Every litigator carries some version of Justice Felix Frankfurter’s warning from the 1939 US Supreme Court decision of Nardone et al. v. United States, 308 U.S. 338: that a tainted process poisons everything it produces; “the fruit of the poisonous tree.” The metaphor is drawn from the law of evidence, where it describes the inadmissibility of illegally obtained evidence, but it captures a broader principle: where the source is adulterated, anything that flows from it may also be tainted.
Constitutional remedies, however, are rarely so straightforward. As public interest jurisprudence evolves, situations may occur where invalidating a flawed government action risks disrupting ongoing public service delivery. In such a case, the “fruit” is preserved despite the flaws in the process that gave rise to it. This presents an important question: must an administrative outcome that is made through a constitutionally flawed procedure inevitably fall with it or can a judge preserve the flawed outcome on account of public interest while requiring prospective compliance? And if a judge chooses the latter, does preserving the outcome excuse the court from ever evaluating whether that outcome could be justified on the merits in the first place?
On 20 August 2026, the Milimani High Court [“the Court”] (Mugambi, J.) grappled with that same question in Nyabuto v Kenya Academy of Sports & another [2026] KEHC 13007 (KLR). The court found that the Respondents did not meet the test of public participation in the process of formulating the criteria for identification and selection of Sports Academies. Yet rather than quashing the whole procurement process, the court proceeded to give flexible preservation orders under article 23(3) of the Constitution, on the ground that the already piloted and established academies were “operational and presumably serving the public interest” (para 93).
By preserving the operationalisation of the already challenged academies and granting the Respondents a latitude to develop, within 3 months, guidelines for future establishment of Sports Academies, this decision sparks debate on whether post regularisation risks converting public participation into the dreaded fear of being a “mere formality”. That risk has a sharper edge here than the general debate admits. The harder question the petition raised was not whether the process for setting future criteria was fair; it was whether the sites already chosen were fair, and specifically whether a home county with 56 registered athletes was ever entitled to four academies while a county with 229 received none. That question was pleaded, argued, and never separately decided. It is the throughline of everything that follows in this piece.
Brief summary
The petition was filed under article 258(2) of the Constitution, urging public interest. The Petitioner, Jared Ngisa Nyabuto, raised several complaints against Kenya Academy of Sports, 1st Respondent, a state corporation established under the Sports Act 2013 and the Ministry of Youth Affairs, Creative, Economy and Sports (2nd Respondent). He cited the lack of public participation in the selection and allocation of academies and failure by the Respondents to engage all relevant stakeholders including local communities, education officials and youth organisations, in a meaningful manner. He challenged the uneven and unjustified distribution of academies across various counties stating that it was politically motivated, thereby violating the principles demanded by article 10 and article 27. Besides, the Petitioner pointed to stark factual inconsistencies and distorted resource allocation. For instance, in Elgeyo Marakwet, the home county for the then Cabinet Secretary for Sports, with only 56 registered athletes, increased its number of academies from one to four (one per constituency) via a readvertised tender notice and without any feasibility study. On the contrary, counties with the highest talent concentrations such as Mombasa (with 229 athletes) and several other high-registration counties were completely excluded from the pilot roll out.
The petitioner also argued that the Respondents breached article 35 of the Constitution by failing to avail the initial tender notice despite his demand issued in December 2024 and failing to provide information about the distribution of the pilot academies. Furthermore, he alleged that the procurement process was carried out in a manner that was illegal and contrary to the laid down procurement processes. Contrastingly, the Respondents contended that the roll out of the academies was gradual and in phases, informed by the economic costs involved. They maintained that they had conducted feasibility studies and engaged key institutional stakeholders (discussed further below). On the unjustifiability of academy distribution, they submitted that their decision was informed by questions of demand and supply. They submitted that site selection followed objective criteria such as land suitability and technical feasibility. They denied that they violated the principles of Access to Information Act and contended that the petition was premature as the roll out of academies was still ongoing having planned to allocate academies in each constituency, in the next ten years.
Of the four distinct grounds the petitioner pleaded, including failure of public participation, discriminatory and politically skewed allocation, breach of the Article 35 right to information, and breach of procurement law, he succeeded on exactly one, and only in part. Ultimately, the court ruled that the Respondents had failed to demonstrate both qualitative and quantitative public participation, violating article 10 of the Constitution. The court noted that there was no evidence to demonstrate that the “general public” was involved beyond the list of officialdom during the consultations (para 66). Crucially, the court rejected the Petitioner’s invitation to have the procurement process conducted afresh and instead opted to issue preservation orders under article 23(3) The court held that the lack of public participation did not mean that the public sport academies already piloted and established were unconstitutional and instead instructed the Respondents to formulate guidelines governing the selection and determination of locations for establishment of Sports Academies within 3 months, subject to public participation before adoption.
The court also declined to fault the procurement process on its own terms. Since the tenders had been openly advertised and re-advertised, and no bidder was shown to have been excluded or unfairly disqualified, the Court found nothing on the record to displace the presumption that an open, competitive process satisfies article 227 (the Public Procurement and Asset Disposal Act, 2015) (para 91). The procurement ground therefore failed alongside the article 35 ground, leaving public participation as the sole basis on which the Petitioner succeeded. Notably, rather than scheduling a formal supervisory return date to monitor compliance, the court affixed a self-executing default clause: “In default of compliance with order 1 & 2 above, any member of the public shall be at liberty to Petition this Court for such further orders as may be just and appropriate.“ (order 3, para 95) By doing so, the efficiency of the constitutional remedy was shifted from the judiciary to the public.
The remainder of this piece critically examines that reasoning ground by ground, starting with what counts as public participation at all, before turning to the remedy the court fashioned and the discrimination question it left untested.
Institutional consultation is not public participation
The Respondents defended the selection process and stated that they had conducted feasibility studies and engaged various stakeholders including county governors, elected members of county and executive committee members, land officers, surveyors, physical planners and the boards of management for the schools. The court was not satisfied that these consultations demonstrated sufficient public participation. Reaffirming the Supreme Court ratio in British American Tobacco Kenya Limited v Cabinet Secretary for the Ministry of Health & another (SC Petition 5 of 2017) (Judgment) (26 November 2019) (para 61) the court restated the guiding principles for public participation noting that public consultations are essential and go to the constitutional tenet of sovereignty of the people.
That holding carries weighty lessons for ministries and state corporations; officialdom alone does not constitute the “general public”. Public institutions that consult with other public offices may later discover that they have conducted institutional consultation rather than public participation. Technical officers may have the information needed to make decisions, and elected officials may be the representatives of the people. However, technical consultation and representation should not be used in place of public participation. The public or impacted communities ought to be informed about the proposed decision and given an opportunity to influence it.
Feasibility studies versus public participation
The judgment also offers a distinction between technical feasibility studies and public participation. Feasibility studies and public participation perform different functions. The first checks whether a project is viable while the other asks whether the people affected have been given a reasonable opportunity to participate in the decision. In this case, the judge ruled that although the feasibility studies and selection criteria were useful in the identification process, they failed because they lacked the input of the public. The finding implies that a government institution may therefore conduct excellent technical studies but still fail to meet the requirement of public participation.
The issue becomes even more critical for projects, which undergo material revisions during implementation. In the present case, the petitioner made two distinct points about the feasibility studies. First, the studies were internally inconsistent on their own terms: commissioned in October 2023, they purported to cover twenty-six constituencies but, on the petitioner’s review, had actually been conducted in only twenty-five. Second, and separately, whatever the studies covered, it was not the eventual scope of the programme: the readvertised tender notices went on to advertise 37 academies across 23 counties, introducing additional sites in Kasarani, Budalangi, Kabuchai, Elgeyo Marakwet, Matuga, Thika Town, Isiolo South, Kitui East, Wajir and Mandera that had not undergone feasibility examination at all. The scope kept moving because the tender itself did: the roll-out went through three successive re-advertisements, on 20 September 2024, 2 December 2024, and 11 April 2025, each widening the list of sites beyond what the original feasibility studies had covered. Further, the petitioner asserted that the feasibility studies failed to provide clear evidence that the process followed the requirements of the Constitution and Article 10.
This raises another important rule for public project management: that public participation applies throughout the life of a project, its formulation, implementation, and completion. If the policy or factual foundation of a public decision is altered in a way that significantly impacts the parties concerned, it is good practice to offer the public another opportunity to participate. Participating in one proposition and then putting another into action violates Article 10.
Can prospective compliance cure previous illegality?
Another important question arising from the judgment is the transient nature of public participation. In paragraph 59, the learned judge states that public participation invokes a participatory model of decision making beyond the periodic electoral cycles. The court clarifies that public participation ensures that the input of the public (or those with stake in an issue) is sought early enough so that their perspective can shape the final outcome. Therefore, public participation should give citizens a voice on proposed decisions before they are implemented.
But the difficulty arises when a public decision is implemented while completely ignoring or inadequately conducting public participation and later faces a legal challenge. The question then becomes: can future participation regularise the previous constitutional flaw?
Article 23 (3) of the Constitution gives the High Court broader discretionary power to grant appropriate and effective reliefs. The court may follow several remedial routes. The first is to invalidate the whole process and order the process to begin afresh. This was a remedy that the court previously considered appropriate for procurement processes in Republic v Independent Electoral and Boundaries Commission (IEBC) Ex parte National Super Alliance (NASA) Kenya & 6 others, Judicial Review No 378 of 2017; [2017] eKLR (para 62). However, on appeal (Independent Electoral and Boundaries Commission (IEBC) v National Super Alliance (NASA) Kenya & 6 others, Civil Appeal No 224 of 2017; [2017]), the court quashed the de novo directive but emphasised that the foundational principles in article 10 (2) are immediately justiciable and enforceable rather than progressive or aspirational.
The second route is to return at the point in which the constitutional flaw might have occurred without undoing what preceded it. This would require the judge to view public participation as stage-to-stage process. The Supreme Court decision in Attorney General & 2 others v Ndii & 79 others [2022] KESC 8 (KLR) recognised that public participation can operate and be conducted at different stages of a continuing process.
The third, which the court adopted, is to preserve whatever has been done while requiring a public body to craft participatory guidelines for future roll outs. This strategy was adopted by Justice Mugambi in the judgement currently under analysis. Although practical, this remedy draws attention to a recurring structural conundrum in Kenyan constitutional law. A similar pattern was highlighted in 8 August 2026 The Mwango Law Review’s case note, (When the Mountain Labours: Reading the Fine Print of Owiso v Attorney General & IEBC) where the court found a law or an act unconstitutional yet refuse to give that finding retrospective effect.
Nyabuto is arguably a more thoroughgoing version of that same pattern. In Owiso, the court found real constitutional infirmity and then suspended its own declaration with no expiry date, so that the fire brigade was, in effect, told to arrive only after the building had already burned down and a replacement built on the ashes. In Nyabuto, no brigade was dispatched to this fire at all: Mugambi J. never got as far as declaring the underlying allocation unlawful in the first place, because, as set out above, the discrimination ground was subsumed into the public participation finding and left there. The respondents were told only to check the wiring before the next building goes up. The remedial instinct is identical in both cases, including finding a pending constitutional problem, then design an order that leaves the facts on the ground squarely as they stood before the petition was filed, even though the two disputes, one about when a nation votes and the other about where a sports academy gets built, look unrelated on the surface.
Notwithstanding, if prospective regularisation becomes the default remedy, public organisations may gain tactical incentive to expedite invalid projects to operational status because judges will be reluctant to impose invalidation once monies have been spent.
A workable dividing line might weigh two factors against each other: the cost of unwinding what has already been built, against the severity and apparent intentionality of the participation failure itself. A minor or inadvertent lapse in stakeholder engagement might tolerably be cured prospectively where reliance costs are genuinely high. A failure that looks less like an oversight and more like a considered decision to consult only officialdom, as arguably happened here, should not receive the same latitude merely because the project has since become operational.
Public interest as a justification
The court, having found that the selection process was flawed for lack of public participation, refused to give orders of certiorari stating that the existing academies were “operational and presumably serving public interest”. It considered it “absurd and unjust (para 93)” to grant the orders. This makes considerable sense. Courts should not issue remedies that disproportionately cause harm to the public. However, public interest must be employed with extreme caution. The fact that a project is beneficial to the public does not retrospectively make its inception or establishment constitutional. Furthermore, it may be conceptually problematic to refer to a public project as serving the public interest if it is established that the public was excluded from participation from the beginning.
Access to information versus Transparency
The judgment differentiates between access to information under Article 35 and transparency as a national value under Article 10 of the Constitution. In the present case, the petitioner argued that the Respondents violated article 35 since he was unable to trace the initial tender notice despite his demand. He also argued that the Respondents failed to provide information about the criteria used to unfairly distribute pilot academies and the procurement process used in the contracts. The court, in paragraph 92, ruled that the process of formulating the selection criteria and site locations lacked constitutional transparency; separately, at paragraphs 80-81 but rejected the Petitioner’s argument that his right to information was violated. This was because the Petitioner could not prove accurate addressing and delivery of the information request letter to the 1st Respondent. From this judgement, it therefore follows that a public institution may be criticised for failing to operate transparently without such failure amounting to a violation of Article 35.
The unaddressed discrimination questions
The judgment was silent on this issue. The Petitioner presented comparative statistics showing the disproportionate allocation of academies across multiple counties. For example, Elgeyo Marakwet with 56 registered athletes, received four academies while Mombasa with 229 athletes received none. The Respondents cited the case of EG & 7 others v Attorney General; DKM & 9 others (Interested Parties); Katiba Institute & another (Amicus Curiae) [2019] KEHC 11288 (KLR), which held that differentiation is permissible where it bears a rational connection to a legitimate governmental purpose. Nevertheless, the court failed to apply the rational-connection test or assess whether the skewed distribution met a constitutional objective.
That silence left a critical gap in the judgment. A public participation finding and a discrimination finding answer different questions: the first asks whether the public was consulted before the criteria were set, the second asks whether the criteria, however made, produced an outcome any rational decision-maker could defend. A respondent can lose on the first and still be entirely right on the second; or, as here, a court can find for a petitioner on the first while leaving the second, on far more specific and unfavourable facts, untested. Indeed, at paragraph 68, the Court itself considers the discrimination complaint as something that could have been avoided by better public participation, rather than as a separately fatal defect in the allocation: the discrimination ground was pleaded, briefed by both sides down to the applicable test, and entirely subsumed into the public participation finding without ever being separately decided. Uneven resource allocation is more than a policy decision; it violates the rules of non-discrimination under article 27 when attributed to political inclination rather than objective demand. In my view, on the numbers the Petitioner placed before the court, that test would have been difficult for the Respondents to satisfy: a rational connection to a legitimate purpose is hard to reconcile with a home county receiving four academies against no feasibility study, while counties with three to four times its registered athlete population received none.
Practical implications for advocates
Firstly, for advocates, this judgement rewards a challenge to projects that are perverse or manifestly irrational (see Judicial Service Commission v Mbalu Mutava & another, Civil Appeal 52 of 2014; [2015] KECA 741 (KLR)) at the earliest practicable opportunity, seeking conservatory relief before the impugned programme becomes operational rather than waiting to challenge it once implementation is complete. This gives the court more room to stop or rectify the process without causing public disruption that makes quashing complicated. Timing matters so much here because Nyabuto confirms that once a programme can be described as already “operational and…serving the public interest,” that description alone works against any later claim for invalidation, whatever the underlying allocation looked like.
Secondly, discrimination and public participation should be pleaded as genuinely separate causes of action, each under its own article and with its own prayer and evidentiary record, rather than left to ride on one another. Nyabuto shows how easily an Article 27 discrimination claim can be filed, briefed by both sides down to the applicable rational-connection test, and still never be decided, once a court is able to dispose of the petition on the easier public-participation ground first. Thirdly, advocates should insist on compliance or oversight mechanisms when the court grants positive or structural orders. They should avoid settling for a proclamation of rights or an order compelling a public body to act. Fourthly, when seeking relief under Article 35, advocates must ensure they address the request to a named, correctly addressed recipient at the appropriate authority, specify the information sought, keep a record of delivery, and preserve any response or lack thereof; Nyabuto was lost on this very ground, at paragraphs 80-81, for want of proof that the request letter had ever been properly addressed or delivered.
Fifthly, advocates alleging discriminatory or politically motivated distribution of a public resource should lead with solid, sourced comparator data, including population or need ratios, timing correlated to a change in officeholders, feasibility-study coverage, because a bare allegation of favouritism, however sound, will not on its own force a court to engage the rational-connection test.
Conclusion
Nyabuto is significant less for its public participation holding, which restates doctrine that has stood for close to a decade, than for its remedy: a court that found a pending constitutional defect under article 23(3) and chose to fix the process going forward while leaving the product of that defective process fully intact. The phrase “the fruits of the poisonous tree” posits that tainted fruit should be thrown away with the tree; this judgment does something more specific than either preserving the fruit or discarding it: it left the fruit standing without cutting it open to check.
Of the four grounds the Petitioner pleaded, he won exactly one, and only in part. The harder question, including whether Elgeyo Marakwet’s four academies were ever defensible against Mombasa’s zero, was pleaded, briefed by both sides, and simply absorbed into the public participation finding without being separately decided. That gap will not close itself, and Nyabuto is not the only recent decision to leave a constitutional finding without consequence for the parties before the court. Advocates who spot that pattern early, and plead accordingly, will do better than those who wait for a court to notice it for them.
Kennedy Baraka is a law graduate and Advocate-designate currently holding over while awaiting admission to the Kenyan Bar. He specialises in constitutional, public policy, and commercial law, with a passion for demystifying legal concepts for the public.