determination, whether the withdrawal application had been brought in bad faith and constituted an abuse of process. 11. By a Ruling delivered on 4th June 2026, this Court dismissed the Petitioner's application for leave to withdraw the Petition. In doing so, the Court made ndings, expressed in unambiguous terms, that “forum shopping is evident”, that “the court process is being abused”, and that the intended withdrawal of the Petition was aimed at aiding proceedings pending before another court. The Court simultaneously noted that in related proceedings before it, being ELRC Petition No. E022 of 2025, Peter Kibelesi Kukubo v Teachers Service Commission, National Assembly & 2 Others, an application by a dierent litigant seeking withdrawal of that petition had, on a materially similar basis, likewise been declined. https://new.kenyalaw.org/akn/ke/judgment/keelrc/2026/2228/eng@2026-07-31 3

page 3 of 16

12. Upon delivery of the Ruling of 4th June 2026, the Court xed the Petition for delivery of judgment on 2nd July 2026. It is a matter of record, and one to which I return later in this Ruling, that the Petitioner, being dissatised with that Ruling, sought and was granted leave to appeal against it. 13. Rather than pursue that avenue of appeal to its conclusion, the Petitioner instead led, on 15th June 2026, the Certicate of Urgency, Notice of Motion and Supporting Adavit that are the subject of this Ruling, seeking, in substance, the recusal of the presiding Judge from any further conduct of this Petition and a stay of the delivery of judgment pending the determination of that request. 14. Mindful of the gravity that attaches to any pending recusal application, and consistently with prudent judicial practice, I did not proceed to deliver judgment on 2nd July 2026 as earlier scheduled. I instead directed that the Application be canvassed by way of written submissions, which directions the parties have complied with: the 1st Respondent ling a Replying Adavit and submissions in opposition, and the Petitioner ling submissions in support, the last of which was led on 9th July 2026. It is that Application which now falls for determination, and upon whose determination the future conduct of this Petition, including the xing of a fresh date for delivery of judgment, depends. C. The Application for Recusal and the Grounds Advanced 15. By the Notice of Motion dated 15th June 2026, expressed to be brought under Articles 25(c), 27, 50(1), 159(2)(a), (b) and (e), and 259 of the Constitution, Sections 1A, 1B and 3A of the Civil Procedure Act, Rule 17 of the Constitution of Kenya (Protection of Rights and Fundamental Freedoms) Practice and Procedure Rules, 2013, the Judicial Code of Conduct and Ethics, 2020, the Bangalore Principles of Judicial Conduct, and all other enabling provisions of the law, the Petitioner/Applicant seeks orders, in summary, that: (a) the Application be certied urgent and service thereof be dispensed with in the rst instance; (b) there be a stay of the delivery of the judgment scheduled for 2nd July 2026 pending the hearing and determination of the Application; (c) Hon. Justice Ocharo Kebira be pleased to recuse himself from any further conduct, hearing and determination of this Petition; (d) the Court le be placed before the Principal Judge of this Court for directions and reallocation to another Judge of competent jurisdiction; and (e) costs of the Application be provided for. 16. The Application is premised on the grounds set out on its face and elaborated in the Applicant's Supporting Adavit sworn on 15th June 2026. Reduced to their essence, and read together with his subsequent submissions, the grounds may be summarised as follows. 17. First, the Applicant avers that the Ruling of 4th June 2026 contained “denitive ndings and conclusions” regarding his conduct, motives and intentions, including the ndings that “forum shopping is evident”, that “the court process is being abused”, and that his intended withdrawal was aimed at aiding proceedings pending before another court, and that these ndings were not conned to the procedural question of withdrawal but extended to contested matters touching on his motives, credibility and the legitimacy of the proceedings generally. 18. Second, he avers that in the related proceedings in ELRC Petition No. E022 of 2025, Peter Kibelesi Kukubo v Teachers Service Commission, National Assembly & 2 Others, an application by a dierent litigant seeking withdrawal of that petition was similarly declined, and contends that this https://new.kenyalaw.org/akn/ke/judgment/keelrc/2026/2228/eng@2026-07-31 4

page 4 of 16

successive pattern of refusal, considered together with the denitive ndings made against him personally, demonstrates that this Court has formed “a settled view regarding the propriety, purpose and legitimacy of the related proceedings”, thereby giving rise to a reasonable apprehension of prejudgment. 19. Third, he avers that the eect of the foregoing is that a fair-minded and informed observer, considering the cumulative dismissal of the interlocutory application, the refusal to permit withdrawal of the present Petition, the refusal to permit withdrawal in the related matter, and the denitive ndings concerning forum shopping and abuse of process, would reasonably apprehend a real possibility that this Court has already formed a concluded view on matters remaining for determination in the substantive Petition. 20. Fourth, he avers, and I record this expressly because it is to his credit that he says so, that his apprehension does not arise merely from adverse rulings as such, but from what he characterises as the cumulative eect of those rulings together with the unequivocal ndings made concerning his motives, conduct and intentions — a distinction to which the law, as will be seen, attaches real signicance. 21. Fifth, he avers that unless the delivery of judgment, xed for 2nd July 2026, is stayed pending the hearing and determination of this Application, the Application will be rendered nugatory and he will suer grave prejudice through the loss of an opportunity to have his concerns regarding judicial impartiality properly adjudicated. 22. On that basis, the Applicant urges that it is necessary, expedient and in the interests of justice that I recuse myself from any further conduct, hearing, determination and/or delivery of judgment in this Petition, and that the le be placed before the Principal Judge for reallocation to another Judge. D. Submissions 23. Because an allegation of this character calls for particular transparency in how it has been met, I set out in this section, at some length and in a manner distinct from my own analysis, the rival submissions led by the parties, so that it is beyond doubt that both positions were fully considered before this Ruling was arrived at. (i) The Petitioner/Applicant's Submissions (led 9th July 2026) 24. Counsel for the Petitioner, Prof. Tom Ojienda & Associates, opens by making plain that the Application “is not premised on dissatisfaction with adverse orders or decisions made by the Honourable Court”, and that the Applicant “appreciates that a judicial ocer is entitled to make determinations against a party and that an unfavourable decision, without more, cannot form the basis for recusal.” The grievance, it is submitted, arises instead from “the cumulative eect of the proceedings herein, the circumstances surrounding the refusal to permit withdrawal of the Petition and the denitive ndings made by the Honourable Court concerning the Applicant's conduct, motives and intentions.” 25. Counsel submits that the constitutional guarantee of a fair hearing under Articles 25(c), 27 and 50(1) encompasses not only actual impartiality but also its appearance, and that the issue for determination is “whether, in the circumstances of this matter, a fair-minded and informed observer would reasonably apprehend that there exists a real possibility that the Honourable Court may not approach the remaining proceedings with an open and impartial mind.” 26. In support of that formulation, counsel invokes the celebrated English authority of Porter v Magill [2002] 2 AC 357, as recited with approval by the Kenyan courts, for the proposition that the applicable https://new.kenyalaw.org/akn/ke/judgment/keelrc/2026/2228/eng@2026-07-31 5

page 5 of 16

test is “whether the fair-minded and informed observer, having considered the facts, would conclude that there was a real possibility of bias.” Counsel further relies on Rono (Suing as the Executrix of the Estate of the Late David Rono) v Almer Farm Limited [2025] KEELC 4403 (KLR), and, through it, on Philip K. Tunoi & Another v Judicial Service Commission & Another [2016] eKLR, for the enumeration of circumstances in which a judge may properly recuse himself, and for the further proposition, drawn from the same authority, that “it is not the mind of the Judge which is considered but the impression given to reasonable people.” 27. Counsel places particular reliance on Rawal & 2 others v Judicial Service Commission & 2 others; Okoiti (Interested Party); International Commission of Jurists & 2 others (Amicus Curiae) [2016] KESC 1 (KLR), quoting the Supreme Court's statement that “recusal is indeed a judicial duty and it does not amount to a judge abrogating his/her duty. Apart from personally recusing to safeguard one's integrity and the sanctity of proceedings, recusal also helps protect the integrity and dignity of the bigger institution, that is, the judiciary and aids in championing its independence.” 28. Counsel also draws attention to Thugi River Estate Limited & another v National Bank of Kenya Ltd & 2 others [2014] KEHC 8278 (KLR), in which the Honourable L.N. Gacheru, J. recused himself notwithstanding an express nding that the apprehension of bias raised before him was “far-fetched”, reasoning that “I would not hesitate to recuse myself, if only my recusal would give the applicant the condence to approach the Court for arbitration of this matter and also to remain assured that justice would indeed be done”, and to Rai & 3 others v Rai & 4 others [2007] KECA 21 (KLR), for the proposition that the issue of recusal is founded on the need to maintain public condence in the impartiality of the judiciary. 29. On that basis, counsel submits that the Applicant has met the threshold for recusal; that the delivery of judgment ought to be stayed pending determination of the Application, failing which the Application would be rendered nugatory; and that the matter should be placed before the Principal Judge for reallocation, an order which, it is submitted, would occasion no prejudice to any party and would instead safeguard condence in the eventual determination of the Petition. (ii) The 1st Respondent's Submissions in Opposition (led 8th July 2026) 30. The 1st Respondent, through Mr. Cavin Anyuor, Advocate, opens its submissions by acknowledging that judicial impartiality is a fundamental pillar of justice under Article 50(1) of the Constitution, but submits that this must be balanced with a judge's duty to sit, citing Shollei v Judicial Service Commission & another [2018] KESC 42 (KLR), where the Supreme Court held: “ Tied to the constitutional argument above is the doctrine of the duty of a judge to sit. Though not profound in our jurisdiction, every judge has a duty to sit, in a matter which he duly should sit. So that recusal should not be used to cripple a judge from sitting to hear a matter. This duty to sit is buttressed by the fact that every judge takes an oath of oce: 'to serve impartially; and to protect, administer and defend the Constitution.' It is a doctrine that recognizes that having taken the oath of oce, a judge is capable of rising above any prejudices, save for those rare cases when he has to recuse himself. The doctrine also safeguards the parties' right to have their cases heard and determined before a court of law.” 31. Counsel submits that recusal is an exceptional remedy which must be considered with great caution, and that the Applicant bears the burden of demonstrating, on objective and reasonable grounds, that a fair-minded and informed observer would conclude that there exists a real possibility of bias. He relies on Kaplana H. Rawal v Judicial Service Commission & 2 others [2016] eKLR, concurring with the https://new.kenyalaw.org/akn/ke/judgment/keelrc/2026/2228/eng@2026-07-31 6

page 6 of 16

decision of the East African Court of Justice in Attorney General of Kenya v Prof Anyang' Nyong'o & 10 Others, EACJ Application No. 5 of 2007, for the proposition that: “ We think that the objective test of 'reasonable apprehension of bias' is good law. The test is stated variously, but amounts to this: do the circumstances give rise to a reasonable apprehension, in the mind of the reasonable, fair minded and informed member of the public that the judge did not (will not) apply his mind to the case impartially. Needless to say, a litigant who seeks disqualication of a judge comes to court because of his own perception that there is appearance of bias on the part of the judge. The court however, has to envisage what would be the perception of a member of the public who is not only reasonable but also fair minded and informed about all the circumstances of the case.” 32. Counsel further relies on the Supreme Court's decision in Rawal & 2 others v Judicial Service Commission & 2 others (the same authority relied upon, for a dierent proposition, by the Applicant), for its adoption of the test articulated in Jasbir Singh Rai & 7 Others v Tarlochan Singh Rai & Others: “Perception of fairness, of conviction, of moral authority to hear the matter, is the proper test of whether or not the non-participation of the judicial ocer is called for”, and submits that the question is therefore whether the Applicant “has demonstrated real likelihood of bias or prejudice”, noting the denitions in Black's Law Dictionary, 7th Edition, that bias is “inclination, prejudice; Judge's bias usually must be personal or based on some extrajudicial reasons”, while prejudice means “a preconceived judgment formed without a factual basis, a strong bias.” 33. Applying that standard to the facts, counsel submits that the Applicant's case for recusal is built solely upon the Ruling of 4th June 2026, and that a fair-minded and informed observer, having regard to the fact that the impugned ndings were made in answer to an issue the 1st Respondent itself had expressly framed for the Court's determination — namely, whether the withdrawal application was made in bad faith and amounted to an abuse of process — would conclude that those ndings were reasoned judicial ndings stemming from pleaded facts, and not evidence of bias. Counsel relies on Hassan Omar Hassan & another v Independent Electoral & Boundaries Commission & 2 others [2018] KEHC 8854 (KLR), where the High Court held that an application amounting to “a critique of the decisions of the Court which have gone against him” is properly pursued in “the appellate arena”, and that “the decisions of the Court per se do not amount to evidence of bias.” 34. Counsel places particular reliance on Kidero & 13 others v Ethics and Anti-Corruption Commission [2023] KEHC 3363 (KLR), citing with approval Attorney General v Anyang' Nyong'o & Others [2009] IEA 21, for the observation that: “ The court must guard against litigants who all too often blame their losses in court cases on bias on the part of the Judge. Success or failure of the government or any other litigant is neither ground for praise nor for condemnation of a court. What is important is whether the decisions are good in law, and whether they are justiable in relation to the reasons given for them. There is a fundamental tendency for the decisions of the courts with which there is a disagreement to be attacked by impugning the integrity of the Judges, rather than by examining the reasons for the judgment......” and that “a 'suspicious mind' in the literal sense will suspect even where no cause for suspicion exists and unfortunately this is a common phenomenon among unsuccessful litigants and that is why the mind envisaged in the test to determine perception of possible or likely bias on the part of the Judge is a reasonable, fair and informed mind. While litigants have the right to apply for the recusal of judicial ocers where there is a reasonable apprehension that they will not decide a case impartially, this does not give them the right to object to their cases being heard by particular judicial https://new.kenyalaw.org/akn/ke/judgment/keelrc/2026/2228/eng@2026-07-31 7

page 7 of 16

ocers merely because they believe that such persons will be less likely to decide the case in their favour.” 35. Counsel also cites this Court's own decision in Macharia & 13 others v Gichana & another [2022] KEELRC 12809 (KLR), where it was observed that: “ ……Applications for the court to recuse itself have become increasingly fashionable of late, regrettably often with no factual or legal justication. It may be tempting for a client to want to recuse the court when he perceives his case if failing, but that is no justication for counsel to make the application … it is for counsel to satisfy himself that there are reasonable grounds for making such an application.” 36. On the strength of the foregoing, and on the further facts deposed to in the Replying Adavit of Dr. Jamleck Muturi John — including that the Applicant, upon delivery of the Ruling of 4th June 2026, sought and was granted leave to appeal against it but elected instead to pursue recusal, and that neither the petitioner in the related ELRC Petition No. E022 of 2025 nor his counsel has taken issue with the analogous nding made in that matter — the 1st Respondent urges that the Applicant has failed to place before the Court any objective facts capable of establishing actual bias or a reasonable apprehension of bias, and that the Application, being an abuse of the court process brought to delay the Petition and to secure a judicial ocer perceived to be more favourable, ought to be dismissed with costs. 37. It is right to record that the 2nd, 3rd and 4th Respondents, and the Intended Interested Party, did not le any response to, or otherwise participate in, this Application, and that the determination that follows has therefore been arrived at on the basis of the submissions of the Petitioner/Applicant and the 1st Respondent alone. E. Issues for Determination 38. Having considered the Application, the grounds advanced, and the rival submissions summarised above, I am satised that the following issues arise for determination: (a) what is the nature and proper scope of the judicial function in an adversarial system such as ours, and what bearing does that function have on how allegations of bias are to be assessed; (b) why, as a matter of sound principle, ought applications for recusal to be brought only sparingly, and what are the likely eects, in the larger scheme of the administration of justice, of such applications being brought too readily; (c) what is the applicable law on recusal, whether recusal follows automatically upon a party moving for it, and in what circumstances a court may or should recuse itself; (d) whether, on a critical assessment of the grounds advanced, the Applicant has demonstrated a proper basis for the recusal of the presiding Judge in this matter; and (e) what orders should issue in the result, including as to the delivery of judgment and costs. F. The Role of a Judge in an Adversarial System 39. It is necessary, before descending into the specic grounds advanced in this case, to pause and reect on the nature of the oce whose occupant is here accused, and on the character of the system within which that oce is exercised, for it is only against that background that the present Application can be properly understood and fairly assessed. https://new.kenyalaw.org/akn/ke/judgment/keelrc/2026/2228/eng@2026-07-31 8

page 8 of 16

40. Kenya, like the great majority of common law jurisdictions, operates an adversarial, rather than an inquisitorial, system of civil and constitutional adjudication. Under that system, it is the parties, and not the court, who frame the dispute: they identify the issues, marshal the evidence, and advance the arguments; the judge's function is to adjudicate between the positions so presented, applying the law to the facts as pleaded and proved, and to decide, as between the parties, who is right and who is wrong on each question properly placed before him. That is not a peripheral or optional feature of the oce; it is its very essence. A judge who declined to decide, or who decided every close question by splitting the dierence so as to avoid ever ruling squarely against either side, would not be discharging the judicial function at all. 41. It follows, ineluctably, from the nature of that function, that in the ordinary course of any protracted litigation — and this Petition, spanning multiple interlocutory applications over more than a year, is protracted by any measure — a court will be called upon, repeatedly, to rule for one party and against the other. Where, as here, a litigant has himself invoked the jurisdiction of the court through successive applications — rst for conservatory orders, then for withdrawal — each application necessarily invites a ruling either in his favour or against him. There is nothing irregular, still less anything suggestive of bias, in a litigant experiencing a run of rulings that do not go his way; it is, on the contrary, a perfectly ordinary incident of adversarial litigation, especially where, as the record here shows, the same litigant has in fact also succeeded on at least one substantial and consequential question — the jurisdiction of this Court to hear his own Petition at all. 42. Nor is it any answer to say that a judge ought to moderate the language of his rulings, or decline to record clear ndings on issues squarely raised by the parties, so as to avoid the risk that a disappointed litigant may later characterise rm ndings as evidence of prejudgment. An adversarial system depends, for its integrity, on judges being willing to decide, plainly and without equivocation, in favour of one party and against another, on the basis of the law, the pleadings and the record before them — including, where the parties have themselves put in issue a litigant's good or bad faith, to make a clear nding on that very question. A judiciary too anxious to appear even-handed by softening its ndings, or by declining to preside further once a pattern of adverse rulings has emerged against a particular litigant, would not be more impartial; it would be less so, having substituted a concern for appearances over a concern for the correct and honest application of the law to the facts of each application as it arises. 43. It is against this understanding of the judicial function — an adjudicator bound to decide each application as framed by the parties, according to law, and without regard to whether the same litigant has previously succeeded or failed before the same court on other, distinct applications — that the propriety of the present Application for recusal must be assessed. G. Why Recusal Applications Should Be Brought Sparingly 44. I turn next, and I do so in the same courteous spirit in which the Applicant has, through counsel, expressly disclaimed any suggestion that this Application is brought out of mere dissatisfaction, to the question why recusal applications ought, as a matter of sound principle and practical wisdom, to be brought only sparingly, and why parties are well advised to reect carefully before moving a court to recuse itself. 45. It is important, at the outset, to arm what is not in question: every litigant is entitled, as of right, to raise a genuine, well-founded concern about a judge's impartiality, and no court should regard the mere raising of such a concern as itself an aront requiring reproach. The right to a fair hearing before an independent and impartial tribunal, guaranteed by Article 50(1) of the Constitution, would be a hollow https://new.kenyalaw.org/akn/ke/judgment/keelrc/2026/2228/eng@2026-07-31 9

page 9 of 16

guarantee indeed if litigants were discouraged, whether by express sanction or by subtle disapproval, from invoking it where a genuine basis exists. I record, expressly and for the avoidance of any doubt, that I do not regard the bringing of this Application, in itself, as improper, still less as an act of disrespect to this Court; it is the merits of the Application, and not the fact of its having been brought, that fall to be assessed. 46. That said, recusal applications are not, and ought never to be treated as, a routine tactical option available to a party dissatised with the trend of rulings against it. Public interest litigation of the kind pursued in this Petition, and litigation as procedurally eventful as this one has been — an ex parte order, an inter partes ruling on jurisdiction and on the conservatory application, a withdrawal sought and refused, and now a recusal application — will almost inevitably produce a sequence of rulings, some or many of which go against one party more often than the other, depending on the relative merit of that party's successive positions. If every such sequence were treated as, of itself, a legitimate springboard for a recusal application, few judges presiding over any protracted, multi-application public interest matter would ever see it through to judgment; parties would be able to engineer a change of judge simply by accumulating a run of adverse, but individually correct, rulings. 47. The likelihood of such accusations arising is not small; it is, if anything, one of the more common hazards of judicial oce in an adversarial system, precisely because every ruling that goes against a party carries with it the seed of a grievance that can be, and sometimes is, reframed as an accusation of bias rather than accepted as an adverse assessment on the merits. It is precisely because such accusations are so readily available to any dissatised litigant that courts, across jurisdictions, have insisted on a rigorous, objective threshold before recusal will be granted — a threshold I address in the next section of this Ruling. 48. The eects, in the larger scheme of the dispensation of justice, of recusal applications being brought too readily and granted too easily are serious and worth stating plainly. They encourage a form of judge-shopping, in which litigants seek, not a fair hearing before whichever judge is properly seised of a matter, but a favourable hearing before a judge of their own eective choosing, achieved by discarding any judge whose rulings prove unwelcome. They multiply delay and cost, since a change of judge in a matter with as extensive a record as this Petition now has would require the incoming judge to familiarise himself afresh with more than a year of pleadings, rulings and submissions, at real cost to the parties, to the public interest this Petition purports to serve, and to the timely administration of justice. They risk rewarding, rather than discouraging, tactical litigation conduct, by allowing a litigant who has properly lost applications on their merits to achieve, through a recusal application, what could not be achieved through the ordinary course of the litigation or, where available, through appeal. And they risk normalising the corrosive notion that a judge's impartiality may be measured by the frequency with which he rules in a particular litigant's favour — a notion entirely at odds with the nature of adjudication described in the preceding section of this Ruling. 49. None of this is to say that the Applicant's present Application is brought in bad faith; I do not nd that it is, and I say so expressly, for the avoidance of any doubt, in the concluding section of this Ruling. It is only to say, with the greatest courtesy but with equal rmness, that parties confronted with a sequence of adverse rulings should pause carefully, and consult authority and counsel closely, before concluding that the correct or proportionate response is an application for the recusal of the presiding judge, as opposed to the ordinary remedies of appeal or review that the law provides for challenging rulings considered to be wrong on their merits — remedies which, as the record in this very matter shows, the Applicant has himself already invoked by seeking, and being granted, leave to appeal against the Ruling of 4th June 2026. https://new.kenyalaw.org/akn/ke/judgment/keelrc/2026/2228/eng@2026-07-31 10

page 10 of 16

50. It is with that caution, oered as much for the guidance of future litigants before this Court as for the disposal of the present Application, that I turn to the settled law on when recusal is, and is not, warranted. H. The Law on Recusal: No Automatic Recusal 51. It is a rst and fundamental principle, common to every authority cited by either party and to every jurisdiction whose learning I have considered, that a court does not recuse itself automatically upon a party moving it to do so. Recusal is not, and has never been treated as, a matter of mere request granted as of course; it is, as the Supreme Court held in Rawal & 2 others v Judicial Service Commission & 2 others [2016] KESC 1 (KLR), “indeed a judicial duty”, one discharged by the very judge whose impartiality is impugned, applying settled legal principle to the facts advanced, and reached only where those principles and facts justify that conclusion. 52. The reason for this is straightforward, if easily overlooked. If recusal followed automatically upon request, the decision whether a particular judge should hear a particular matter would pass, in substance, from the judiciary to the litigants — precisely the mischief of judge-shopping already discussed. That is why the Supreme Court, in Shollei v Judicial Service Commission & another [2018] KESC 42 (KLR), was at pains to arm the “doctrine of the duty of a judge to sit”, holding that “recusal should not be used to cripple a judge from sitting to hear a matter”, and that a judge, having taken the judicial oath, “is capable of rising above any prejudices, save for those rare cases when he has to recuse himself.” 53. The objective standard by which those rare cases are identied is settled, and was correctly stated by both parties in their submissions, albeit for dierent ends. It was expressed by the English House of Lords, and adopted throughout the Commonwealth, in Porter v Magill [2002] 2 AC 357, as the question whether “the fair-minded and informed observer, having considered the facts, would conclude that there was a real possibility of bias”, a formulation this Court's own sister court applied in Rono v Almer Farm Limited [2025] KEELC 4403 (KLR), and one to which the Kenyan Court of Appeal gave further content in Philip K. Tunoi & Another v Judicial Service Commission & Another [2016] eKLR, holding that “in considering the possibility of bias, it is not the mind of the Judge which is considered but the impression given to reasonable people.” 54. That same test was restated, in materially identical terms, by the East African Court of Justice in Attorney General of Kenya v Prof Anyang' Nyong'o & 10 Others, EACJ Application No. 5 of 2007, as adopted by our own Court of Appeal in Kaplana H. Rawal v Judicial Service Commission & 2 others [2016] eKLR: “do the circumstances give rise to a reasonable apprehension, in the mind of the reasonable, fair minded and informed member of the public that the judge did not (will not) apply his mind to the case impartially”, a test the Supreme Court in Rawal & 2 others v Judicial Service Commission & 2 others further glossed, by reference to Jasbir Singh Rai & 7 Others v Tarlochan Singh Rai & Others, as one of “perception of fairness, of conviction, of moral authority to hear the matter.” 55. It is critical to underline what that formulation, consistently repeated across every authority cited by the parties, does not ask. It does not ask whether the litigant himself sincerely believes the judge to be biased; both the Supreme Court in Kaplana H. Rawal and the High Court in Kidero & 13 others v Ethics and Anti-Corruption Commission [2023] KEHC 3363 (KLR) — the latter citing Attorney General v Anyang' Nyong'o & Others [2009] IEA 21 — were emphatic on this point, the latter authority observing that “the court must guard against litigants who all too often blame their losses in court cases to bias on the part of the Judge”, that “success or failure of the government or any other litigant is neither ground for praise or for condemnation of a court”, and that “a 'suspicious mind' in the https://new.kenyalaw.org/akn/ke/judgment/keelrc/2026/2228/eng@2026-07-31 11

page 11 of 16

literal sense will suspect even where no cause for suspicion exists and unfortunately this is a common phenomenon among unsuccessful litigants.” The mind to be consulted is that of the reasonable, fairminded and informed observer, not that of the disappointed party. 56. Nor does an adverse judicial nding, without more, satisfy that objective standard, however rmly it may be expressed. The Rono decision, itself following Philip K. Tunoi & Another, records the grounds recognised under Regulation 21 Part II of the Judicial Code of Conduct and Ethics as warranting recusal where a judge: “ a. 57. Is a party to the proceedings. b. Was or is a material witness in the matter. c. Has personal knowledge of disputed evidentiary facts. d. Has actual bias or prejudice concerning a party. e. Has personal interest or is in a relationship with a person who has a personal interest in the outcome of the matter. f. Had previously acted as counsel for a party in the same matter. g. Is precluded from hearing the matter on account of some other sucient reasons. h. A member of his family has economic or other interests in the outcome of the matter in question.” It will be observed that every one of those grounds looks to a source of partiality extraneous to the judge's proper adjudication of the matter before him — a personal relationship, a nancial interest, prior involvement as counsel, personal knowledge of disputed facts acquired outside the proceedings. None of them is satised merely by the fact that the judge has, in the ordinary exercise of his adjudicative function upon issues the parties themselves placed before him, reached and recorded a rm conclusion adverse to one of them. That, indeed, is precisely the distinction drawn by the Supreme Court of the United States in Liteky v United States, 510 U.S. 540 (1994), in a passage this Court nds equally persuasive and consistent with the reasoning of our own courts, holding that “judicial rulings alone almost never constitute a valid basis for a bias or partiality motion”, and that opinions a judge forms on the basis of facts introduced or events occurring in the course of the very proceedings before him: “ ...do not constitute a basis for a bias or partiality motion unless they display a deep-seated favoritism or antagonism that would make fair judgment impossible.” 58. This is the doctrine sometimes described as the requirement of an “extrajudicial source”: absent some inuence originating outside the four corners of the litigation itself, a judge's assessment of a party's conduct or credibility, however rmly expressed, formed in the course of deciding the very matters properly before him, is not evidence of disqualifying bias; it is simply adjudication. The English Court of Appeal reached the same position in Locabail (UK) Ltd v Bayeld Properties Ltd [2000] QB 451, holding that the fair-minded and informed observer test is not satised merely because a judge has, in the course of current or related proceedings, made ndings adverse to a litigant; something more — some extraneous inuence, interest or association bearing on the judge's capacity to approach the matter with an open mind — must be shown. The Constitutional Court of South Africa, in President of the Republic of South Africa & Others v South African Rugby Football Union & Others 1999 (4) SA 147 (CC), articulated the same safeguard by insisting on a double reasonableness — both the person apprehending bias and the apprehension itself must be reasonable — holding that “the test for apprehension of bias must be a reasonable one, held by reasonable and right minded persons, applying themselves to the question and obtaining thereon the required information.” 59. It is instructive, nally, to recall the origin of the appearance-of-bias principle itself, in the celebrated statement of Lord Hewart CJ in R v Sussex Justices, ex parte McCarthy [1924] 1 KB 256, that “justice should not only be done, but should manifestly and undoubtedly be seen to be done.” That https://new.kenyalaw.org/akn/ke/judgment/keelrc/2026/2228/eng@2026-07-31 12

page 12 of 16

statement, properly understood and consistently applied by the later courts cited above, is directed at circumstances extraneous to the merits of the case before the court — in that case, an undisclosed conict on the part of the clerk advising the justices. It was never intended, and has never been applied, to mean that justice is not “seen to be done” merely because a court has, in the ordinary course of a protracted and procedurally eventful matter, made ndings against the same litigant on more than one occasion, upon issues the parties themselves placed before it. 60. There remains the question of when, consistently with these principles, a judge may or should recuse himself. The answer, drawn from the authorities above and from this Court's own decision in Macharia & 13 others v Gichana & another [2022] KEELRC 12809 (KLR), is that a judge should recuse himself where an objective, fact-specic basis of the kind described in the preceding paragraphs is shown to exist — a disqualifying interest, a close personal or extrajudicial association bearing on the case, or conduct or language demonstrating a deep-seated favouritism or antagonism transcending ordinary adjudication — and equally, out of respect for the “sanctity of proceedings” and the “integrity and dignity” of the judiciary spoken of in Rawal & 2 others v Judicial Service Commission, where continuing to sit would in fact undermine rather than serve public condence in the impartial administration of justice, as Gacheru, J. recognised in Thugi River Estate Limited & another v National Bank of Kenya Ltd & 2 others [2014] KEHC 8278 (KLR) when he elected to recuse himself even upon a nding that the apprehension raised before him was “far-fetched”, so as to give the applicant in that case “the condence to approach the Court for arbitration of this matter.” 61. A judge should not, however, recuse himself where, as this Court's own decision in Macharia & 13 others v Gichana & another cautions, the application is simply one of the increasing number brought “with no factual or legal justication”, tempting though it may be for “a client to want to recuse the court when he perceives his case is failing.” To do so would not protect the integrity of the judiciary; it would reward exactly the kind of judge-shopping described earlier in this Ruling, and would, moreover, be inconsistent with the “duty to sit” armed in Shollei, and in Hassan Omar Hassan & another v Independent Electoral & Boundaries Commission & 2 others [2018] KEHC 8854 (KLR), the latter holding that an application amounting to no more than “a critique of the decisions of the Court which have gone against” the applicant is properly ventilated “in the appellate arena”, and not by way of recusal. 62. Drawing these authorities together — Kenyan and comparative alike — the following propositions may be taken as settled: recusal is not automatic upon request, but requires a judicial determination against an objective legal standard; that standard is whether a reasonable, fair-minded and informed observer, and not the sincere but subjective apprehension of the party raising the objection, would conclude that there is a real possibility of bias; a judge should recuse himself where a disqualifying interest, extraneous association, or deep-seated favouritism or antagonism is shown, or where the interests of public condence in the administration of justice otherwise so require; but a judge should not recuse himself merely because adverse ndings, however rmly expressed, have been made against a litigant in the ordinary and proper exercise of the adjudicative function upon issues the parties themselves placed before the court. 63. It follows, and I state this proposition plainly for the benet of the parties in this and any related cause, that a court is not obliged to recuse itself, and ought not to do so, merely because it is asked. It is against that settled standard, and not against the Applicant's own understanding, if in my nding ultimately mistaken, sense of grievance, that the grounds advanced in the present Application must now be tested. https://new.kenyalaw.org/akn/ke/judgment/keelrc/2026/2228/eng@2026-07-31 13

page 13 of 16

I. Judicial Candour in Interlocutory Rulings 64. Before undertaking that assessment, it is necessary to address a related but distinct point, because it bears directly on how the language of the Ruling of 4th June 2026 ought properly to be read. A court engaged in determining an interlocutory application — including an application, as here, in which one party has itself expressly put in issue whether its opponent's conduct amounts to bad faith or an abuse of process — is not merely entitled, but often duty-bound, to state its ndings on that issue in clear and unambiguous terms. 65. A court should not be shackled from strongly expressing itself on an interlocutory question where the circumstances of the matter genuinely require it. Litigants are entitled to know, in plain language and not in euphemism, what a court has found and why; equivocal or deliberately softened language, adopted merely to avoid giving oence, serves neither the parties nor the wider public interest in transparent adjudication, and would, if anything, invite the very suspicion it is designed to avoid, by leaving litigants to speculate as to what a court truly meant. 66. That principle, however, carries an important corollary, and it is one to which this Court has been, and remains, scrupulously attentive: rmness of expression must be, and remain, tethered to the record. A nding that “forum shopping is evident” or that “the court process is being abused” is legitimate, however forcefully stated, where it is reached upon a fair evaluation of the pleadings, the chronology of related proceedings, and the submissions properly placed before the court by the parties themselves — as, in the present case, it plainly was, the 1st Respondent having expressly framed bad faith and abuse of process as an issue for the Court's determination. What the law will not countenance, and what would properly found a recusal application, is rmness of expression that draws upon matters extraneous to that record — private knowledge, personal antagonism, or considerations having nothing to do with the case as pleaded and argued. 67. It is that distinction, between robust fact-based adjudication on the one hand, and impermissible extrajudicial antagonism on the other, that the doctrine of the “extrajudicial source” discussed in the preceding section is designed to capture. A court that speaks plainly, on the strength of the record and in answer to an issue the parties themselves have raised, has not thereby disqualied itself; it has simply done its duty. 68. This Court accordingly does not resile, and sees no principled reason to resile, from the ndings made in its Ruling of 4th June 2026. Those ndings were made upon a considered evaluation of the chronology of this Petition — including the timing of the withdrawal application relative to the vacation of the ex parte conservatory orders, and its relationship to proceedings pending in another court — and upon the specic issue of bad faith and abuse of process that the 1st Respondent had itself placed before the Court. They were, in short, an exercise of the ordinary judicial function, robustly performed; they were not, and are not, evidence of bias. J. Critical Analysis of the Grounds Advanced 69. I turn, then, to a critical assessment of the specic grounds advanced by the Applicant, mindful throughout that the question is not whether the Applicant sincerely holds the apprehension he professes — I have no reason to doubt that he does — but whether a fair-minded and informed observer, apprised of the whole record and not merely of the extracts most favourable to the Applicant's case, would conclude that there is a real possibility of bias. 70. First, and centrally, the Applicant's case rests substantially on the ndings made in the Ruling of 4th June 2026. But an observer apprised of the whole record would also recall the Ruling of 29th https://new.kenyalaw.org/akn/ke/judgment/keelrc/2026/2228/eng@2026-07-31 14

page 14 of 16

January 2026, in which this Court, over the determined objection of all four Respondents, held that it possessed jurisdiction to entertain this very Petition and the constitutional question at its heart — a ruling squarely in the Applicant's favour on the single most consequential question then before the Court, since a contrary nding would have terminated the Petition in limine. A judge harbouring the antagonism now alleged against him would have had every opportunity, and no shortage of respectable legal argument advanced by the Respondents, to dispose of this Petition adversely to the Applicant at that earlier stage; that the Court instead ruled for him on jurisdiction, and against him only on the narrower and distinct question of the continuation of ex parte orders, sits uneasily with the suggestion that the Court has, across these proceedings, evinced a settled disposition against him. 71. Second, the ndings of “forum shopping” and “abuse of the court process” of which complaint is made were not free-oating pronouncements volunteered by the Court; they were the Court's answer to an issue the 1st Respondent had itself expressly framed for determination — whether the withdrawal application was made in bad faith and constituted an abuse of process — argued on the pleadings and submissions of both sides. A court that answers, one way or the other, a question the parties themselves have placed before it has not thereby demonstrated bias; it has discharged its adjudicative function. Were the rule otherwise, no court could ever determine an allegation of bad faith or abuse of process without thereby disqualifying itself from the remainder of the case in which that allegation was raised — a proposition that would paralyse the ordinary conduct of interlocutory litigation. 72. Third, the Applicant places weight on the fact that a similarly-framed withdrawal application was declined in the related ELRC Petition No. E022 of 2025, Peter Kibelesi Kukubo v Teachers Service Commission, National Assembly & 2 Others, submitting that this demonstrates a “settled view” on the Court's part regarding the related proceedings generally. On reection, however, that very fact points the other way. A judge personally antagonistic toward this Applicant would have no discernible motive to rule, on a materially analogous application, against an entirely dierent litigant, in a separate cause, represented by dierent counsel. The consistency of outcome across two unrelated litigants is more naturally explained by the even-handed application of a single legal principle to materially similar facts than by personal antagonism directed at this Applicant specically — and it is notable, as the 1st Respondent's Replying Adavit records without contradiction, that the petitioner in that related matter has himself raised no complaint of bias, notwithstanding an outcome identical in substance to the one of which this Applicant now complains. 73. Fourth, the Applicant's own Supporting Adavit and submissions candidly, and to his credit, disclaim reliance on the adverse rulings as such, resting instead on the “denitive ndings” made concerning his motives and conduct. But that concession, properly understood, brings the Application squarely within the “extrajudicial source” doctrine discussed above: the ndings of which complaint is made were reached, on the Applicant's own account, entirely upon the record of these very proceedings — the pleadings, the chronology of the withdrawal application, and the submissions of the parties — and not upon any matter extraneous to it. No pecuniary or proprietary interest, personal relationship, prior involvement as counsel, or extrajudicial knowledge on the part of the presiding Judge has been alleged, still less demonstrated. On the authorities reviewed in Section H above, that is fatal to the Application. 74. Fifth, it is a matter of record that upon delivery of the Ruling of 4th June 2026, the Applicant sought, and was granted, leave to appeal against it — the ordinary and constitutionally appropriate remedy for a litigant who considers a ruling to be wrong on its merits. Rather than pursue that remedy, the Applicant elected to bring the present Application instead. While I do not nd, and do not suggest, that the Application was brought in bad faith, its eect, whatever the intention behind it, would be to achieve by way of recusal what the law more properly commits to the appellate process: a fresh look at ndings the Applicant considers erroneous. That is precisely the substitution of recusal for appeal https://new.kenyalaw.org/akn/ke/judgment/keelrc/2026/2228/eng@2026-07-31 15

page 15 of 16

against which Hassan Omar Hassan & another v Independent Electoral & Boundaries Commission & 2 others cautions. 75. Sixth, and nally, I have considered whether the cumulative eect of the matters relied upon — even if none, individually, meets the threshold — might nonetheless, in combination, satisfy the fairminded and informed observer test. Having done so, I am satised that it does not. The cumulative narrative advanced by the Applicant is, in substance, a narrative of adverse rulings, correctly made upon the record and upon issues the parties themselves raised, in a Petition that has also produced at least one substantial ruling in the Applicant's favour, and whose only common thread across the related proceedings is the even-handed application of settled principle to more than one litigant. That narrative, however sincerely felt, does not disclose a real possibility of bias; it discloses the ordinary, sometimes unwelcome, texture of adversarial litigation conducted according to law. 76. For all the foregoing reasons, I nd that the Applicant has failed to establish any ground upon which this Court could properly recuse itself from further hearing, determination, and delivery of judgment in this Petition. K. Disposition 77. 78. For the foregoing reasons, I make the following orders: (a) The Petitioner/Applicant's Notice of Motion dated 15th June 2026, seeking the recusal of the presiding Judge and the reallocation of this Petition to another Judge, is hereby dismissed. (b) The prayer for a stay of delivery of judgment pending the hearing and determination of the Application is spent, this Ruling having now nally disposed of the Application. (c) The presiding Judge shall proceed to prepare, and, in due course, deliver judgment in this Petition. (d) The costs of this Application shall be borne by the Petitioner/Applicant. It is so ordered. DATED, SIGNED AND DELIVERED AT MOMBASA THIS 31st DAY OF JULY, 2026. OCHARO KEBIRA JUDGE https://new.kenyalaw.org/akn/ke/judgment/keelrc/2026/2228/eng@2026-07-31 16

page 16 of 16