[2016] KECA 581 (KLR)

IN THE COURT OF APPEAL

AT NAIROBI

(CORAM: GITHINJI, OKWENGU, MWERA, GBM KARIUKI & AZANGALALA, JJ.A)

CIVIL APPEAL NO. 64 OF 2012

BETWEEN

REPUBLIC…………………….………...........................……APPELLANT

EXPARTE

THE MINISTER OF STATE FOR PLANNING

NATIONALDEVELOPMENTANDVISION 2030

THE DIRECTOR KENYA NATIONAL BUREAU FOR STATISTICS

AND

THE INTERIM INDEPENDENT BOUNDARIES

REVIEW COMMISSION…….………................................1stRESPONDENT

NOOR MAALIM HUSSEIN..……………..……….....…...2ndRESPONDENT

ABDI BUUHOW ABDI……………………………......…...3rdRESPONDENT

HASSAN HUSSEIN MOHAMED…………………......…..4thRESPONDENT

ABDULLAHI ABDI HUSSEIN……………………............5thRESPONDENT

GEOFFREY IKATUKON LOBOKOT………..…....…….6thRESPONDENT

AHMED MOHAMED NOOR……………………......……7thRESPONDENT

HAMDI ALI MOHAMED…………………………......…..8thRESPONDENT

MOHAMED ALI GURE………………………….....…….9thRESPONDENT

MOHAMED ALI ABDI………………………………......10thRESPONDENT

AND

THE COUNTY GOVERNMENT OF MANDERA}

THE COUNTY GOVERNMENT OF GARISSA }….INTERESTED PARTY

(Being an appeal from the Judgment and Order of the of the High Court of Kenya at Nairobi (Hon Mr. Justice Mohamed Warsame) delivered on 7thFebruary 2012 at Nairobi in HC (JR) Misc Civil App No. 309 of 2010)

BETWEEN

REPUBLIC………………………………………………..APPLICANT

EXPARTE

NOOR MAALIM HUSSEIN..……………..…………..}

ABDI BUUHOW ABDI……………….………………...}

HASSAN HUSSEIN MOHAMED…….………………..}

ABDULLAHI ABDI HUSSEIN……….……..……......}

GEOFFREY IKATUKON LOBOKOT...} APPLICANTS

AHMED MOHAMED NOOR HAMDI ALI MOHAMED INTERESTED PARTIES MOHAMED ALI GURE

MOHAMED ALI ABDI

JUDGMENT OF THE COURT

Introduction

[1] This appeal arises from judicial review proceedings that were initiated in the High Court by Noor Maalim Hussein, Abdi Buuhow Abdi, Hassan Hussein Mohamed, Abdullahi Abdi Hussein and Geoffrey Ikatukon Lobokot who are now 2nd, 3rd, 4th, 5th and 6th respondents respectively. Following leave granted by the High Court under Order 53 of the Civil Procedure Rules, the five respondents who described themselves as registered voters and residents of eight constituencies within the Republic of Kenya, sought orders of certiorari and prohibition against the Minister of State for Planning, National Development and Vision 2030, the Director of Kenya National Bureau of Statistics (now the 1st and 2nd appellants respectively), and the Interim Independent Boundaries Review Commission (now the 1st respondent). During the proceedings in the High Court four other persons Ahmed Mohamed Noor, Hamdi Ali Mohamed, Mohamed Ali Gure and Mohamed Ali Abdi were joined as interested parties. They are now 7th 8th 9th & 10th respondents, respectively to this appeal. During the hearing of this appeal theCounty Governments of Mandera and Garissa applied and were joined in this appeal as interested parties.

Background

[2] On the night of 24th/25th August 2009 at midnight the National Government commenced the national census and enumeration process pursuant to the Statistics (Census of Population) Order 2008, issued by 1st appellant. The exercise was undertaken by the 2nd appellant as provided by section 4 of the Statistics Act. For the first time in the history of our nation the exercise took place over a period of seven days ending on 31st August 2009. A year later, on 31st August 2010, the results were released. While releasing the results the 1st appellant announced that results from eight districts had been canceled. These included Lagdera, Mandera East, Mandera Central, Mandera West, Wajir East, Turkana North, Turkana South and Turkana Central.

[3] Under the Districts and Provinces Act, 1992, Kenya was divided into Nairobi Area and 46 administrative districts under the national government. By April 2010, the number of districts had been increased to a total of 267 through Executive fiat. The eight districts whose census results are subject of this appeal fell within that category. Under the current Constitution, the Constitution of Kenya 2010, the country is now divided into 47 counties, with governments at the county and national levels being distinct but interdependent. The eight districts now fall within 4 counties, that is Mandera, Garisa, Wajir and Turkana Counties and this explains the interest of the County Governments of Mandera and Garissa who successfully applied to be joined in this appeal. Although in the proceedings and judgment of the High Court, the eight districts are also sometimes referred to as constituencies, the constituency boundaries were not necessarily the same as district boundaries. Article 89 of the Constitution established two hundred and ninety constituencies, the boundaries of which were to be reviewed by the 1st respondent. At the time the litigation in this matter arose, the constituency boundaries had not been finally settled. We shall therefore refer to the eight districts whose results were subject of controversy as districts and not constituencies.

[4] On the same day that the census results were announced, the 1st appellant appeared before Parliament and issued a statement on the census results. In the statement, the 1st appellant explained that results from the eight districts were irregular and therefore, necessitating the cancellation. The irregularities included understatement of population in some areas, and overstatement of population in other areas. The 1st appellant further stated that the enumerated figures fell way above the projected figures that were based on the previous census (that is 1999) results. The 1st appellant therefore indicated that the Government wanted to repeat the exercise in the affected areas.

[5] It is this announcement by the 1st appellant cancelling the results of the eight district that aggrieved the 2nd to 6th respondents and triggered the judicialreview proceedings against the 1st respondent, 1st & 2nd appellants in which orders were sought to quash the decision of the 1st appellant cancelling the census results for the eight districts, and prohibit the 1st and 2nd appellants from publishing the projected results for the eight districts, or circulating any other figures apart from the published 2009 census results. In addition, an order was sought to prohibit the 1st respondent from relying on the decision of the 1st appellant cancelling the results or from relying on any projected results other than the published 2009 census results.

[6] The 1st and 2nd appellants responded to the suit through a replying affidavit sworn by Wycliffe Ambetsa Oparanya who was then holding the position of the Minister of State for Planning, National Development and Vision 2030, and Antony KM Kilele who was then holding the position of the Director-General of Kenya National Bureau for Statistics. In a nutshell, the deponents of the replying affidavits justified the data evaluation and assessment of the 2009 Kenya Population and Housing census results on the grounds that the exercise was undertaken in accordance with standard procedure and the results were part of the checks and balances required to understand the nature and magnitude of errors in the census exercise in order to facilitate the proper utilization of the data by stakeholders; that the assessment included the census for the entire country and not just the eight districts where the anomalies were noted; that the appellants are obligated by law to ensure the release of accurate data, and advise where data is not accurate; and that the advice for cancellation of the results for the eight district was given by the appellants to the Government in good faith and in accordance with tested census systems applied locally and internationally. In time, other parties applied and were joined in the suit as interested parties as already observed.

The Judgment of the High Court

[7] In his judgment, the learned judge noted that what was in question was the power of the 1st appellant to cancel the results of the eight districts after the census exercise was completed, vis-a-vis the rights of the inhabitants of the eight districts in so far as the cancellation was concerned. In his view, a reading of section 23of theStatistics Actdid not show or demonstrate that the 1st appellant could exercise any power in respect of the census results as the section only talked of the power of the Director-General to grant, permit, restrict and prohibit request or information under his custody. The learned judge stated that statutory powers could only be exercised reasonably, and for the intention for which the powers were donated; that if any power was exercised in a manner contrary to the drafters’ intention or against public interest, the power was exercised capriciously, irrationally or unreasonably.

[8] The learned judge held that the responsibility of carrying out a credible census that meets international standards was that of the 1st and 2nd appellants; that there was no blame attributed to the inhabitants of the eight districts to warrant the cancellation of the census results to their disadvantage; and that the negligence, whether deliberate or an inadvertent omission on the part of the 1st and 2nd appellants, could not be shifted to the inhabitants of the eight districts. The judge accepted the argument that cancellation of the results of the eight districts could only be done if there was a clear and deliberate error or omission, linked to the inhabitants in the area. He added that the inhabitants of the eight districts had a legitimate expectation that the 1st and 2nd appellants would act fairly with respect to the enumeration, analysis and dissemination of the 2009 population and housing census results.

[9] The learned judge further held that the 1st appellant had no powers under section 17 or 23 of the Statistics Act to cancel the results in the eight districts and that his action amounted to an illegality. As regards the interested parties, the learned judge found that they had not established any significant relationship between the decision of the 1st appellant and their grievance. The learned judge, therefore, concluded that the motion before him was merited and granted the orders sought. It is this decision that resulted in the present appeal.

The Memorandum of Appeal

[10] In their memorandum of appeal, the appellants have raised six grounds. In short, the appellants have faulted the learned judge for issuing the order of prohibition against the appellants, maintaining that the effect of the order was to fetter the discretion of the appellants from correcting errors and mistakes made in the census report, contrary to the principle in Administrative Law that courts cannot determine the manner in which an administrative body should exercise its discretion; that the order of prohibition resulted in the appellants being prohibited from performing their statutory duties as provided under section 23 of the Statistics Act which requires the Director-General, with the approval of the Board, to disseminate to the public statistical data collected by the Bureau upon ascertainment of the data’s accuracy; and that the order of prohibition resulted in the appellants circulating the censures figures without corrections or verification of their accuracy or correction of errors contrary to section 23 of the Statistics Act. In addition, that the learned judge misapprehended and misapplied the law by issuing an order of prohibition that contradicted his findings. On these grounds, the appellants seek to have the judgment and orders issued by the High Court on 7th February, 2012 set aside and substituted with an order dismissing the respondents’ application for judicial review.

Appellants’ Submissions

[11] In support of the appeal, the appellants filed written submissions that were highlighted during the hearing of the appeal by Mr. Mwangi Njoroge of the State Law Office. Briefly, it was the appellant’s position that the nature and magnitude of the census dictated that it be conducted through other people namely, local leaders and the citizenry; that the locals had as much responsibility as the Bureau to ensure the success of the exercise; that national mobilization and sensitization for the exercise was undertaken targeting the provincial administration, local authorities, local leaders and citizens; that owing to the nature and magnitude of such census exercise, errors were not uncommon; that this necessitated census data evaluation to be undertaken as a standard procedure to understand the nature and magnitude of errors in order to facilitate the proper use of the data by stakeholders; that the advice given by the 2nd appellant was based on its evaluation and assessment of the data done according to scientific and internationally acceptable standards; and that the data evaluation results of the eight districts revealed anomalies as the rate of growth was higher than what the documented population dynamics (birth and death rates, and migration tides) could support, and were also way above the projected figures based on previous census.

[12] It was argued further that the granting of the prohibition order was contrary to the principle of administrative law that courts cannot determine the specific manner in which an administrative body should exercise its discretion; that the order of prohibition offended the provision of section 23 of the Statistics Act that gave the Director-General of the Kenya National Bureau of Statistics, residual powers to correct any inaccurate information; that the correction of mistakes and errors in the census report could not therefore amount to abuse of powers; and that the import of Section 23 of the Statistics Act is that the Minister and the Director General have a residual power to correct errors made in the census results.

[13] The learned judge was also faulted for issuing the order of prohibition after finding that the concerns of the appellants, on the unusual demographic patterns, trends and dynamics, were grounds for carrying out further assessment to determine the source of variation or change in the trends and dynamics. Relying on Kenya National Examination Council (exparte) Geoffrey Gathinji Njoroge & 9 OthersNairobi Civil Appeal No. 266 of 1996(unreported), it was conceded that the 1st appellant did not have the powers to cancel the results of the census but reiterated that the 2nd respondent had powers and responsibility to correct the errors and anomalies. The appellants argued that the order of prohibition had the effect of tying their hands to the initial results making it impossible to correct the errors in the census results. That the spirit of section 23 of the Statistics Act requires that only accurate information is given to the public and therefore the decision of the appellants to carry out evaluation and re-assessment of the results was not unreasonable nor beyond the appellant’s power as per the standard set out in Associated Provincial Picture Houses Limited vs Wednesbury Corporation[1948] 1KB 223.

[14] Further it was submitted that the cancellation of the 2009 census results was not meant to frustrate the legislative purpose but to deal with the inaccurate information and ensure that the results were analyzed in accordance with the Statistics Act, section 4, which establishes the 2nd appellant as the principal agency of the Government for collection, analyzing and dissemination of statistical data.

[15] On legitimate expectation, the appellants relied on Schmidt vs Secretary of State for Home Affairs [1969]wherein it was held that a legitimate expectation arises where the citizen has been led to believe by a statement or other conduct of the Government that he is singled out for a benefit or advantage of which it would be unfair to deprive him. It was argued that the correction of the errors in the census results was not unfair as it was meant to prevent inaccurate information from being disseminated to the public or being erroneously used by public institutions to the detriment of the public. Further, that there was no promise or assurance given to the respondents that could generate a legitimate expectation. The Court was therefore urged to allow the appeal and set aside the judgment and orders of the High Court.

1stRespondent’s Submissions

[16] Although the 1st respondent did not file any written submissions, during the hearing of the appeal, Mr. Saluny, who was holding brief for Mr. Nyamodi, appeared for it and indicated that the 1st respondent was supporting the appeal. Learned counsel associated himself with the submissions made on behalf of the appellants, and faulted the learned judge for forcing the 1st respondent to rely on the initial results published as the 2009 census results. Counsel argued that the 1st respondent had the mandate under Article 41(c) of the Constitution, to make recommendations on the delimitation of boundaries and that in so doing the 1st respondent had an obligation to make sure that the number of inhabitants was accurate, and that this depended on the accuracy of the exercise undertaken by the 1st and 2nd appellants.

Submissions for the 7thto 10thRespondents

[17] The 7th to 10th respondents did not file any written submissions but relied on the submissions they had filed in the High Court. Ms Makobu, who appeared for these respondents, made oral submissions in which she supported the appeal and associated herself with the submissions made by the advocate for the appellants. Counsel pointed out that under Section 23 of the Statistics Act the appellants had a duty to publish accurate figures after correcting all the errors; that in this case the data was not verified but reliance was being placed on unverified qualified population data, which means the data was not complete. She gave example of the data of Hambalash sub location in regard to which there was zero entry showing no enumeration, whereas the results for the census for 1999 had shown a figure of 3,409. Ms Makobu reiterated that reliance on the unqualified data was prejudicial to the residents of those districts and a violation of Article 89 of the Constitution which requires the boundaries of constituencies to be pegged on the population quota. She urged the Court to take note of the fact that the population data is very crucial in allocation of national resources, social infrastructure and the delimitation of boundaries. Finally, counsel reiterated that the appellants had powers to verify and correct the census results.

Submissions in opposition to the appeal

[18] Mr. Ahmednasir, who appeared for the 4th and 5th respondents, presented the main oral arguments for the respondents who were opposing the appeal. In addition, the 3rd & 6th respondents, and the two interested parties, filed joint written submissions that were duly highlighted by Mr. Issa, Mr. Biriq for the interested parties. Mr. Keenhwe for the 3rd respondent, also made submissions reiterating the submissions made by Mr. Ahmednasir for the 4th and 5th respondents.

[19] In opposing the appeal, Mr. Ahmednasir noted that the appellants did not appeal against the substantive order of certiorari quashing the 1st respondent’s decision cancelling the 2009 census results for the eight districts. He maintained that the orders of prohibition upon which the appeal was anchored were mere consequential orders. Counsel argued that the appellants having conceded that the learned judge was right in quashing the decision of the 1st appellant, there was nothing to appeal against as the orders of prohibition flowed from the order quashing the 1st respondent’s decision.

[20] In addition, Mr. Ahmednasir asserted that there was no justification for the appellants to use the projected results; that in any case the 1st appellant had no powers to correct errors in the census results; that the 2nd appellant was not prevented from carrying out its duty as the order of prohibition issued against it was only limited to eight districts; and that the ascertainment of errors and corrections ought to have been done before the release of the census results.

[21] Further, counsel argued that neither the 1st appellant nor the 2nd appellant had any statutory power or discretion under the Statistics Act to use projected figures once a census had been undertaken, results analyzed and disseminated; that the appellants had not identified any omissions or duplications in the coverage or content of the census results to warrant cancellation of the census results. He urged the Court to dismiss the appeal as it was devoid of merit.

The 2nd, 6thto 10thRespondents

[22] In a nutshell, it was argued in the written and oral submissions made on behalf of the 2nd, 6th to 10th respondents that no evidence was laid before the Court to show that there were any irregularities affecting the eight districts or to justify any conclusion that the results posted for any of the administrative units in the eight districts, were either falsified or exaggerated; that the reasons given for the cancellation of the results by the 1st respondent were absurd; that the appellants not having challenged the order of the court quashing the decision of the 1st appellant to cancel the results for the eight districts, the results of the said districts remained as enumerated and released on 31st August, 2010; that in the circumstances the appellants have no justification for using projected results and therefore the order of prohibition was properly issued.

[23] It was also argued that the inhabitants of the eight districts would lose their rightful share of allocation of resources if projected figures are used, since this would have significant effect on the population, poverty index, land area and basic equal share that have to be taken into account in allocation of resources. It was maintained that the projected figures lowered the census-enumerated figures significantly, and that reliance and use of projected results would disadvantage the eight districts in the future review of electoral boundaries.

[24] With regard to the growth rate, it was submitted that the apparent cultural traditions and nomadic lifestyle contributed to the huge difference noted; that under the United Nations Principles and Recommendations for Population and Housing Census Revision 2, 2008, the demographic patterns, trends, dynamics and futures were grounds for carrying out further studies and assessment to determine the source of variation and changes in the trends and dynamics, but not a basis for cancellation of results or use of projected figures.

[25] Further, that although under section 33 of the Statistics Act, the 2nd respondent has a discretion in disseminating information, that discretion must be properly exercised and the court has jurisdiction to intervene if the discretion is abused. In this regard, the following cases were cited: Republic vs Non-Governmental Organization Coordination Board & Another[2014] eKLR; Republic vs Minister of Home Affairs exparte Sitamze [2008] 2 EA 59;andMunicipal Council of Mombasa vs Republic[2002] eKLR.

[26] It was pointed out that the 1st respondent did not oppose the application in the High Court and that in any case it had a statutory obligation to use enumerated census results as published. The support of the appeal by the 1st respondent was therefore shrugged off as based on a misapprehension. The Court was therefore urged to dismiss the appeal.

The Interested Parties

[27] The Mandera County Government and Garissa County Government, who were interested parties in this appeal, also opposed the appeal through written and oral submissions. In the submissions, the arguments made by the other respondents in opposition to the appeal were reiterated. In addition, it was noted that the 1st appellant made a statement in Parliament on 21st February, 2012 after the delivery of the judgment subject of this appeal in which it claimed that after a process of smoothening of figures, the national population figures had been significantly adjusted and that the figures for the counties of Garissa, Mandera and Wajir were reduced by 916,982. It was argued that the decision of the Minister was unreasonable, arbitrary and discriminatory, it being maintained that the 1st respondent did not have any discretionary powers under the Statistics Act and that the decision was illegal, arbitrary and an abuse of his office. In addition it was asserted that the order of prohibition issued by the High Court was necessary to forestall the establishment of a dangerous precedent whereby Government agencies conjure up population figures to fit into a political arrangement.

[28] The 3rd respondent also associated himself with the submissions made on behalf of the other respondents and urged the Court to dismiss the appeal as frivolous and lacking merit.

The order of Certiorari.

[29] In the motion before the High Court, the bone of contention was the exercise of the 1st appellant’s powers in cancelling the 2009 census results for the eight districts. It was maintained that the 1st appellant purported to cancel the results although he had neither statutory power nor jurisdiction to do so under the Statistics Act. A look at paragraphs 3,4,5,6,7 and 8 of the statutory statement that accompanied the motion, shows that the complaint was against the 1st appellant whose decision to cancel the 2009 census results was alleged to be illegal, unreasonable, an abuse of power and contrary to the legitimate expectation of the inhabitants of the eight districts.

[30] Following the enshrinement of the right to fair administrative action in the Constitution of Kenya, through Article 47, the exercise of administrative powers has now come under closer scrutiny. Anybody or person taking administrative action must in the first place have the authority or jurisdiction to take such action. If the body or person has the jurisdiction to take the action then other factors come into play. This includes a consideration of the decision or action taken as to whether the same is efficient and reasonable, and whether the process was expeditious and procedurally fair.

[31] Where an issue arises concerning the exercise of administrative powers, the first consideration would therefore be whether the person exercisingthe administrative authority had the jurisdiction or mandate to exercise such authority. Indeed in the application before the High Court, one of the main grounds upon which the administrative action taken by the 1st appellant was challenged was that the 1st appellant (who was 1st respondent in the High Court), did not have any statutory power to cancel the results after the publication. The learned judge captured this issue as follows:

“…It is the case of the applicants that the decision of the 1strespondent made on 31stAugust 2010 cancelling the results should be brought and quashed as the same was made ultra vires and an abuse of statutory power, the decision was unreasonable, it was meant to frustrate legislative purposes and lastly it frustrated legitimate expectation and consistency. It was contended that the Statistics Act and Statistics Order 2008 do not donate any power to the 1strespondent to cancel any part of census report. And that the 1strespondent in cancelling part of 2009 Population and Housing Census results was acting without jurisdiction and therefore his decision was ultra vires the Statistics Act...”

[32] The learned judge’s findings on that issue were as follows:

“…by cancelling the results of the 8 areas, without any justification and without any statutory underpinning (sic) its decision, then it acted in excess of its jurisdiction and without authority. In essence the Minister exceeded the limits and the confines of the law, thereby trespassing on the rights and the interests of the eight districts.

There is no power donated to the Minister under section 17 of and/or 23 of the Statistics Act to enable him to cancel the results of the eight districts. In my view therefore the Minister had no statutory power or authority to cancel the results of the eight districts. Consequently he crossed the boundary and he must be shown a red card for exceeding the limit of his powers. Secondly it is clear that the 1stand 2ndrespondents made a promise and/or a practice that induced a legitimate expectation that they would carry out a credible exercise to the satisfaction of all citizens of this country. It was mandatory for them to keep their promise or assurance. To do otherwise would amount to discrimination and violation of the fundamental principle of equality and fairness.

I have considered the adequacy of the reasons advanced for the cancellation of the results and it is my determination that the same amounts to an illegality in all aspects of ‘Wednesbury Standard’which applies to the generality of executive decisions. The respondent cannot be allowed to frustrate the expectations of the inhabitants of the eight districts. They can also not be allowed to take a new and different course for that would amount to an abuse of power...”

[33] It is worthy of note that in in this appeal the appellants have not challenged the decision of the learned judge in granting the order of certiorari quashing the decision of the 1st appellant to cancel the census results of the eight districts. In other words, the appellants have conceded that the 1st appellant did not have the powers to cancel the results. Notwithstanding this concession, the appellants are challenging the orders of prohibition that were issued by the learned judge.

The Prohibition Orders.

[34] The challenge by the appellants of the orders of prohibition begs the question: what is an order of prohibition? This question was answered in theKenya National Examinations Council (exparte) Geoffrey Gathinji Njorogecase (supra) wherein this Court addressed it as follows:

“What does an ORDER OF PROHIBITION do and when will it issue? It is an order from the High Court directed to an inferior tribunal or body, which forbids that tribunal, or body to continue proceedings therein in excess of its jurisdiction or in contravention of the laws of the land. It lies, not only for excess of jurisdiction or absence of it but also for a departure from the rules of natural justice. It does not, however, lie to correct the course, practice or procedure of an inferior tribunal, or a wrong decision on the merits of the proceedings – See

HALSBURY’S LAW OF ENGLAND, 4thEdition, Vol.1 at pg.37 paragraph 128. ….... The point we are making is that an order of prohibition is powerless against a decision, which has already been made before such an order is issued. Such an order can only prevent the making of a decision. That, in our understanding, is the efficacy and scope of an order of prohibition.”

[35] In the matter before us the impugned administrative decision was the decision to cancel the census results of the eight districts. This decision was made by the 1st appellant and communicated through a ministerial statement made in Parliament on 31st August, 2010. The learned judge having found that the 1st appellant had no powers to cancel the census results an order of certiorari quashing the decision was appropriate, and it is not surprising that the appellants have not challenged the same.

[36] However, in addition to the orders of certiorari, the 1st and 2nd respondents sought and obtained orders of prohibition in the following terms:

“2. That an order of prohibition be and is hereby granted prohibiting the 1stand 2ndrespondents from publishing, issuing, or gazetting projected results of Lagdera, Mandera East, Mandera Central, Mandera West, Wajir East, Turkana North, Turkana South and Turkana Central Constituencies, from circulating any other figures other than the published 2009 population and Housing Census Results to any other organ of the Government, Constitution Commission, offices or organizations;

3. That an order of prohibition be and is hereby granted prohibiting the 3rdrespondent, whether through its Commissioners, Secretariat, Officers, agents, employees or howsoever from acting on any other census data relating to Lagdera, Mandera East, Mandera Central, Mandera West, Wajir East, Turkana North, Turkana South and Turkana Central Constituencies other than the published 2009 Population and Housing Census Results in the determination of boundaries review and/or determination of new boundaries.” (emphasis added)

[37] In our view, several issues arise with regard to these orders. For instance what are “projected results” and what is their purpose and impact on census figures? How are the orders of prohibition related to the decision of the 1st appellant that was quashed? Was there any basis for issuing the orders of prohibition as sought? What is the effect of the prohibition orders and would it hinder the Government from carrying out future projections for purposes of government planning?

[38] In addressing these issues the statements made by the 1st and 2nd appellants at the time of releasing the 2009 Population and Housing Census Report are important. The first indication of the action that the 1st appellant was contemplating in regard to the census results for the eight districts is in the Foreword to the 2009 Kenya Population and Housing Census (KPHC) Report Volume 1A wherein he stated:

“The preliminary analysis of the data indicates that generally the KPHC data are complete and accurate. However data from eight districts namely Mandera Central, Mandera East, Mandera West, Lagdera, Wajir East, Turkana Central, Turkana South and Turkana North are anomalous, and exhibit population growth rates that deviate significantly from the patterns portrayed not only by the rest of the country but by their respective neighboring districts as well. The Government regardless has decided to release the results as enumerated. However, the data will be assessed further using internationally acceptable procedures and suitable remedial measures undertaken. Meanwhile the stakeholder and the public at large are advised to exercise caution while using the interim results for the affected districts.”

[39] At this stage the action envisaged was further analysis of the data even though the official results remained as announced. The 1st appellant was alive to the fact that there were stakeholders who needed to use the census results for purposes such as planning and policy formulation hence the caution. On 31st August, 2010 the same day that 1st appellant announced the 2009 KPHC results, the 1st appellant made a ministerial statement in Parliament in which he stated inter alia as follows:

“…The data for 2009 census is fairly accurate. However after performing routine data evaluation and assessment as recommended by the United Nations before releasing of any results, we have observed that the results of the following eight districts are irregular. …..The government wanted this exercise to be as transparent as possible and, therefore, has proceeded to release the results for all areas enumerated. However, the Government has decided to repeat the exercise in the affected districts with a view to ascertaining the results at a date that will be announced soon.”

[40] In the above quoted extract of the ministerial statement the 1st appellant, was clear on the further action to be taken. The action was not further evaluation and assessment of the data as intimated in the Foreword, but a repeat of the census exercise in the eight districts. One of the bodies that needed to use the census results was the 1st respondent that was required under Article 89 (5) and (6) of the Constitution to take into account the population when delimiting or reviewing constituency boundaries.

[41] These constitutional provisions provide as follows:

“(5) The boundaries of each constituency shall be such that the number of inhabitants in the constituency is, as nearly as possible, equal to the population quota, but the number of inhabitants of a constituency may be greater or lesser than the population quota in the manner mentioned in clause (6) to take account of –

(a) geographical features and urban centres;

(b) community of interest, historical, economic and cultural ties; and

(c) means of communication.

(6) The number of inhabitants of a constituency or ward may be greater or lesser than the population quota by a margin of not more than—

(a) forty percent for cities and sparsely populated areas; and

(b) thirty percent for the other areas.

(7) In reviewing constituency and ward boundaries the Commission shall --

(a) consult all interested parties; and

(b) progressively work towards ensuring that the number of inhabitants in each constituency and ward is, as nearly as possible, equal to the population quota.”

[42] These provisions placed a constitutional obligation on the 1st respondent to take into account population figures. These figures were crucial in establishing the number of inhabitants in each constituency and the population quota, factors that were imperative in reviewing or delimiting constituency boundaries, and this made the 2009 KPH census results crucial to the 1st respondent. This appears to be the context within which the Interim Independent Boundaries Review Commission (“IIBRC”), the predecessor to the 1st respondent, addressed a letter to the Minister of State for Planning, National Development & Vision 2030 following the announcement of the KPH census results by the 1st appellant. The letter written by IIBRC was not contained in the record of appeal, as it was apparently not availed to the High Court. The response to that letter, which was dated 17th September, 2010 and signed by the Permanent Secretary, was an annexure to the affidavit of Ahmed Mohammed Noor. It stated in part as follows:

RE: INTERIM POPULATION ESTIMATES FOR

DISTRICTS WITH NULLIFIED RESULTS

Reference to your letter Ref: IIBRC/1346/09/10 dated 2ndSeptember 2010 to the Hon Minister. The Kenya National Bureau of Statistics has now provided estimates for the eight (8) districts whose results were nullified by the Government.

The projections were done at constituency level, which are fairly congruent with the current district boundaries, especially for these districts, and then adjustments made as appropriate.

Please note that estimating the population of Northern Kenya is fraught with formidable challenges and, therefore, the verdict cannot be based on only one source. Specifically, please note the following:

Projecting highly transient (unstable) populations such as Northern Kenya is often very challenging and requires several parameters as well as other corroborating sources

The accuracy of the base population on which the projections are based (1999 Census) may not be established with high certainty given the enumeration patterns for this region in previous censuses or surveys.

We are still to examine other sources of information as well, namely mapping estimates, Post Enumeration Survey (PES) and even the population dynamics.

There is also need to examine the data for this region over time and smaller geographic areas (such as locations or even sub-locations) to isolate the problem areas.

Lastly the government intends to undertake a recount in the affected districts.

Therefore these estimates below are not the final figures for these districts. We will get the final, more authoritative figures once the foregoing suggestions are followed through.

Estimated Population for districts with nullified results

District Projected Based on

Difference 1999 census enumerated Difference

(Enumerated)

Lagdera 207,697 245,123 - 37,426

Wajir East 154,338 224,413 - 70,080

Mandera

Central 131,034 417,294 -286,260

Mandera

East 110,185 288,687 -178,502

Mandera

West 107,374 319,775 -212,401

Turkana

Central 237,198 254,606 - 17,408

Turkana

North 313,748 374,414 - 60,666

Turkana

South 101,510 226,379 -124,869

TOTAL 1,363,084 2,350,696 -987,612

I would like to caution that these estimates be treated as interim, and only used internally to facilitate government activities. They should not be published as the final adjustments to the enumerated population until we re-examine data for the region and other parts of the country critically- when we have more information in our hands.

[43] The above background puts the orders of prohibition that were sought against the 1st respondent, the 1st and 2nd appellant in context. It is evident that following its decision to nullify the results of the eight districts, the 1st and 2nd appellants purported to give estimated population figures of the eight districts based on projected results from the 1999 census. The order of prohibition as couched in prayer 2 had two aspects. First, to restrain the 1st and 2nd appellants from announcing or making available projected results for the eight districts, and secondly to restrain the two from disseminating any figures other than the published 2009 Population and Housing Census Results. Prayer (3) simply prohibited the 1st respondent from using any census data relating to the eight districts in the review and delimitation of new boundaries, other than the published 2009 Population and Housing Census Results. Although the learned judge granted these orders, he did not specifically address the orders in his judgment.

[44] In our view by providing estimated district population figures for the eight districts based on projected results, the response to the letter from IIBRC reveals an attempt to put into effect the decision by the 1st appellant to cancel the 2009 KPH census enumerated results of the eight districts. As already noted the Constitution obligated IIBRC to take into account population figures in delimiting or reviewing constituency boundaries. There is a clear difference between projected results and enumerated results. “The United Nations Principles and Recommendations for Population and Housing Censures, Revision 2”a document that was relied upon by the appellant, defines a “Population Census” at page 7 as:

“The total process of collecting, compiling evaluating, analyzing and publishing or otherwise disseminating demographic economic and social data pertaining, at a specified time to all persons in a country or in a well delimited part of a country”

[45] The same document identifies the essential features of a population census as individual enumeration, universality within a defined territory, simultaneity, and defined periodicity. What this means in simple terms is that a Population and Housing Census must of necessity involve enumeration of each person, within a defined area, at the same well defined point in time, and repeated at regular intervals so that comparable information is made available in a fixed sequence. A population census exercise is focused around enumeration, which is divided into 3 main phases. These are the pre-enumeration stage, the enumeration stage and the post-enumeration stage. Black’s Law Dictionary Ninth Edition defines “enumerate” as a verb meaning “to count off or designate one by one; to list”and identifies enumeration as the noun derivative of enumerate. This means that population enumerated results are results based on the actual counting of individuals under circumstances that meet the essential features of a population census as above stated.

[46] On the other hand, the United States Census Bureau, an agency of the US Federal Statistical System, in its publication under the title “Population Estimates Terms and Definitions”(accessed at http://www.census.gov/popest/about/trems.html) defines “Population projection” as:

“Estimates of the population for future dates. They illustrate plausible courses of future population change based on assumptions about future births, deaths, net international migration, and net domestic migration. Projected numbers are typically based on an estimated population consistent with the most recent decennial census as enumerated, projected forward using a variant of the cohort-component method.’’

[47] Population projection is thus an estimate of future population based on assumption on certain scientific factors. The same source also defines “Population Estimates” as – “The calculated number of people living in an area as of a specified point in time. Usually 1stJuly. The estimated population is calculated using a component of change of model that incorporates information on natural increases (births, deaths) and net migration (net domestic migration, net international migration) that has occurred in an area since the latest decennial census”

[48] Thus enumerated results are not the same as projected results, the latter are a projection of the expected future population. The 1st respondent could not use estimated or projected population results in the review or delimitation of the constituency boundaries as this was contrary to the express provisions of the Constitution that required use of the number of inhabitants in a constituency, and use of the population quota based on that number, and this called for population census enumerated results and not estimated or projected figures. The contention that the data that was announced by the 1st appellant as the enumerated results of the 2009 census, was inaccurate and required verification, cannot hold because the 2nd appellant announced the results exactly one year after the date of the actual enumeration.

[49] This means that the 2nd appellant had one year to do the evaluation, analysis and verification of the enumerated census results. This period was enough for the 2nd appellant to subject the enumerated results for the eight districts to scrutiny, verification, further study and assessment, so that the enumerated results ought not to have been announced if there were such serious anomalies as to negate the enumerated results. The order of prohibition barring the 3rd respondent from using data other than the published 2009 Population and Housing Census Results in the determination of boundaries review and/or determination of new boundaries was thus proper within that context, and the learned judge cannot be faulted.

[50] But the matter does not rest there. The learned Judge did not restrict the prohibition to the 1st respondent only, but gave it a wider application by prohibiting the appellants from publishing or gazetting, or circulating to any otherorgan of the Government, Constitutional Commission, offices or organizations the projected results for the eight districts or circulating any other figures other than the census population results for 2009. The learned judge failed to take into account that population census results are required by different bodies for different purposes.

[51] In the United Nations Principles and Recommendations for Population and Housing Censures, Revision 2,at Chapter II paragraphs 1. 21 -various uses of population census results are given. These include providing the facts that are essential to the government for policymaking, planning and administrative purposes; providing data for scientific analysis and appraisal of the composition, distribution and past prospective growth of the population for research purposes, providing data for use in the business, industry and labour market in determining consumer demand for goods and services anchored on the population size; and providing data that is necessary for the electoral boundary delimitation.

[52] The importance of population projections are underscored by the Population Reference Bureau, a private US non-profit organization that informs people around the world about population health and the environment. In its policy brief entitled “Understanding and Using Population Projections” (accessed on 7th April 2016 at http://www.prb.org/pdf/UnderStndPopProj_Eng.pdf), the Population Reference Bureau States:

“Government policymakers and planners around the World use population projections to gauge future demand for food, water, energy and services and to forecast future demographic characteristics. Population projections can alert policy makers of major trends that may affect economic development and help policymakers’ craft policies that can be adapted for various projection scenarios”

[53] The brief further indicates that most national governments make population projections for their own countries, and that the United Nations issue revised global population projections regularly, that are relied on worldwide by many national governments, international agencies, the media, researchers and academic institutions. In addition, the World Bank also issues population projections for the world major regions and individual countries that are relied upon for planning and managing projects.

[54] Thus the population projection exercise undertaken by the 2nd appellant cannot be said to be an abuse of discretion nor an exercise that was beyond the powers of 2nd appellant, nor was it an exercise that contravened the law as population projection is an exercise that is accepted world-wide. Indeed the population projection of all districts and not just the eight districts is an exercise that is crucial to the national government for policy decision-making, planning and resource allocation. The prohibition order that was issued by the learned Judge restraining the 2nd appellant from publishing or gazetting, or circulating the projected results for the eight districts or circulating any other figures other thanthe census population results for 2009 to any other organ of the Government, Constitutional Commission, offices or organizations, in effect made it impossible for the 2nd appellant to provide population projections for the eight districts, thereby denying government policymakers, and other research institutions information that is crucial for their operations. The requirements of Article 89(5),(6) & (12)with regard to the use of enumerated results in establishing the population and the population quarter, was specific to the 1st respondent in so far as the delimitation of electoral boundaries was concerned. The requirements could not be extended to other organs of Government or Constitutional Commissions or other organizations, nor justify preventing the Government and other bodies from using projected figures in appropriate circumstances.

[55] We come to the conclusion that the learned judge was wrong in issuing a blanket order of prohibition restraining the publication and circulation of the projected results to bodies other than the1st respondent. Accordingly we allow the appeal to the extent of setting aside the order prohibiting the 1st and 2nd respondents from publishing, issuing, or gazetting projected results of Lagdera, Mandera East, Mandera Central, Mandera West, Wajir East, Turkana North, Turkana South and Turkana Central districts, or from circulating to any other organ of the Government, Constitutional Commission, offices or non-governmental organizations, any other figures other than the published 2009 population and Housing Census Results. We order that each party shall bear its own costs of the appeal.

Those shall be the orders of this Court.

Dated at Nairobi this 13thday of May, 2016.

E. M. GITHINJI

…………………..

JUDGE OF APPEAL

H. M. OKWENGU

……………………

JUDGE OF APPEAL

J. W. MWERA

……………….….

JUDGE OF APPEAL

G.B.M. KARIUKI

………………….

JUDGE OF APPEAL

F. AZANGALALA

…………………...

JUDGE OF APPEAL

I certify that this is a true copy of the original.

DEPUTY REGISTRAR