STATE

CAPTURE
INSIDE KENYA’S
INABILITY TO FIGHT
CORRUPTION

a
STATE
CAPTURE
INSIDE KENYA’S
INABILITY TO FIGHT
CORRUPTION
Wachira Maina

This is a ‘State of the Nation’

report by the Africa Centre for
Open Governance (AfriCOG),
explaining why President Uhuru
Kenyatta’s current anti-corruption
efforts are unlikely to work. The
problem, the study shows, is state
capture; that is, the repurposing
of state institutions for private
profiteering. The study concludes
that publicity-driven prosecutions
are likely to deepen rather
than undermine corruption, as
indictments and prosecutions get
weaponised to partisan ends for
the 2022 elections.
Design & Layout | STUDIO INFERNO CO LTD | [email protected]
CONTENTS
Abbreviations and Acronyms......................................................................................................................iv

Executive Summary........................................................................................................................................ v

Part 1: Corruption, Governance and the State........................................................................................1

The global corruption problem.......................................................................................................................1
Kenya’s corruption problem.............................................................................................................................2
The problem with the conventional approach to corruption....................................................................5
State capture......................................................................................................................................................6
Why the state capture perspective makes sense.........................................................................................9

Part 2: Tales of State Capture: Goldenberg, Anglo Leasing, Eurobond.........................................12

Corruption and politics, never the twain shall part...................................................................................12
Goldenberg – designing the methods of state capture............................................................................15
The Anglo Leasing SCANDAL.........................................................................................................................18
The Eurobond SCANDAL................................................................................................................................24

Part 3: The Mechanics of Capture.............................................................................................................27

Capture technique 1: ensure that the electoral management body is compromised.........................27
Capture technique 2: undermine law enforcement..................................................................................32

Part 4: Undoing State Capture: In Search of Opportunities and Reform......................................42

Democracy retains potential to undo state capture.................................................................................42
How might the democratic process undo state capture? . ......................................................................43
What civil society organisations can do.......................................................................................................46
Conclusion.......................................................................................................................................................52

iii
ABBREVIATIONS AND ACRONYMS
AG Attorney General
AfriCOG Africa Centre for Open Governance
CKRC Constitution of Kenya Review Commission
CPI Corruption Perceptions Index
CPIB Corrupt Practices Investigation Bureau
CS Cabinet Secretary
CSO Civil Society Organisation
DCI Directorate of Criminal Investigations
DPP Director of Public Prosecutions
DRC Democratic Republic of the Congo
EACC Ethics and Anti-Corruption Commission
ECK Electoral Commission of Kenya
EMB Electoral Management Board
EVID electronic voter identification devices
IEBC Independent Electoral and Boundaries Commission
IIEC Independent Electoral Commission
IMF International Monetary Fund
IREC Independent Review Commission
ISDCS Immigration Security and Document Control System
ICTJ International Centre for Transitional Justice
JSC Judicial Services Commission
Kshs Kenya Shillings
KPLC Kenya Power and Lighting Company
LSK Law Society of Kenya
NIS National Intelligence Service
NPSC National Police Service Commission
NSIS National Security Intelligence Service
OCPD Officer in Charge Police Division
PCGG Presidential Commission on Good Government
PPARB Public Procurement Administrative Review Board
PS Principal Secretary
TI Transparency International
US$ US Dollars
WDF World Duty Free

iv
EXECUTIVE SUMMARY
President Kenyatta has launched
a highly visible anti-graft war ...

Two new sheriffs in town

It has been a frantic period since 13 March 2018, when President Kenyatta appointed Noordin Haji
of the National Intelligence Service (NIS), as the Director of Public Prosecutions (DPP), to replace
the lacklustre Keriako Tobiko, who is now Cabinet Secretary for Environment and Forestry. Almost
simultaneously, Mr Kenyatta also appointed Paul Kihara, former President of the Court of Appeal as
the Attorney General (AG) to replace Githu Muigai, completing what was seen as both a makeover
and a clean-up of the ineffectual State Law Office1. On the political front, Uhuru Kenyatta’s
reconciliation with his main rival, Raila Odinga of the National Super Alliance (NASA), restored some
measure of the political legitimacy that he lost in the deeply flawed presidential elections of 2017.
Because the appointments and the making-up were so unexpected, so seemingly out of character,
and so removed from the usual hands-off, see-no-evil, hear-no-evil modus operandi of Mr Kenyatta’s
first term, many observers were convinced that the President had made a break with the past; that
he was primed to battle the twin bugbears of his administration: the toxic post-election politics –
which were both polarising and debilitating – and institutionalised corruption.

Kihara & Haji sworn in

Paul Kihara, Attorney General Noordin Haji, Director of Public Prosecutions

1 See The Star, “Uhuru nominates Noordin Haji to succeed former DPP Tobiko”, 13 March 2018 at https://www.the-star.co.ke/news/2018-03-13-uhuru-nominates-
noordin-haji-to-succeed-former-dpp-tobiko/

v
A handshake with his opponent
On this reading, finding common ground with Mr Odinga normalises politics and gives Mr Kenyatta
the legitimacy he needs to make difficult political choices associated with fighting corruption.
The new appointments, the story goes, should revitalise the State Law Office and give the war on
corruption the impetus it needs to succeed. The narrative is that Mr Kenyatta is looking beyond
2022, to his legacy, which he now sees as imperilled by corruption. In short, the President’s change
of heart is real.

To be sure, something has changed: for the first time, the government has launched, and more
important, sustained (so far) an aggressive campaign of high-profile arrests and indictments. The
DPP may sometimes sound sanctimonious but he has been consistent, even with the inevitable
energy slump that comes with time.

An invigorated State Law Office

For his part, Mr Kihara has brought more energy and less showmanship to the AG’s office: he has
invigorated the 2015 Multi-Agency Task Force on Corruption – comprising the National Intelligence
Service, the Directorate of Criminal Investigations (DCI), the Anti-Money Laundering Unit, the Asset
Recovery Agency, the Anti Banking Fraud Unit, the Financial Reporting Centre, the Kenya Revenue
Authority, the Cybercrime Unit, the Anti-Counterfeit Body and the Ethics and Anti-corruption
Commission (EACC). Though it has not made public its operations since March last year, the AG’s
office has reportedly given the President a confidential list of high-level government officials
involved in graft. This list is ‘rumoured’ to be the basis of the latest spate of arrests2.

In June 2018, President Kenyatta sent more than 1,000 procurement officers and accountants
home, fulfilling his Madaraka Day promise that they needed to be vetted afresh. His promise was
that this revetting would include lie detector tests.

Initiated a money repatriation effort

Simultaneous with the arrests and indictments, the government says that it has launched a ‘bring-
back-our-cash’ initiative by writing to seven countries, including Britain, Mauritius, Dubai and
Switzerland, for details of money – said to be in trillions of shillings – hidden there by politicians
and connected businessmen. Australia may also be on that list, given the claim by the EACC that it is
following Kshs 4.5 million of corruption money that Migori governor, Okoth Obado and his children
have squirrelled away and laundered in a local casino in the country. Whether this initiative will yield
a money return remains to be seen (see below).

Impressed the diplomatic corps

President Kenyatta’s moves have impressed the diplomatic corps and created a buzz among some
Kenyans. In August 2018, the United Nations Economic Commission for Africa’s (UNECA) Executive
Secretary, Vera Songwe, was particularly fulsome, christening Mr Kenyatta ‘an African champion’
for his war on corruption.

2 “Special anti-graft unit sends shivers as more governors put on arrest list”, 7 July 2018 https://www.standardmedia.co.ke/article/2001287032/special-anti-graft-unit-
sends-shivers-as-more-governors-put-on-arrest-list

vi
This accolade may have been encouraged by the fact that, as part of the anti-corruption drive,
Kenya has appointed a National Task Force on Anti-corruption to support the UN Collaborative
Programme on Reducing Emissions from Deforestation (UN-REDD). Nonetheless, Mr Kenyatta’s
measures are being watched with interest, even by the sceptical.

Yet the President has left intact

the roots of corruption in Kenya …
What are we to make of this latest ‘war on corruption’? The conventional account may be summarised
as follows. The government’s efforts – though rather modest when set against the scale of the
problem – deserve support. Yet such support should be based on a critical understanding of the
nature of Kenya’s corruption problem and strategic advice on tackling it. The current approach
is largely tactical, too legalistic and prosecution-driven and unfortunately, counter-corruption
has been ‘weaponised’ to resolve the Jubilee party’s (the ruling party) factional infighting. In the
main, the government has up to now pursued a series of unconnected anti-corruption activities,
mainly prosecutorial, when it really needs an anti-corruption strategy based on a proper diagnosis
of the problem and the drivers of that problem. High profile arrests and indictments; ‘aggressive’
repatriation attempts; revetting and suspending of procurement officers; and revitalising and
funding blue-ribbon multi-sectoral agencies have all been tried before and proven to be ineffectual.
There is something to be said for this conventional summary but even if the government adopted
the advice it gives, it is unlikely to dig out corruption’s deep roots.

This is because the difficulties of fighting corruption in Kenya lie in the union of corruption and
politics; a union in which, at least since the Goldenberg scandal, a power elite has captured the
state, especially the Presidency and the Treasury and repurposed the machinery of government
into a “temporary zone for personalised appropriation”.3 The object of this ‘repurposing’ is to gut
state resources for electioneering and thus maintain power. In this dispensation, politics is a zero-
sum game of “competitive aggression”4 in which “the principal victim” is “the state itself”5 and
politics is “a pursuit requiring ever faster forms of enrichment”. For state capture to succeed, four
things must happen.6

One, oversight institutions must be eviscerated and hollowed out. This means that the offices of
Controller of Budget, the Auditor General, the Kenya National Commission on Human Rights and
parliamentary committees must be totally compromised and wholly ineffectual in their oversight.

Two, law enforcement and rule of law institutions – the police, the judiciary, the EACC and the
prosecution agencies – must be weakened or captured and redirected as ‘weapons’ with which to
fight political opponents.

Three, the ordinary channels of political change and accountability through periodic elections must
be blocked, either by compromising the electoral management board (EMB) or through violent
intimidation of political opponents.

3 Burbidge, D (2015), The Shadow of Kenyan Democracy: Widespread Expectations of Widespread Corruption at p. 35 Ashgate Publishing
4 ibid p.33
5 ibid p.33
6 ibid p.33

vii
Finally, the space for countervailing institutions to function – especially civil society and media
– must be shrunk until these pose no threat to capture. Alternatively, these institutions are also
compromised and redirected to the state capture agenda.

This study has four parts. Part 1 has four segments: the first summarises the nature of the corruption
problem, the second outlines its Kenyan dimensions, the third discusses the conventional approach
to dealing with it, and the final segment outlines the state-capture theory that this study adopts. Part
2 has detailed descriptions of the three definitive cases of state capture in Kenya: the Goldenberg
scandal; the Anglo Leasing scandal and the Eurobond scandal. It draws lessons from these case
studies to set the stage for Part 3, which shows why the reforms so far adopted won’t work. Part 4
lays out realistic options for reform.

viii
Part 1

CORRUPTION, GOVERNANCE
AND THE STATE
The global corruption problem
Addressing the UN Security Council in September last year, Secretary General António Guterres
decried the huge and growing cost of corruption to the global economy. Estimates from the World
Economic Forum, he told the Council, put the annual global cost of corruption at “US$2.6 trillion,
or 5 per cent of the global Gross Domestic Product.”7 Out of this, the World Bank estimates that
more than one third (US$1 trillion) is composed of the bribes that businesses and individuals pay to
access services.

The world can ill afford such leakages and yet these estimates may still be gross underestimates
of the actual cost that corruption imposes on global governance. Some of the more significant
opportunity costs are often incommensurable. One can measure how much of the money meant for
schools, drugs, hospitals and roads may have been diverted through bid rigging and kickbacks, but
what is the cost to society when the police look the other way as criminals cannibalise institutions?

Such costs have become evident. Indeed, the Secretary General’s address arose from the fact that
the Security Council had finally cottoned onto the idea that there are links between corruption
and global conflicts, the Council’s core mandate. The pathways are not always explicit, but
corruption undermines states by weakening institutions in ways that erode both their resilience and
effectiveness, thereby making a country susceptible to conflict. On the demand side, corruption
corrodes social norms, such as trust, mutual faith and forbearance; in short, it destroys values that
foster social capital, the glue that holds communities together. Externally, many of the mechanisms
that finance conflict are spawned by corruption. The global networks of arms dealers, illicit drugs,
human traffickers, terrorists and money launderers are increasingly connected to international
conflict entrepreneurs and all thrive in the illicit spaces that corruption opens up.

More prosaically, corruption aggravates inequality and is a massive tax on the private sector.
Hobbling the private sector undermines growth and poverty reduction, the two main interventions
for addressing the drivers of conflict. Corruption deepens inequality in two ways, one of which,
according to the World Bank, is direct. The Bank cites studies that show that the poor pay a higher
percentage of their income in bribes8 than the rich. It notes, for example, that in Paraguay, the
poor pay 12.6 percent of their income in bribes while high-income households pay only 6.4 percent.

7 See António Guterres, Remarks to the Security Council on Corruption in Conflict to the UN Security Council on 10 September 2018 https://www.un.org/sg/en/content/
sg/speeches/2018-09-10/corruption-conflict-remarks-security-council
8 See World Bank Brief, Combating Corruption at http://www.worldbank.org/en/topic/governance/brief/anti-corruption

1
In Sierra Leone, the numbers are 13 percent and 3.8 percent, respectively. In addition, the poor rely
disproportionately on public services. When these are subverted by corruption, as they often are,
the poor pay a disproportionate price in foregone services.

The old canard that corruption is not universally bad – that it can be beneficial when it ‘oils the
wheels of commerce’ – must be rejected, even though there are many people who genuinely
believe that paying bribes or peddling influence speeds things up. There is a form of the prisoner’s
dilemma involved in this thinking. As Rotberg points out, an “occasional corrupt act, in isolation, may
be efficient” but “routinized corruption never is.” Pervasive corruption, such as is entailed in state
capture, destroys the state. Even payments that seem petty and benign can create a moral hazard:
paying for services that should be available for free creates the risk that what was initially a cost for
speeding up a service becomes a mandatory private fee, generating as in Bangladesh, an official
stance of, ‘no money no service’.9 Moreover, once a climate of impunity is created, services can be
sped up for the ‘right ethnic group’ and slowed down for the wrong one, generating or sharpening
the kind of inequalities that poison communal relations and stoke deadly conflicts. More poignantly,
even when corruption has oiled the machinery, corrupt officials – already morally compromised –
will rarely agonise over what they are giving a licence for: it may be relatively harmless to speed up
a licence to open a bar, but a corrupt licence to practice medicine, open a clinic or erect a building
can have deadly consequences.

SUB-SAHARAN AFRICA in the CPI

32/100
Sub-Saharan Africa is the lowest scoring region on the index, and has failed to translate its anti-
corruption commitments into any real progress. A region with stark political and socio-economic
contrasts and longstanding challenges, many of its countries struggle with ineffective institutions
and weak democratic values, which threaten anti-corruption efforts.
AVERAGE SCORE
SCORE COUNTRY/TERRITORY RANK 40 Benin 85 30 Sierra Leone 129 20 Democratic Republic of 161
66 Seychelles 28 38 Swaziland 89 30 Togo 129 the Congo
61 Botswana 34 37 Gambia 93 28 Guinea 138 19 Angola 165
57 Cabo Verde 45 36 Tanzania 99 27 Comoros 144 19 Chad 165
56 Rwanda 48 35 Côte d’Ivoire 105 27 Kenya 144 19 Congo 165
53 Namibia 52 35 Zambia 105 27 Mauritania 144 17 Burundi 170
51 Mauritius 58 34 Ethiopia 114 27 Nigeria 144 16 Equatorial Guinea 172
46 Sao Tome and Principe 64 34 Niger 114 26 Central African Republic 149 16 Guinea Bissau 172
45 Senegal 67 32 Liberia 120 26 Uganda 149 16 Sudan 172
43 South Africa 73 32 Malawi 120 25 Cameroon 152 13 South Sudan 178
41 Burkina Faso 78 32 Mali 120 24 Eritrea 157 10 Somalia 180
41 Ghana 78 31 Djibouti 124 23 Mozambique 158
41 Lesotho 78 31 Gabon 124 22 Zimbabwe 160
Credit – Transparency International
http://venturesafrica.com/corruption-perception-index-2019-how-sub-saharan-africa-fared/

Kenya’s corruption problem

Corruption has been a persistent problem in Kenya since before independence but it has flourished
and put down robust roots since the country’s return to multiparty politics in 1992. With more
democratisation, the government’s infirmity in fighting corruption has also grown proportionately.
The cost of corruption to Kenya is a much-debated figure, but some experts say it is up to 1 percent of
GDP per year.10 Corruption cases are routinely reported in the press, in the Auditor General’s reports

9 Rotberg, R. I., (2017) The Corruption Cure: How Citizens and Leaders Can Combat Graft p. 33 Princeton, NJ: Princeton University Press, in which he reports survey
results showing that 71 percent of households paid bribes in order to obtain official services.
10 In 2016, the newly appointed chair of the EACC told Reuters that Kenya is losing a third of its state budget - the equivalent of about $6 billion - to corruption every
year. https://www.reuters.com/article/us-kenya-corruption/third-of-kenyan-budget-lost-to-corruption-anti-graft-chief-idUSKCN0WC1H8

2
and to the Ethics and Anti-Corruption Commission (EACC) but these are rarely fully investigated let
alone resolved satisfactorily. Commissions of inquiry11, such as the Inquiry into the Goldenberg Affair,
the Inquiry into Illegal and Irregular Allocation of Public Land, and the Gicheru Commission of Inquiry
into Robert Ouko’s death, have achieved little. Even where a commission proposes extensive reforms
– as the Inquiry into Illegal and Irregular Allocation of Public Land did12– these are implemented
in a patchy manner, seemingly ‘triaged’ to exonerate the powerful or to punish their enemies.
Key sectors of the economy – food, land and oil for example – are vulnerable to periodic heists and
systematic mendacity and cover-ups. Reports of investigations done by Committees of Parliament –
such as the Musikari Kombo list of shame13– are either totally rejected or doctored before they are
laid before the House, ensuring that no action is ever taken. Prosecutions fare no better: small fish
are nabbed, big fish never. Even where, uncommonly, an indictment followed by a conviction, has
happened, the stolen money is rarely recovered. In the handful of cases where monies have been
returned, it is usually from lowly officials. In some exceptional cases, the bigwigs will make secret
‘sweetheart deals’ in which they return some token assets but get to keep the greater loot.14

All this motion without movement is possible because corruption is deeply

embedded in politics, which it both funds and subverts.

As the report of the Task Force on Public Collections or Harambees, 2003, showed clearly, politicians
are the largest donors to ‘charitable’ causes – churches, schools, higher education and funerals are
firm favourites – to which they give fortunes that are many times more than their known legitimate
incomes. Such charity is, in truth, a bait and switch ploy: once moral institutions buckle to the lure of
corruption money, the corrupt buy absolution and are free to dip deeper into public coffers.

Reforms come in unsustained and unsustainable spurts, usually after an election, as in 2003, or in a
moment of fiscal crisis, as in 2013 just before Kenya went to the market to issue its first Eurobond.
In the early phase of the Kibaki administration in 2003, there was frantic action that, for a moment
seemed to herald a fresh beginning in the fight against graft. The national mood was optimistic,
exultant and supportive. In short order, 1,000 procurement officers were suspended after an audit
showed widespread irregularities.15 Between 2003 and 2007 a harvest of new laws were made,
buttressing a new set of anti-corruption institutions.16 Commissions of inquiry and task forces17
were set up to investigate past scandals and propose additional reforms. They were expectant
times: in a 2003 Gallup International survey of more than 67,500 people in 65 countries, Kenyans
were ranked the most optimistic people in the world. This result was not a response to any of these
reforms, merely the after-glow of the fall of the Moi autocracy. What it says, though, is that the
government was on a huge wave of popular ‘feel-good’ and legitimacy.

11 On the general weaknesses of commissions of inquiry as a public policy tool see AfriCOG, (2010) Postponing the Truth: How Commissions of Inquiry are Used to
Circumvent Justice in Kenya.
12 AfriCOG, (2009) Mission Impossible? Implementing the Ndung’u Report
13 See http://news.bbc.co.uk/2/hi/africa/841579.stm
14 Cf. the recent decision of the National Land Commission allowing Deputy President Ruto to keep a hotel he had built on illegally acquired land, The Star, “How NLC
arrived at Weston ruling”, 29 January 2019.
15 AfriCOG, (2008) Shattered Dreams: An Audit of the Kibaki Government’s Anti-Corruption Drive 2003-2007 at p.4
16 By May 2003, the government had enacted the Anti-Corruption and Economic Crimes Act and the Public Officer Ethics Act. Three more laws - the Public Procurement
and Disposal of Assets Act, the Public Finance Act and the Government Financial Management act – were passed by 2005. The Witness Protection Act and Political
Parties Act were enacted by end 2007.
17 The Bosire Commission of Inquiry Into the Goldenberg Scandal, 2003; the Kiruki Commission of Inquiry into the Affairs of the Artur Brothers 2006 and the Ndungu
Commission of Inquiry into Irregular and Illegal Allocation of Public Lands, 2004.

3
It didn’t last. The laws were no sooner enacted than ignored. Task forces and commissions
completed their tasks but nothing happened. Notwithstanding the baleful effect on the national
economy of the Goldenberg scandal, the commission of inquiry into the scandal, led by Justice
Bosire was irresolute in its recommendations, most recommending further investigations.18
The commission had been tasked with investigating major financial fraud in the early 1990s, involving
a company called Goldenberg Ltd. and the Central Bank of Kenya (CBK), in which the CBK and the
local banking sector made losses of over US$1 billion. The report of the Kiruki Commission (2006)
on the activities of the notorious Artur Brothers19 never even saw the light of day. The two brothers
had been involved in a series of high-profile criminal activities and security breaches in collusion
with highly placed Kenyan officials, including a raid on the premises of the Standard Media Group.
The Task Force on Public Collections or Harambees 2003, revealed how intimately corruption was
coupled to politics and laws were amended to stop civil servants from getting involved in fund-
raising activities. By 2010, the law was being honoured more in the breach than observance.

As under the Moi regime, the much beloved commission of inquiry proved to be an exceptionally
weak instrument against graft. Few real changes20 have come out of the more than 30 commissions
of inquiry Kenya has had in a century. And yet tax-payers invest millions in these commissions
expecting a real return by way of official integrity. As one AfriCOG report21 notes, reckoning the
costs of these anti-corruption commissions shows:

“The Goldenberg Commission cost Kshs 503 million of which Kshs 200 million was
spent on operations and Kshs 303 million on allowances. The Ndung’u Commission
cost Kshs 78.1 million of which Kshs 7.4 million went to operations and Kshs 70.7
million was paid out as allowances. The Kiruki Commission cost Kshs 19.97 million
of which Kshs 13.37 million was paid out in allowances and Kshs 6.6 million went
to operations.” 22
Yet despite all the efforts, President Kibaki failed to actually reduce corruption. Instead, barely a
year into the Kibaki presidency, in 2004, the still unresolved Anglo Leasing scandal broke out.23
By the mid-point of President Kibaki’s first term, most Kenyans had given up hope that corruption
could be meaningfully tackled. To date, that has not changed. The widespread nature of corruption
means that Kenya routinely performs very poorly in international and local measures of corruption,
integrity and good governance. Since the Transparency International (TI) Corruption Perceptions
Index was launched in 1995, Kenya has invariably been in the bottom half of the countries surveyed.
Sometimes, the country records some improvement on particular indicators, like the World Bank’s
Government Effectiveness and Doing Business surveys, but generally it soon slips back to its bad
ways, or just stagnates. Locally, corruption surveys by TI’s national chapter report widespread
bribery. Some institutions – the police, land registries and county licensing services – are notoriously
predatory.

18 See AfriCOG, (2011) All that Glitters? An Appraisal of the Goldenberg Report, p 18ff at https://africog.org/wp-content/uploads/2011/03/Goldenberg_
Report2029_03_20111.pdf
19 Report on the Investigation into the Conduct of the Artur Brothers and Their Associates . https://www.scribd.com/document/238299479/Report-on-the-Investigation-
Into-the-Conduct-of-the-Artur-Brothers-Their-Associates-July-2007
20 Indeed it is difficult to justify spending public money on a Commission of Inquiry into Devil Worship as happened in 1994 appointed by President Moi under pressure
from churches and media alleging that devil worship was rife in schools.
21 See AfriCOG, (2008) Postponing the Truth: How Commissions of Inquiry are used to Circumvent Justice in Kenya, at https://africog.org/wpcontent/uploads/2008/09/
Commissions20Of20Inquiry20Full20Report1.pdf.
22 ibid p 7.
23 Anglo Leasing was a series of procurement contracts in which the government ‘contracted’ fictitious or dummy companies to provide security and passport related
services worth over US$700 million. The scandal broke after an MP queried a Kshs 2.7 billion contract for a new passport and immigration system in April 2004.

4
As this report argues, the problem is two-fold: the nature of the corruption problem that Kenya
faces; and the reform theory behind the anti-corruption measures that donors push a reluctant
Kenyan government to implement. Mostly, the government embraces ‘appearances’ of reform,
rather than the fact of reform. The report argues that this is so because deep reforms would loosen
the ruling elite’s grip on power and severely subvert politics as played in Kenya.

The problem with the conventional approach to corruption

What, then, is to be done? A generation of reforms has not dented the corruption edifice or undone
its rhizome-like penetration into the body politic in Kenya. Part of the problem is conceptual: how
we name corruption and how we understand its character.

First, the naming. Corruption studies – and therefore corruption reforms – typically analyse
corruption in terms of ‘petty’ or ‘grand’ corruption: the one pervasive, the other destructive.24
However, and here we come to the impact on our understanding of the character of corruption,
these simple but loaded terms present a false dichotomy. ‘Petty’ suggests that the corruption is
merely an irritant: something, as Rotberg points out, that people who want to speed things up
or evade a long queue do to “lubricate the system”.25 The term suggests an expedient with trivial
effects, considered case by case. In fact, that characterisation is deeply mistaken:

“… if everyone, public official and consumer alike, expect transactions to be

lubricated with an inducement or consideration, then petty corruption becomes
customary, sustainable, and endurable”.26

Most important, it becomes a fee, because it guarantees that what was initially a free service is no
longer so. From a macro-economic perspective, its distortionary effect could be at least as impactful
as grand corruption.

The term ‘grand’ corruption, on the other hand, can also be misleading if grand suggests debilitating
to the state. Implicit in the term is the notion of a corrupt deal of significant size, involving senior
officials and high-ranking politicians. Such corruption involves large-scale stealing of state resources
and, the theory goes, it erodes confidence in government, undermines the rule of law and spawns
economic instability. Corruption of this magnitude should cause popular shock and be obvious to
all. Yet, some highly corrupt states seem able to sustain economic and political stability for very
long periods of time, with populations that appear oblivious to the deleterious effects such graft
has on them. This suggests that a form of ‘corrupt legitimacy’ is possible. In turn, this suggests that
there are cases in which the term ‘grand’ corruption fails to communicate the moral shock and
magnitude that seems implicit. ‘Grand’ then becomes merely an audit term that simply describes
financial scale. If that conclusion is right, it would explain the frequent lack of moral outrage about
widespread theft in government, with the result that there will be cases in which characterising
corruption as petty or grand implies nothing about its impact or the social and political levers one
can push to eliminate it.

24 Rotberg, R. I., (2017) The Corruption Cure: How Citizens and Leaders Can Combat Graft p. 11 Princeton, NJ: Princeton University Press.
25 ibid p. 43
26 ibid p. 32

5
The problem of naming is then compounded by medical or sociological language that pathologises
corruption. The touchstone of much anti-corruption reform in Kenya is the assumption that
government is trying to govern but is somehow side-tracked by corruption, understood as a
malignant institutional failure that frustrates the governing effort. Therein lies the problem: anti-
corruption programmes ‘pathologise’ the relationship between corruption and the state, deploying
medical terms like ‘cancer on the body politic,’ a ‘disease that we must cure’ or ‘a pervasive ill’
potentially responsive to curative interventions.

Even when the language used is sociological rather than medical, the pathological dimension
stays. Corruption is a ‘perverse culture’ or a ‘negative norm’. Both the medical and the sociological
language mobilise a deep-seated “conviction that there is something pathological – an illness –
within [Kenyan] politics and culture”.27 This suggests that what reformers must do is, “to identify
this pathology” and formulate a diagnosis that examines Kenyan society and brings to the surface
the “fissures and contradictions”28 that explain the graft. The medical perspective that implies that
the state has gone awry and can be put to rights with an appropriate intervention is pervasive.
Implicit in the diagnosis and the proposed cure is the thought that the state is constructed for some
legitimate – or benign – purpose that has been perverted by corruption. What if we attribute no
benevolent purpose to the state? What if we assume instead, that governing is not the government’s
objective?

State capture
Analysing Kenya’s inability to fight corruption since 1992 and the emergence of both corruption
and showy, but sham investigations since, this study argues that the assumption that state exists
for benign ends, but is debauched by corruption, is deeply mistaken.

What is at play in Kenya is ‘state capture’, defined as a political project in which a well-organised elite
network constructs a symbiotic relationship between the constitutional state and a parallel shadow
state for its own benefit.29 As defined by Catrina Godinho and Lauren Hermanus,30 state capture is:

“… a political-economic project whereby public and private actors collude in

establishing clandestine networks that cluster around state institutions in order
to accumulate unchecked power, subverting the constitutional state and social
contract by operating outside of the realm of public accountability.”

If capture is successful, “state institutions, governance, and functions are repurposed and
re-engineered over time”, and the constitutional state sheds many of its substantive, but not
formal, democratic features and becomes increasingly autocratic, at least in the ways in which
power is exercised de facto.31

27 Bratsis, P 2003 Corrupt compared to what? Greece, capitalist interests, and the specular purity of the state. Discussion paper (8) p.6. Hellenic Observatory, London
School of Economics, London, U.K.
28 ibid
29 In this we follow the perspective developed in the report of South Africa’s State Capacity Research project, Betrayal of the Promise: How South Africa is being Stolen,
May 2017.
30 Godinho, C & Hermanus L 2018 (Re)conceptualising State Capture - With a Case Study of South African Power Company Eskom, p. 3 Conference Paper prepared for
the Public Affairs Research Institute’s State Capture and Its Aftermath: Building Responsiveness Through State Reform, 22-24 October 2018, Johannesburg.
31 ibid p. 4

6
Essential to state capture is the existence of, “a crooked, conniving and reciprocal relationship”32
between certain types of businesses and politicians. State capture creates a two-government
country: “there is an elected government, and there is a shadow government – a state within a
state.”33 Capture networks radiate, web-like, from two centres: the Presidency, where the power is
concentrated, and the Treasury, where the money management is centralised.

Successful state capture networks in Kenya have had two elements. On the bureaucratic side, there
is usually a coterie of favoured officials who are allowed to accumulate, concentrate and exercise
power in completely unaccountable ways, often behind the shield of presidential privilege, state
security or defence procurement. On the business side, there is often a clique of local businessmen
allied to political insiders, or alternatively, the favoured groups are shadowy, international companies
whose shareholders are usually unknown. Capturing and controlling the Presidency – the source
of power – and the Treasury – the source of money – is essential to fashioning the ‘criminal web’
necessary to repurpose government for the benefit of rent seeking elites.34

This state capture perspective requires us to see the state elite, not as a government at all, but as
“a vertically integrated criminal organization”35 that operates in the shadow of the constitutional
state. Taking that view, even political rivals are allies who co-exist uneasily,36 not principally for
the purpose of “exercising the functions of a state” in the abstract, but primarily and concretely
for “extracting resources for personal gain”.37 Given its private objectives, the state, seen this
way, has no interest in public purposes, such as development, education or health.

Thus, we see the institutional dimensions of state capture. But capture has operational and human
resource management dimensions, too. On the human resource side, captured systems routinely
select and employ the unfit, even as they drive honest people out of government. As Chayes asks,
why is it “so hard to find honest people to serve in government?”. The answer is straightforward:
if government operates as a criminal syndicate this should not be a surprise. This perspective
questions the idea that the Kenyan state is recognisable as such, suggesting instead that institutions
are “run as noxious, corrupt criminal fiefs”. The government, “from the presidency to the police” is
‘privatised’ with “criminalized elements operating under unofficial licence from the civil authority”
and the appropriate way to think of the authorities is as “organized criminals” who have perverted
state institutions “to maximize predation.”38 “Mafias”, Chayes notes, “select for criminality, by
turning violation of law into a rite of passage, by rewarding it, by hurting high minded individuals
who might make trouble”.39 Thus, the crooked thrive and the conscientious perish.

As Jacques Pauw says of South Africa, the state capture elite around former President Jacob Zuma
– what Pauw called the ‘President’s keepers’ – drove honest public servants from office “through
discrediting campaigns, trumped up charges, false allegations, malicious rumours and fake dossiers.

32 See Pauw, J (2017) The President’s Keepers: Those Keeping Zuma in Power and Out of Prison p. 72. Tafelberg, Cape Town, South Africa
33 ibid p.78.
34 Betrayal of the Promise: How South Africa is being Stolen, May 2017 p. 2. State Capacity Research Project.
35 Chayes, S (2015) Thieves of State: Why Corruption Threatens Global Security p. 41. W. W. Norton & Company, New York, US.
36 ibid
37 ibid. p.62.
38 ibid p. 25.
39 ibid. p. 62.

7
Their careers were ruined, they were humiliated and shamed, persecuted and prosecuted, and
had their savings exhausted because of malicious litigation”.40 This serves two purposes: to create
‘rational disorder’ in which performance standards and bureaucratic probity can be allowed to
‘slide’, and to create an atmosphere of flux in which all employees are uncertain about their jobs
and careers and are thus susceptible to pressure. High staff turnover at relatively senior levels, is a
good indicator of capture: bureaucratic instability is a litmus test that civil servants must pass – one
shuffles officials until reliable “apparatchiks” who are nothing but “lame ducks and weaklings”41 can
be found.

Why has state capture been so stable in Kenya, even seemingly able to transition through elections?
The capture elites are not self-annihilating and an unstated rule of capture is that successor regimes
will not disturb their predecessors. Corrupt systems have an incentive to “distill and purify their
own criminality”.42 As John of Salisbury wrote in ‘Policratius’43, “the raven rejoices in the work of the
wolf, and the unjust judge applauds the minister of injustice … in lands in whose princes are infidels
and companions of thieves; they hasten to embrace those whose misdeeds they observe, thus
adding their own share of iniquity in the hope that they may gain for themselves some portion of
the spoil”.44 In this sort of state, governing is really “just a front activity” and to that extent the state
is only “failing at being a state” but is otherwise “remarkably effective in achieving its objective”:45
enriching the state capture elite.

To be able to transition ‘capture’ across elections, from one regime to another, as Kenya did in 2002
and again in 2013 and 2017, involves repurposing “politics” so as to limit “the political agency of
citizens.”46 To do this successfully, democracy has to be reframed in purely formal and procedural
terms. The political class is then able to use the democratic process, especially elections, to
frustrate what Michael Johnston calls ‘deep democratization’. Deep democratization as opposed
to procedural democracy is:

“[The] process whereby citizens become able to defend themselves and

their interests by political means. It is “democratization”, not in the sense of
establishing formal democratic institutions for their own sake, but rather in
the sense of broadening the range of people and groups with some say about
the ways power and wealth should – and should not – be pursued, used and
exchanged.”47 (Emphasis added.)

In this way, the procedural elements of democracy are used to hollow out its substantive
commitments whilst keeping the diplomatic respectability that is conferred by regular elections.

40 Pauw, J (2017) The President’s Keepers: Those Keeping Zuma in Power and Out of Prison p.72. Tafelberg, Cape Town, South Africa
41 ibid p. 79.
42 Chayes, S (2015) Thieves of State: Why Corruption Threatens Global Security p. 62. W. W. Norton & Company, New York, US.
43 https://archive.org/stream/JohnOfSalisburyPolicraticusJohnOfSalisbury/John%20of%20Salisbury_%20Policraticus%20-%20John%20of%20Salisbury_djvu.txt
44 Chayes p. 62.
45 ibid
46 Burbidge, D (2015) The Shadow of Kenyan Democracy: Widespread Expectations of Widespread Corruption at p. 20 Ashgate Publishing.
47 ibid

8
In addition to sequestering democratic concepts for private use in this way, political elites also
appropriate moral language and social norms to ‘conventionalise’ corruption, fashioning a vocabulary
that removes the moral sting and opprobrium from corruption and its various forms. Corruption is
‘traditionalised’ and reframed as gift-giving or a form of socially recognisable reciprocity. Corrupt
practices are then expressed in the language of moral obligation. No moral wrong is involved when
an official or politician from one’s village violates conflict of interest rules or other laws to provide
some “token benefit”; or to take up some “obligation to provide mutual assistance” to aunties and
cousins; or to pull strings on behalf of other “kin and friends of friends”.48 Once a moral bond is
accepted as legitimate “a civil servant”49 or politician cannot “refuse to profit” from ‘juicy’ postings”
that might benefit him and his people or “fail to ‘spread the benefit around’ to his relatives”.50 In
this way, the state capture elite syncretically amalgamates traditional practices of gift-giving and
reciprocity with their own corrupt and predatory practices51 that are based on smash, grab and a
dollop of tokenism for friends and relatives.

This leads to another feature of captured states that runs contrary to conventional explanation:
namely, that though corruption generally undermines social trust, it is also possible to find nested
networks of very high-trust relations within a highly corrupt environment characterised by social
anomie and low trust.

This combination of a democracy drained of substance and communal mores, which are purloined
to give social legitimacy to vice, erodes democracy’s emancipatory power and robs the public of
the moral resources they need to confront bad governance. As George Steiner points out, political
falsehood of this magnitude cuts language from its roots in “moral and emotional life”52 so that it
becomes “ossified with clichés, unexamined definitions, and left-over words”.53 According to Steiner,
this ‘anaesthetizing’ drains “the life-force of the language”54 and dissolves the society’s moral and
political values.55 Indeed, language is in a parlous condition when the bribe a judge takes to free a
dangerous criminal is named chai, like a nice ‘cuppa’ tea between intimates.56

Why the state capture perspective makes sense

Seeing corruption as a problem of state capture solves many of the persistent puzzles of anti-
corruption reform in Kenya. Why do the emblematic corruption cases – the Goldenberg, Anglo
Leasing or Eurobond scandals – never get resolved even though they never really die? They are
not meant to be resolved: to resolve them would undo the ‘implicit transition bargain’ of Kenyan
politics that successors will not harm the interests of their predecessors. And yet, these emblematic
cases cannot really be allowed to die because that would expose the capture racket. Therefore,
such cases are kept interminably in the public eye, partly as evidence that ‘action is live’ and partly
as a fig leaf to keep the machinery of larceny functioning under cover.

48 Rotberg, R. I. (2017) The Corruption Cure: How Citizens and Leaders Can Combat Graft p. 28. Princeton, NJ: Princeton University Press.
49 ibid
50 ibid
51 Ibid p. 29
52 See George Steiner, (1984), A Reader. Philology, Addresses Essays Lectures, The Retreat from the Word at p.296.
53 ibid
54 ibid
55 ibid
56 ibid

9
But the real contribution of the state capture perspective lies in its ability to illuminate alternative
paths to reform. The pathological view – that there is a decent state trying to do a good job but
sundered by institutional infirmities – typically suggests solving corruption on some variant of either
the principal-agent theory or collective-action theory.

Principal-agent theories explain the problem in a straightforward manner: leaders (the ‘agents’) are
given authority by voters (the principals), with the expectation that they will act in the best interests
of society. However, because the effort of the agents is not always observable, in that what they
do is not transparent, agents are able to act in ways that are beneficial to them and against the
interests of the principals. Anti-corruption reform programmes that are rooted in the principal-
agent theory57 focus on making the agents’ efforts observable.

In concrete terms, this means enacting policies that do the following: enhance the ability of
principals to monitor and sanction agents; reduce the discretion of agents; support civil society
organisations to be both watchdogs and a countervailing power against state functionaries;
promote transparency; and create mechanisms that alter the individual incentives of agents.
There is compelling evidence that the flood of such reforms that have been adopted over the last
decade or so have been effective only in public finance management.58 Marquette and Peiffer are
surely right in saying that once the principal-agent perspective is adopted then both the ability and
willingness of principals to monitor and hold agents accountable for their actions becomes the
crucial gateway to reform. The theory takes it for granted that principals have the ‘will’ and does
not, therefore, spell out how one may foster this ‘willingness’ where it is absent. It also ignores the
fact that sometimes the principals may want to act but lack the levers of action.

It seems, then, that the principal-agent perspective illuminates some dimensions of corruption but
fails to generate reform measures that work across government or in high impunity environments.

An alternative theory, the collective-action theory, sees corruption as a collective-action problem.

Honest behaviour by one, or a few individuals, rarely makes a difference. In fact, an economy with
high levels of corruption “is bound to move towards high corruption stable equilibrium”.59 In its
most basic formulation, a collective action problem is said to occur when, “even if it is in the best
interest of all individuals in a group to act collectively towards a common goal, group members do
not do so; instead, group members find it in their individual interest to not contribute at all or to
limit their contributions, ensuring that the collective benefit is not realised to its fullest potential”.

On the collective action perspective, high corruption environments generate, according to

Burbidge, “widespread expectations of widespread corruption”, pushing everyone to act corruptly.
This includes those whom the principal–agent theory “refers to as ‘agents’” and those it “refers to
as ‘principals’”. The result is a low trust society, which helps sustain “systemic corruption”. Honest
government is a public good. Public goods have two features that make them prone to collective-
action problems. One, it is impossible to exclude a member of the public from consuming them.

57 Marquette H & Peiffer C (2015) Corruption and Collective Action, DLP Research Paper 32, University of Birmingham at p.2
58 ibid p. 2
59 ibid p. 44

10
They are non-excludable. Two, one person’s use of a public good does not reduce the amount
available for other people to enjoy. They are non-rivalrous. Public roads are a good example of such
goods.

The psychological insight of collective-action theories is the fact that consumers of goods who
know that they cannot be excluded from benefit have no incentive to invest in producing the goods
– they are free-riders. People “free-ride when they know their level of contribution towards the
collective good does not impact their beneficiary status”.60 The risk that some people will free-ride
is the nub of the collective-action problem.

Corruption poses free-riding problems in two ways: on the service delivery side and on the reform
side. On the service delivery side, people engage in corruption because they put personal interest
above collective interest. Taking this view, systemic corruption persists wherever “corruption is
widely perceived to be the norm … and individuals gain little from abstaining from or resisting
corruption if they cannot trust that others will do the same”.61 As Marquette and Peiffer point out,
corruption often provides solutions to real problems that people face. This can institutionalise free-
riding and deprive society as a whole of a ‘corruption-free environment’ or of quality government,
defined as one that treats all citizens impartially and equally. This can be particularly insidious when
the group that benefits from corruption is able to transfer the costs of corruption to a weaker
or excluded political group. In this way, corruption becomes an “institution that feeds off and
reinforces itself”,62 possibly acting to maintain the corrupt status quo.

On the reform side, people abstain from reform advocacy and free-ride on the efforts of the
collective, hoping to benefit from a low corruption environment when and if it is ever achieved.
This is one reason why democracy has been particularly ineffectual in solving collective problems
like corruption. It works in two ways: one, if individuals think that their vote won’t make a difference
they may be discouraged from voting; two, and more insidious, is the political impact of low levels of
social trust. In high impunity, corrupt environments, many voters prefer the corrupt politician from
their in-group to the honest politician from the out-group – an institutionalised form of adverse
selection.

What this means is that the arsenal that democracy mobilises to fight corruption and
institutional failure (political legitimacy; robust laws; a mobilised and informed population
opposed to graft; vibrant media and civil society; a fearless and independent judiciary and
effective prosecutorial and oversight agencies63) is not available if democracy itself has already
been subverted. This, in fact, is probably the explanation for the blossoming of corruption and
its ever more ghastly character with each successive election since 1992.

60 Marquette H & Peiffer C (2015) Corruption and Collective Action, DLP Research Paper 32, University of Birmingham p. 3
61 ibid p. 1 University of Birmingham, UK
62 Rotberg, R I (2017) The Corruption Cure: How Citizens and Leaders Can Combat Graft p. 44. Princeton, NJ: Princeton University Press .
63 ibid p. 78. (Rotberg’s checklist)

11
Part 2

TALES OF STATE CAPTURE:

GOLDENBERG, ANGLO LEASING,
EUROBOND
Corruption and politics, never the twain shall part
Politics and corruption have always been intimates in Kenya since independence. Little wonder
that the first commission of inquiry appointed after independence, the 1965 Chanan Singh Maize
Commission of Inquiry, was triggered by a corruption scandal involving Paul Ngei, the then Minister
for Marketing and Cooperatives. Mr Ngei had permitted his wife Emma Ngei, through her company
Uhuru Millers of Kangundo (commonly referred to at the time as Emma Stores) to directly buy maize
from farmers, bypassing the Maize Marketing Board, which he chaired. This was despite the fact
that the law did not allow Kenyans to buy maize straight from farmers (cheaper than buying from
the government). Worse still, Ms Ngei was permitted to buy 2,000 bags of maize, but she refused to
pay for them, marking payment demands as ‘return to sender’. She refused to remit the differential
between the farmer’s and the government prices to the Board, which too was against the law.
Widespread speculation in maize by well-connected individuals, combined with official failure to
import more maize in time led eventually to a national shortage.64 The commission of inquiry was
appointed by President Kenyatta to investigate the cause of the maize shortage. Mr Ngei, because
of his relationship with Uhuru Millers (owned by his wife) was briefly suspended from the Cabinet,
but was later reinstated.

Maize, then, before and since has had a long career in both politics and corruption.65 That first
scandal set the tone for future graft: the politically connected rig the system to benefit themselves,
their relatives and their cronies. When they are outed, they resort to inconclusive methods of
investigation such as commissions of inquiry, task forces or inept prosecutions. The difference
between that early corruption and the corruption described in this study as state capture is the fact
that most of it involved abuse of discretion and conformed closely to Robert Klitgaard’s definition:
Corruption = “Monopoly + Discretion – Accountability”.66

The first corruption scandal encompassing major characteristics of state capture was the
Turkwel Gorge hydro-electric power project between 1986 and 1991. Many aspects
of the process of contracting for this project entailed rigging and repurposing legal
processes for the benefit of President Moi and his cronies. According to an internal

64 http://kenyalaw.org/kl/fileadmin/CommissionReports/Report-of-the-Maize-commission-of-Inquiry.pdf
65 See AfriCOG (2009) The Maize Scandal.
66 Klitgaard R (1988) Controlling Corruption. University of California Press, US.

12
European Commission Memorandum of March 198667 written by Achim Kratz, the then
Commission’s delegate to Kenya, the contract price for the project was more than double
the amount Kenya’s government would have paid under an international competitive tender.
The memo stated that the government knew that Spie Batignolles, the French contractor’s price
was extortionate, but hired them nevertheless, “because of high personal advantages”. Those
“personal advantages” were millions of dollars paid to President Daniel Arap Moi and to the then
Minister of Energy, Nicholas Biwott. Moreover, companies associated with people close to Moi
and Moi’s family were sub-contracted to execute many elements of the Spie Batignolles contract.
The effect of the combination of personal interest and inattention to geological and hydrological
factors was that when the project was finally commissioned by President Moi in October 1993, the
reservoir was under 25 percent full and the project had already cost three times the estimated cost.
The knock-on effect was probably even greater: the Turkwel corruption provoked donors to cut
funding to the energy sector, which would eventually generate the crippling power outages of the
mid 1990s to the early 2000s.

Some of the lessons learnt from the Turkwel Gorge saga on repurposing state institutions and
lawful processes to extract regime and personal gain would be applied with a vengeance to the
first unambiguous case of state capture: the Goldenberg scandal.

Goldenberg and other post-1992 state capture scandals marked a watershed in corruption politics:
that is, “a shift from a politics of state-led control” to “a politics of competitive aggression, the
principal victim of which has been the state itself”.68 In this dispensation, politics “has become
a pursuit of ever faster forms of enrichment”.69 These scandals challenge both the theory of
democratic consolidation and the supposed ‘curative’ effects of democracy on political corruption.
Kenya under democracy has transformed from “a moderately corrupt society to a pervasively
corrupt society”70 and corruption has moved stepwise from “petty or bureaucratic”71 to “grand
or political”72 and eventually to state capture. With this shift, the state has become a “zone for
personalised appropriation”.73

The success of state capture rests on the ability of a small group of powerful
and rich operatives to take over and pervert the institutions of democracy, while
keeping the façade of a functioning democracy. Thus, oversight institutions are
weakened; law enforcement is partisan and in the pockets of politicians; civic space
is asphyxiated; free elections are frustrated and are typically won by the most
violent or the most corrupt, or those who are both violent and corrupt. Arrests
and indictments are often precursors of inaction, not proof of official will to fight
corruption.

67 Details are set out in appendices to a document laid before the House of Commons at https://publications.parliament.uk/pa/cm200001/cmselect/cmintdev/39/39ap06.
htm
68 Burbidge, D (2015) The Shadow of Kenyan Democracy: Widespread Expectations of Widespread Corruption p. 33 Ashgate Publishing.
69 ibid p. 33
70 ibid p. 34
71 ibid p. 34
72 ibid p. 34
73 ibid p. 34

13
A scandal, once exposed, is followed by a tumult of indignant editorials. Often a period of nervy
and protracted official stonewalling follows. Politicians make fiery anti-corruption speeches,
eventually, a few high profile arrests and indictments are made and jubilant headlines follow.
Over a period of months, the cases are throttled and indictments quietly dropped. Senior officials
previously interdicted are quietly reinstated. If a new government is elected, it aggressively restarts
investigations into the mega-scams of its predecessor partly to cover up its own mega-scams, partly
to build its anti-corruption credentials, and partly to eliminate potential competitors in the state
capture game.

Lessons from state capture scandals

What the stories of procrastination on the Goldenberg, Anglo Leasing and Eurobond scandals show
is the evolution of Kenya as an institutionalised kleptocracy; that is, “a system of state capture
in which ruling networks and commercial partners hijack governing institutions and maintain
impunity” for the purpose of raiding the budget to sustain themselves in power and for the security
of the regime.74 We can distill the following conclusions about the principal characteristics of state
capture. These are:
1. Capture depends on control of the Presidency and operates on the rule that no one should
be allowed to threaten the President and other men of power. Goldenberg drew in the most
powerful men in the Moi government: the President himself, his deputy, his security chief and
some of the most powerful members of Cabinet. Given that powerful insiders risk their all once
they are out of power, succession planning and careful preselection of one’s successor is a vital
part of what keeps capture alive.
2. Once the Presidency is compromised by mega-corruption – as Moi was by Goldenberg, Kibaki
by Anglo Leasing and Uhuru by the Eurobond scandal – the whole government machinery
becomes completely permissive towards corruption. Impunity becomes the glue that holds the
system together and the impunity of low-level functionaries is the price the bosses pay to avoid
subversion from within.
3. The success of state capture relies on parallel informal structures, based on the Presidency,
that subvert the constitutional institutions whilst maintaining the outward forms of those
institutions. These informal networks are serviced with cash and loyalty.
4. Control of the judiciary and all independent offices is crucial to successful state capture. If these
cannot be controlled – through bribes, if possible, or coercion if bribes fail – then budget cut-
backs can be used to undermine their effectiveness. Personalised attacks on individual judges
can also be considered.
5. The police are the eyes of the captured state and must be kept sweet, which typically implies
looking the other way as the police supplement their incomes by predation. As Mobutu once
told his soldiers in the DRC, “You have guns, you don’t need a salary”. Police can become as
wealthy as their appetites allow by preying on the little guys as the police vetting exercise, held
from 2013- 2018, discussed below, revealed.
6. Regime fixers, high-level officials and close allies must benefit from corruption and junior
functionaries must understand that their ability to pass payments up the chain is their sacrament
of submission and without it they are vulnerable.

74 See Sasha Lezhnev, A Criminal State: Understanding and Countering Institutionalized Corruption and Violence in the Democratic Republic of Congo, a report from
Enough Project, 27 October 2016.

14
7. Big projects are good politics and good business for capture. State capture depends on financing
mega-projects and under-spending on services unless these services have ‘high spend’ on
equipment.
8. When in doubt, the state capture elite gets busy. Hectic, but controlled actions on corruption
ensue, creating uncertainty and simultaneously multiplying the veto points at which real action
on corruption can be plausibly thwarted, (for example, by bogus investigations that guarantee
acquittals and thus deflect blame to the courts). Alternatively, informal institutions or quasi-
judicial mechanisms, such as commissions of inquiry, are created that cannot make binding
decisions.

With that in mind, we may now turn to events themselves, to illustrate the ways in which these
features are embedded in the three largest scandals in Kenya’s history.

Goldenberg – designing the methods of state capture

In 1991 and 1992 Kenya underwent a foreign exchange crunch. The proximate cause for this
was mounting pro-democracy pressure by the opposition and civil society groups, to which the
government responded with violent crackdowns. Political repression and donor concern about
corruption, combined with poor export performance of the leading foreign exchange earners
of coffee, tea and tourism, led to a significant drop in hard currency reserves. The government
responded to this with an export promotion scheme in which exporters who deposited their hard
currency earnings would not only receive the Kenya shilling equivalent of their deposits, but also
an additional 20 percent ‘export incentive’. Goldenberg International, a company jointly owned by
Kamlesh Pattni and the then director of the special branch, (Kenya’s secret service) James Kanyotu,
concocted a scheme to export gold and diamonds to three companies in Dubai and Switzerland
on an understanding that they would be paid 35 percent ‘export compensation’. The problem was
that gold and diamonds were not covered in the Export Compensation Act and the ‘incentive’ paid
to the company was 15 percent above the lawful limit. The real scandal, though, was that Kenya
had no diamonds and its gold mining was insignificant. In the beginning, Goldenberg International
exports turned out to be entirely made up of gold smuggled from the Democratic Republic of
the Congo (formerly Zaire). Later, the company stopped smuggling gold altogether and merely
completed export declaration forms, produced fake hard currency deposit slips and got paid, not
only the coupon amount on the fake deposit slips, but also the 35 percent export compensation.

The total cost of the scandal is unknown, but some estimates are that up to 10 percent of Kenya’s
GDP was lost. The 2006 Bosire Commission of Inquiry into the scandal concluded that up to Kshs
158.3 billion of Goldenberg money was transacted with 487 companies and individuals. This is
probably a gross underestimate, as in fact Goldenberg was a series of inter-connected financial
scandals rather than the phantom exports of gold and diamonds that most investigations have
focused on since 1992. (The scandal was first revealed in the Controller and Auditor General’s reports
for 1991 and 1992.) According to various affidavits sworn by the main suspect in Goldenberg and
associated scandals, the beneficiaries of these dealings included President Moi, his vice-president
and his business associates.

15
Notwithstanding revelations in the Controller’s and Auditor General’s Reports together with whistle-
blower accounts covered in the media, the government initially stonewalled. This prompted the
Law Society of Kenya (LSK) to seek the permission of the High Court to file a private prosecution to
remedy the inaction of the Attorney General (AG). The AG, Amos Wako, suddenly bestirred himself,
asking to join the LSK case as a friend of the court. He promptly opposed the LSK’s application,
arguing that he had been delayed by investigation reports and requesting the LSK to hand him
such evidence as they had so that he may act. Backed by an affidavit by Japhet Masya, Clerk to the
National Assembly, the AG also argued that the High Court had no jurisdiction on Goldenberg given
that the issue was before a committee of Parliament.

Mr Wako’s pleas were both inexplicable and disingenuous: Parliament has no criminal jurisdiction
and any policy issue on Goldenberg pending before one of its committees can have no effect on an
indictment for corruption. The AG sounded more like a defence attorney than the head of public
prosecutions and guardian of public interest that he was. Dr Willy Mutunga, then chair of the LSK,
feared that Mr Wako’s ruse was proof that the government was “determined to complete the
Goldenberg cover-up”. Mr Wako, he predicted, would continue to act like “counsel for all the accused
persons” and would engineer “protracted delays” “mention after mention, adjournment followed by
adjournment”, ending in a “dramatic withdrawal of the cases”. So it proved. The magistrate, Uniter
Kidullah, appointed the Director of Public Prosecutions after her decision in this case, rendered a
rude and intemperate judgment, combining otiose proceduralism with personalised insults against
the LSK: Mr Mutunga’s pleadings were inadmissible because he, rather than the secretary, had
signed them; the LSK had no legal standing to file a private prosecution since it could not show how
its interests had been harmed by the Goldenberg scandal and, so far as she could see, the LSK had
acted outside its statutory mandate; finally, she concluded that the only knowledge LSK seemed to
have was that of “stealing from ... clients”.

There the Goldenberg scandal would have died but for the government’s continuing hard currency
crisis. The International Monetary Fund (IMF) and the World Bank warned Kenya that no new
programme would be agreed with the country until the government took credible action on
corruption in general and on Goldenberg in particular. It was this threat that spurred Attorney

David Sadera
Munyakei
Goldenberg
Whistleblower
1968 - 2006
16
General Amos Wako to indict Pattni and his co-accused in 1997, five years after the scandal first
broke. But the charge was not meant to result in effective prosecution. Against the advice of
his DPP, Bernard Chunga, the AG framed more than 90 counts in one charge in the face of clear
precedent that so many counts would invalidate the charges. Knowing this, in July 1997 Kamlesh
Pattni challenged the charges as illegal and was granted an order of prohibition by the High Court,
stopping the trial. Donors, aghast at this turn of events, refused to lift the conditions they had
imposed on aid to Kenya until Goldenberg was properly prosecuted.

A chastened AG filed new charges in August 1997, calculated to be good optics for an IMF mission
that was expected in Nairobi in early 1998. In the meantime, Mr Pattni had concocted a new fraud
to defeat any fresh charges that the AG might bring against him. Using forged papers, fake sale
agreements back-dated to 1992 and with the connivance of the Registrar of Companies – in the
Attorney General’s Chambers – Mr Pattni purported to be the owner of World Duty Free (WDF), the
Isle of Man company to which he claimed to have sold the gold and diamonds. He then obtained
court orders allowing him to take over management of WDF shops in Kenya. The point of this devious
scheme was that in a future prosecution Pattni could argue that as the owner of WDF he couldn’t be
forced to testify against himself. Armed with this new civil suit, he challenged the fresh indictments,
claiming these charges should be stopped as prejudicial to the WDF civil case. The court agreed with
this risible claim, even though legal principle works the other way: where a criminal case raises the
same issues as a civil case, the criminal case is heard first. There are two reasons for this: one, the
public interest should be vindicated before the private interest; and, two, given that the standard
of proof in criminal cases – beyond reasonable doubt – is much higher than the standard in civil
cases on a balance of probabilities, it is more efficient to hear the criminal case first, since facts
proved need not be proved again in the related civil case. This botched 1998 prosecution was the
last action that the Moi government took to resolve the Goldenberg scandal.

In 2003, President Mwai Kibaki succeeded Daniel arap Moi. He quickly set up a commission of inquiry
into the Goldenberg scandal, ironically just about the same time that his own cronies were busy
siphoning monies out of Kenya under the Anglo Leasing scandal. The commission was chaired by
Justice Samuel Bosire, later to be pushed out as unfit to be a judge during the vetting of magistrates
and judges mandated by the 2010 Constitution. The Bosire Inquiry established what everyone always
knew but could not prove, because the AG, Amos Wako had developed feet of clay. Goldenberg,
the commission concluded, involved the highest levels of President Moi’s government and Moi
had personally authorised two Goldenberg related payments. After the inquiry, the government
imposed travel bans on 21 people75 named by the commission as connected to Goldenberg. Bosire
also recommended that retired President Moi’s role in Goldenberg be investigated. Nothing came
of either the travel ban or the Moi investigation. In August 2006, the credibility of the report
was seriously dented when Professor George Saitoti (formerly vice-president to Moi), who the
commission had found culpable enough to warrant an indictment, got a court order expunging his
name from that list of shame.

75 1) Gideon Moi, retired President Moi’s son, then MP for Baringo Central; 2) Philip Moi, retired President Moi’s son; 3) Mr Moi’s lawyer, Mutula Kilonzo; 4) Mr Moi’s
former personal assistant, Joshua Kulei; 6) former Central Bank of Kenya governor Eric Kotut; 7) former Central Bank of Kenya deputy governor Eliphas Riungu; 8)
former Central Bank of Kenya employee Job Kilach; 9) former Central Bank of Kenya employee Tom Werunga; 10) former Central Bank of Kenya employee Michael
Wanjihia; 11) former Director of Public Prosecutions Philip Murgor; 12) former Treasury Permanent Secretary Charles Mbindyo; 13) former Treasury Permanent
Secretary Wilfred Karuga Koinange; 14) former Treasury Permanent Secretary Joseph Magari; 15) Prof George Saitoti, who resigned as education minister; 16)
Goldenberg architect Kamlesh Pattni; 17) Pattni’s business partner and former Special Branch chief James Kanyotu; 18) former commissioner of Mines and Geology
Collins Owayo; 19) Arthur Ndegwa, senior mining engineer in the Commissioner of Mines Nairobi office; 20) former commissioner of Customs and Excise Francis
Cheruiyot and 21) former Kenya Commercial Bank general manager Elijah arap Bii.

17
In the end, no one was ever convicted for any of the Goldenberg crimes. In 2006, six months after
the release of the Goldenberg Report, the man who first blew the whistle on the scandal only to be
hounded into destitution for his efforts, David Munyakei, died, a lonely and forgotten victim of the
forces of state capture.

The Anglo Leasing Scandal

The Goldenberg script would be reprised in the second state capture case, the biggest scandal
of the Kibaki era: the 2003 Anglo Leasing scandal. Anglo Leasing was a series of security related
scandals involving 18 state security contracts, collectively worth about US$770 million (Kshs 55
billion), in which the government entered lease finance and suppliers’ credit agreements to pay
for the following: forensic facilities; security equipment and support services for Kenya Prisons;
the Police Airwing; the police force; the Directorate of Criminal Investigation; the administration
police; the National Security Intelligence Service, (NSIS), and the National Counter Terrorism Centre.
Thirteen of the eighteen contracts were made under President Daniel arap Moi, the other five
after 2002, under President Mwai Kibaki. The true identities and whereabouts of the companies
remained unclear. Though the immediate investigation that blew open the scandal involved the
Anglo Leasing and Finance Company, in truth the scandal involved many more companies owned
by the same set of individuals: Deepak Kamani; Anura Perera; Amin Juma; Merlyn Kettering and
Ludmilla Katuschenko.

Within these 18 generally irregular contracts, individual contracts were even more blatant in their
irregularity: the contract for tamper-proof passports granted to Anglo Leasing and Finance Company
was described by the Public Accounts Committee (PAC), ironically chaired by Uhuru Kenyatta,
as “an organised, systematic and fraudulent scheme designed to fleece the government through
the so called special purpose finance vehicles for purported security contracts”.76 How exactly Anglo
Leasing got involved in these security contracts is unclear from the records, but the pattern is clear.
Let’s start with the immigration contract.

In 2000, the Department of Immigration did a ‘computer needs assessment’ that concluded that to
eliminate fraud, forgery, inefficiencies and revenue loss it would need to procure a passport issuing
system. This was to be done by restricted tender. The Ministerial Tender Committee invited five
international firms to submit bids: two British firms, De La Rue Identity Systems and AIT International
PLC; South Africa’s Face Technologies; Setec OY of Finland and Johannes Enschede of Netherlands.
Three firms responded. The decision was that AIT International PLC met both commercial and
technical specifications for the award.

However, the ministry’s budget for the 2000/2001 financial year did not cover the Kshs 622,039,944
contractual sum that MIS AIT International PLC gave as the cost of the system. The procurement
was deferred to 2002/2003. Six international firms were now invited to bid, the initial five and GET
Group of the USA. Once again, three responded: De La Rue Identity Systems; South Africa’s Face
Technologies and GET Group. The previously successful group, AIT International PLC, did not submit
a bid. A technical committee of the Government Information Technology Services concluded that
none of the bids were responsive and subsequently recommended that they not only be disqualified

76 This account is summarised from the Report on Special Audit on Procurement of Passport Issuing Equipment by the Department of Immigration, Office of the Vice-
President and Ministry of Home Affairs, pp. 1-5

18
but also that, “the system be redesigned and expanded to cover other aspects of the work of the
Immigration Department, such as border controls and immigration monitoring”.

It was now agreed that the expanded system would have five components: 1) high security new
generation passports; 2) secure passport issuing system; 3) high security new generation visas; 4)
high security visa issuing system; and 5) computerisation of machine readable immigration records.
One consequence of expanding the system was a spiking of costs, which would require the Treasury
to seek donor funds.

That is how matters stood when on the 1 August 2003, a firm named Anglo Leasing and Finance Ltd
of Alpha House, 100 Upper Parliament Street, Liverpool L19 AA, UK, sent an unsolicited technical
proposal to the permanent secretary (PS) in the Vice-President’s Office to supply and install an
“Immigration Security and Document Control System, (ISDCS)”. The installation would be done by
a sub-contractor of Anglo Leasing, Francois-Charles Obethur Fiduciare of Paris,77 France. To ease
the funding problem, Anglo Leasing would offer a facility of Euros 31,890,000 (equal to Kshs 2.67
billion) to be repaid at an interest of 5% (later 4%) over a 62-month period.

On review, the PAC thought this highly irregular: a financing firm had prepared a detailed proposal
for a project very similar to the one recommended by the Government Information Technology
Services without a request from the government and most curiously, in a manner that strongly
suggested that the firm “had fore-knowledge of the recommendation to enhance and expand the
system”. Nonetheless, a month later, on 5 September 2003 the Vice-President’s Office asked the
Treasury to contract Anglo Leasing. That permission came through on 25 November 2003. Also on
5 September, the Vice-President’s Office sought legal clearance from the AG’s Chambers and in a
letter of 18 September 2003, the AG advised the ministry to do due diligence. For example, how
many projects of this magnitude had Anglo Leasing successfully undertaken? What was the firm’s
credit rating? The PAC did not see any evidence that tests had been undertaken or that the ministry
had assessed the “authenticity, capacity, experience and track record of Francois-Charles Obethur
Fiduciare”.78

Even with all these things still outstanding, the government signed the Suppliers Services and
Financing Credit Agreement for the ISDCS on 4 December 2003 and two months later, on 4
February 2004, a sum of Kshs 91,678,169.25 described variously as “arrangement” “commitment”
and “administration” fees was paid out to Anglo Leasing.79 According to John Githongo’s dossier to
the President80, all the Anglo Leasing type shell companies were probably established by one Pritpal
Singh Thethy, an accountant and engineer, who was associated with Anura Perera. Thethy’s job was
to set up shell companies, as he had done for General Kibwana, Anura Perera and Deepak Kamani.

77 Owned by the Savare Family of Switzerland, Francois-Charles Oberthur Fiduciaire was later involved with Safran Morpho, the contractor for the IEBC servers for
the 2017 elections. The firm would then merge its activities and card systems to form Oberthur Technologies in 2011 with third generation, Thomas Savare as CEO.
At the time the company had global revenues of €979 million and its 2011 rationalisation plans were initially meant to fund a bid to buy out De La Rue, a UK-based
banknote printing company.
78 Report on Special Audit on Procurement of Passport Issuing Equipment (2006) by the Department of Immigration, Office of the Vice-President and Ministry of Home
Affairs, Public Accounts Committee.
79 See Report on Special Audit on Procurement of Passport Issuing Equipment by the Department of Immigration, Office of the Vice-President and Ministry of Home
Affairs, p. 4.
80 After going into exile on 24 January 2005, John Githongo, the former Permanent Secretary for Governance and Ethics in the Office of the President sent a written
report to then President Mwai Kibaki on 22 November, detailing his investigations of the Anglo Leasing scandal, which became known as the ‘Githongo Dossier’. See
http://news.bbc.co.uk/2/shared/bsp/hi/pdfs/09_02_06_kenya_report.pdf

19
These companies routinely won large contracts to supply goods and services at inflated prices to
the security services and were famed for paying generous kickbacks.

The unravelling of Anglo Leasing began when Maoka Maore, MP for Ntonyiri, tabled documents
in Parliament in April 2004, showing that Anglo Leasing and Finance Company Limited had been
paid a Kshs 91 million commitment fee, amounting to 3 percent of a Kshs 2.7 billion contract to
produce the tamper-proof passports. The department of governance and ethics, headed by John
Githongo, tried to get to the bottom of the affair. In that same month, whilst on a visit to the United
Kingdom he asked Kroll Associates to do some due diligence on Anglo Leasing and discovered
that no such company existed. Githongo had begun to suspect that very senior officials in the
Kibaki administration were involved. Early suspects included Moody Awori, Vice-President; Kiraitu
Murungi, Minister for Justice and Constitutional Affairs; David Mwiraria, Minister for Finance; Chris
Murungaru, Minister for Internal Security; PS Home Affairs, Sylvester Mwaliko, PS Finance, Joseph
Magari, PS Internal Security David Mwangi, Alfred Getonga, Deepak Kamani and Jimmy Wanjigi.

From an early stage in a series of private meetings, the Vice-President as well as the ministers for
justice and for finance, assiduously tried to stop the investigation, partly based on the theory that
“the Vice President had already given a parliamentary statement”. The scale of Anglo Leasing and
the depth of its penetration into the inner sanctum of power would become much clearer over
the next few months. It turned out that even as investigations kicked off, additional payments and
commitment fees were being processed.

When these stories hit the media, the then Secretary to the Cabinet, Francis Muthaura, said that
Anglo Leasing had contacted him and promised to repay the monies they had already received.
Shortly thereafter, on 14 May 2004, Anglo Leasing and Finance Ltd, wired back Euros 956,700 from
Schroder & Co Bank AG in Zurich. Investigations would reveal even more dirt. By early June, inquiries
had established that Anglo Leasing had been paid US$5 million for a forensic laboratories contract
for which they had done no work. The brains behind the revival of this Moi-era contract were Deepak
Kamani, Jimmy Wanjigi, Chris Murungaru, Dave Mwangi, Alfred Getonga, and C. Oyula, the Financial
Secretary. It was clear that there were many more Anglo Leasing type contracts, and eventually 16
of them would become public. The case of two of these Anglo Leasing type companies – Sound Day
Corporation and Apex Finance Corporation – closely followed the conspiratorial modus operandi of
the contracts for the tamper-proof passports. The two companies, which were managed by Brian
Mills, a US national, had signed four contracts, cumulatively worth more than US$145 million.81
According to newspaper accounts, the three Kamanis – Chamanlal Kamani, Deepak Kamani and
Rashmi Kamani – became directors of Sound Day in April 1990. Sound Day, like other Anglo Leasing
companies, was to provide credit, as well as supply the equipment to be financed through that
credit. However, the contract terms were that the equipment would not be supplied until the
government paid the first instalment. Sound Day provided no credit, but charged 3 percent interest
on this ‘financing’ whilst in fact, the financing was the money that had been advanced by the Kenyan
government in the first place. This Byzantine arrangement was later described in court as a “classic
case of reverse financing”.82

81 That is according to the Special Audit Report of the National Assembly.

82 See “Kamani family linked to ghost firm, Sound Day Corporation”, Daily Nation, 4 April 2016 at https://www.nation.co.ke/news/Family-at-centre-of-Anglo-Leasing-
linked-to-ghost-firm/1056-3144806-u00xja/index.html

20
As Anglo Leasing unravelled, the attempts to stop investigations got both frantic and menacing.
Minister for Finance Mwiraria indicated that he would not lay before Parliament a damning special
audit report compiled by the Controller and Auditor General until the Treasury had made some
‘major changes’. The Minister for Justice, Kiraitu Murungi, weighed in with the caution that
Mr Githongo should be careful not to ‘knock out key political people’ like Alfie (Alfred Githonga)
and Murungaru given that both were “key players at the very heart of government”. He would
later add that, “if Chris (Murungaru) is dropped and Alfie (Gitonga) is dropped we are in trouble,
the enemy will have won”. According to him, people were concerned that John Githongo “did not
appreciate the political costs of his work”.

A different politician was later to emphasise these warnings, saying that if Githongo’s investigations
threatened the “stability of the regime” then the President would stop backing him. Both Mwiraria
and Kiraitu said that they hoped that the investigations would stop as soon as Anglo Leasing repaid
the money. In time, the cover up efforts would turn bizarre: Francis Muthaura even questioned
the legal authority of the Kenya Anti-Corruption Commission (KACC) to conduct the investigation
and implied that the Anti-Corruption and Economic Crimes Act was not reasonable legislation,
ostensibly because of the broad powers it gave to KACC.

What the pressure on Githongo and the repayment of the money on the publicly known contracts
revealed, was a clever ploy to fob off investigators on the more numerous, yet unknown contracts
by issuing a mea culpa on what was then publicly known.

One issue surrounding the scandal is what President Kibaki knew and when he knew it. For instance,
on the forensic labs contract, the secretary to the Cabinet had indicated to Githongo that he had
briefed the President on this contract, but when Githongo met the President on 29 May, 2004
Kibaki said that no one had briefed him and asked to be furnished with a copy of the contract. Two
days later, Muthaura would insist that the President had been fully briefed and that it had been
agreed (i) all payments were to be stopped and (ii) that the authorities must establish who Anglo
Leasing was.

John Githongo, Permanent

Secretary for Governance and
Ethics, 2003-2005

21
Later still, Mwiraria would claim that the President had requested that they `go easy’ on Anglo
Leasing given that the money had now been returned. Mwiraria and Kiraitu would argue that if
the public were to know that there were other corrupt deals of this magnitude, “our government
would fall”. Had the President in fact said this or were Mwiraria and Kiraitu using the authority of
the Presidency to smother inquiries? Had the President lied when he said to Githongo that he had
not been briefed?

From the determined opposition to his inquiries, the lukewarm support he got from the President
and threatening messages that he received throughout this early phase of the investigation,
Githongo feared for his life and went into self-imposed exile in the UK in 2005. His conclusion was
that the Anglo Leasing scandal went all the way to the top and that its baseline was a scheme to
finance the 2007 election.

In November 2005, President Mwai Kibaki finally acted. He dropped Chris Murungaru from the
Cabinet. On 1 February, he dropped David Mwiraria and a fortnight later he had ‘accepted’ Kiraitu
Murungi’s resignation. Though 80 MPs demanded that the President fire his Vice-President, Moody
Awori, the President demurred. As with Goldenberg, the government imposed the usual travel
bans on the principals and announced that it would also freeze their assets. Whether this happened
or not is unclear; there is no official indication that it did.

In 2007, the UK’s Serious Fraud Office tried to get to the bottom of a US$30 million transfer by
Apex Finance, one of the Anglo Leasing companies, made between April 2002 and February 2004
through the Channel Island tax havens of Jersey and Guernsey. But by 2009 this effort had petered
out, partly due to obstruction by Kenya. That same year, authorities in Switzerland launched
investigations into Swiss companies named in the scam and froze their bank accounts. It, too, came
to naught. By the time President Kibaki served out his two terms in 2013, no action had been taken
on Anglo Leasing.

The next time Anglo Leasing was in the news, it was early 2014, ahead of the country’s debut launch
of a US$2 billion sovereign bond. Half of this amount would disappear into thin air in the biggest
scandal of the Uhuru presidency, as described below. The facts were as follows. Kenya had lost a
lawsuit in Geneva filed by two Anglo Leasing companies linked to Anura Perera, First Mercantile
Securities Corporation and Universal Satspace. Perera was one of the suspects named in the 2006
special audit of Anglo Leasing. It then turned out that the country had to pay Kshs 1.4 billion to
improve its credibility with international markets by clearing its (ostensible) debts in preparation for
the launch of its debut in the foreign sovereign bond market, the Eurobond. This was odd for two
reasons. First, there was also a contrary judgment from the High Court in Kenya. Justice Mathew
Anyara Emukule had ruled in 2012 that the two companies were non-existent entities that could
not sue. Second, the government had claimed that the contract was vitiated by bribery and there
was a PricewaterhouseCoopers (PWC) audit showing that the goods were over-priced and some
had never been delivered, even though payments had been made83. The Geneva court rejected
these PWC findings.

83 Daily Nation, 15 May 2014, “Government statement on settlement of Anglo Leasing debts” https://www.nation.co.ke/news/1056-2316374-my00txz/index.html

22
As a matter of Kenyan law, the government had paid this large sum to non-existent parties.
According to Treasury Cabinet Secretary Henry Rotich, it was necessary to pay out this amount
lest the country suffer huge interest penalties. The Deputy Solicitor General, Muthoni Kimani,
buttressed the Treasury’s argument with the claim that the Anura Perera litigation in Switzerland
had adversely affected the issuing of the sovereign bond. Hot on the heels of this payment,
National Treasury PS, Kamau Thugge, told the Public Accounts Committee that Mr Perera was now
demanding an additional Kshs 3.05 billion for services given to the National Security Intelligence
Service, now the NIS. (According to Thugge, Perera’s new demand related to another project,
Flagstaff National Counter Terrorism Centre, that the government had contracted in 2004 at a cost
of US$41,800,000.84)

A payment of US$16.4 million to Deepak Kamani in 2014, also purportedly to facilitate the launch
of the Eurobond, seems to have triggered the government’s interest in prosecuting the Anglo
Leasing principals. In March 2015, 11 years after the scandal broke, 13 people connected to Anglo
Leasing including businessman Deepak Kamani and former minister Chris Obure, now a senator,
were indicted.

The prosecution might be explained by President Kenyatta’s fury at the US$16.4 million (Kshs1.6
billion) Kamani payment and the extra Kshs 3.05 billion being demanded by Perera. In addition,
some pressure seems to have come from Switzerland. Jacques Pitteloud, Swiss ambassador to
Kenya, told the Financial Times that Switzerland was tired of suffering reputational loss as a safe
haven for stolen money. But the real political reason could well be that prosecuting Anglo Leasing
deflected attention from scandals involving the friends and relatives of Mr Kenyatta. None of the
targets of the Anglo Leasing indictments were connected to the Kenyattas.

As with Goldenberg, none of the arrests and indictments have so far led to convictions. This script
of never holding to account those involved in state capture scandals, would be replayed by Uhuru
Kenyatta, as President, when he was himself caught up in the Eurobond scandal.

President
Kenyatta
authorises
Anglo
Leasing
payments
84 The Standard 23 May 2014, “Shock as Treasury Reveals Anglo Leasing firm now wants Sh3b more” https://www.standardmedia.co.ke/article/2000122079/shock-as-
treasury-reveals-anglo-leasing-firm-now-wants-sh3b-more/?pageNo=3

23
The Eurobond Scandal
Less than a year after the election of President Uhuru Kenyatta in March 2013, Kenya went to
international money markets to issue Kenya’s first sovereign bond worth US$2.75 billion. This
was done in two tranches. The first issue raised US$2 billion (Kshs176 billion at the time) and the
second another US$815 million (Kshs 74 billion) for a total of US$2.8 billion (Kshs 250 billion). The
government said that the money would be used to reduce official borrowing from the domestic
market which would spur private investment by lowering interest rates. According to an analysis
by economist, David Ndii, the government executed two transactions from the offshore account
into which the US$2 billion had been credited. It paid off a pending loan of US$604 million (Kshs 53
billion) and then transferred US$394 million (Kshs 35 billion) to the exchequer. That left US$1.002
billion (Kshs 88 billion) in that account. The government has never accounted for this money.

When inconsistencies were pointed out the government responded with both lies and insults. The
lies were that up to Kshs 120 billion had been used partly to pay pending bills to road contractors
and partly for budget support. But as Ndii points out, the recurrent budget for the 2014/2015
financial year was funded by domestic revenues: the government raised Kshs 1.106 trillion in
revenues, of which Kshs 229 billion was transferred to the counties. That left Kshs 877 billion for
national government functions. The national government’s recurrent budget for that year was Kshs
897 billion, a mere Kshs 20 billion more than the revenue, reflecting no inflow of the Kshs 120
billion as claimed. According to this logic, the national government required only Ksh 20 billion
more than what it had earned through revenue, so there was no way it could have used the Ksh 88
billion from the bond.

In its first public statement on the matter, the Treasury promised to give information on the projects
that the Eurobond money had funded. It subsequently gave ministries three weeks to furnish the
relevant information. Five weeks later, in an interview with Business Daily, the CS Finance lamented
that, “The ministries cannot differentiate whether the money they have received from the Exchequer
came from VAT, income taxes, customs duties, excise taxes, domestic borrowing or the Eurobond”.
This is true but irrelevant to the issue. Treasury should have been able to provide the answer. As Ndii
points out, government has a monitoring and evaluation responsibility:

“For the Treasury to disburse a huge external loan, the biggest ever, without
expenditure tracking seems downright irresponsible”.85

In the coming months the government would ‘torture’ the figures to show that the missing
Eurobond money had indeed financed development projects. This was done by “wildly” (Ndii’s
word) inflating the cost of nine projects in the energy sector that showed overruns of nearly Kshs
67 billion. Rural electrification of primary schools was said to have cost Kshs 34 billion rather than
the Kshs 9.9 billion that had been budgeted. An unbudgeted item for the financial year, military
modernisation, gobbled up another Kshs 62.8 billion. The point of the cookery, Ndii surmised, was
to create a plausible storyline to explain the missing Eurobond money, “How high up does this fraud
go?” he asked.

85 See Eurobond Accountability: Facts, Figures, Questions by David Ndii, https://kptj.africog.org/eurobond-accountability-facts-figures-questions-by-david-ndii/

24
The government couldn’t – or rather wouldn’t – answer this question directly but its conduct in the
coming years had the guilty air of an adulterer caught in flagrante delicto.

By way of explanation, the Treasury posted seven memos on its website which ostensibly proved
it had transferred all the Eurobond receipts to the Exchequer. It ought to have been a simple case
of authenticating the deposits and then confirming the transfers as alleged by the Treasury. In the
meantime, the Treasury was actively trying to conceal that it had squandered the public’s money.
It uploaded a document purportedly showing that US$998 million (Kshs 101.1 billion) had been
remitted, but the bottom part of the document was redacted.

As David Ndii explained. the government’s real problem was that it could not account for
the Eurobond money that it never spent and still manage to balance its accounts.86 In the
2014/15 financial year, it partially pulled off this miracle by reducing domestic borrowing
for the year from Kshs 251 billion to Kshs 110 billion. The Kshs 140 billion reduction covered
the exact amount of Eurobond money that it claimed to have carried forward from 2013/14.
Unfortunately, this voodoo accounting was undone by the Central Bank accounts on domestic
borrowing and was flatly contradicted by the interest that the government reported to have
paid on domestic borrowings for the year.

In 2016 the Auditor General, Edward Ouko, tried to get to the bottom of the affair by conducting
a forensic audit of Eurobond transfers from the Federal Reserve Bank of New York. As part of
his preparations, he told Parliament that he had already made appointments with top US and UK
financial institutions involved in the transactions. Mr Ouko promised to send forensic auditors to
scrutinise transaction data at JP Morgan, Federal Reserve Bank, City Transaction Services New York,
JP Securities, Barclays Bank, ICB Standard Bank, Qatar National Bank and other banks that had
handled the US$2 billion Eurobond transactions.

Mr Kenyatta promptly blocked the investigation, arguing, implausibly, that by saying that “the
Eurobond money was stolen and stashed in the Federal Reserve Bank of New York”, Mr Ouko
was implying that the Kenyan government and United States had colluded. “Who is stupid here?”
the President scornfully asked. In the next few years, the government became cockier and more
belligerent. With the Auditor General not allowed to follow the international trail of the money, he
would be reduced to informing Parliament at the end of each audit year that, “Investigations into
the receipts, accounting and use of funds related to the Sovereign/Eurobond are still ongoing and
the accuracy of the net proceeds of Kshs 215,469,626,035.75 is yet to be ascertained”.

As Ndii’s analysis pointed out, unravelling this mystery should not have been as complicated as the
Auditor General’s laconic conclusion might suggest. On the deposit side of the Eurobond, there
should have been only two entries: one for the balance of the first issue in June 2014, US$2 billion,
and a second one for the US$ tap sales87, issued later in the year. The Treasury’s effort to explain
the mystery only compounded it, even with the IMF weighing in to support the official explanation.

86 See Daily Nation, 4 December 2015 https://www.nation.co.ke/oped/opinion/Eurobond-billions-a-curious-incident-of-a-dog-in-the-nighttime/440808-2984510-

dbpm9qz/index.html
87 See https://www.investopedia.com/terms/t/tap_issue.asp. The tap sale allowed the government to hold back part of the original bond and issue it over a period of time
at the original face value, maturity and coupon rate, but sold at the reigning market price. Theoretically, this would allow the government to issue bonds when market
conditions were favourable, and lock in agreements, redemption schedules and interest payment dates.

25
But as the Mozambique Eurobond story shows88, the IMF has been criminally negligent on these
matters.

In this case, the IMF’s attempt to aid the government was unavailing. The Fund showed that
Eurobond money was received and spent in the 2013/14 financial year. But given that the Eurobond
money was received in the last week of that financial year, it would not have been possible for it to
be spent in that year. There was no drawdown until the first week of July, which was the start of
the 2014/15 financial year. The difference between the Fund’s fiddling and the Treasury’s fiddling
was that the IMF reported a domestic borrowing figure of Kshs 251 billion for 2014/15 domestic
borrowing, whilst the Treasury showed one of Kshs 110 billion. As Ndii noted, “The IMF cooks the
books one way, and the Treasury, the other”.

But the Treasury’s lies were also compounded by the mandarins’ bad memory. By 2015/2016, they
seemed to have forgotten the 2014/2015 numbers. Now the Treasury reported Kshs 251 billion
as the correct domestic borrowing figure. With Kshs 251 billion confirmed as the correct amount,
the only way to account for the Eurobond Kshs 140 billion was to show the projects in which it was
invested. That no such projects have been named implies that at least US$1 billion of the Eurobond
money has disappeared into thin air. The conclusion that it has most likely been stolen by some very
senior untouchables is compelling.

With investigations never even started, the Auditor General beaten down by the President and
a marked lack of enthusiasm from the US end (especially by the New York Federal Reserve) it is
unlikely that we will know who stole nearly US$1billion of tax payers’ money.

Eurobond
billions:the
anatomy of a
grand heist
https://www.nation.co.ke/oped/Opinion/The-anatomy-of-a-grand-heist/-/440808/3035238/-/l1er9dz/-/index.html

David Ndii Economist and policy analyst.

88 See Cate Reid, Mozambique: The anatomy of corruption, Africa Report at http://www.theafricareport.com/Southern-Africa/mozambique-the-anatomy-of-corruption.
html

26
Part 3

THE MECHANICS OF CAPTURE

As Burbidge argues in The Shadow of Kenyan Democracy89, the point of the aggressive avarice of
Kenya’s corrupt leaders is to maintain power and privilege. That depends not just on effective
control of the Presidency and the Treasury, but also of the electoral process and law-enforcement
agencies – especially the police, the judiciary and the anti-corruption authorities, coupled with
the weakening of oversight institutions, especially Parliament, the Auditor General, civil society
and the media. Beginning with electoral capture, this chapter provides details of the mechanics,
that is, the tools and methods of capture.

1. Capture technique 1: ensure that the electoral management body is compromised

In the early 1990s, the primary method for controlling the electoral process was through use of
public order laws, such as banning public meetings, arresting and detaining regime opponents,
and control of the electoral management body (EMB) through the President’s power to appoint
commissioners. Once appointed, the commissioners were nominally independent, but were
almost immediately compromised by being allowed to draw illegal payments and allowances.
A 1996 analysis of the Controller and Auditor General’s Report for 1993/1994 by the Institute
of Economic Affairs showed that the chairman and commissioners of the Electoral Commission
of Kenya (ECK) had been paid Kshs 38,443,800 (equivalent to Kshs 375 million today) in sitting
allowances, subsistence allowances and accommodation. They were paid whether they were on
duty or not, even on public holidays. They had also been allowed to use privately registered cars
that had no work tickets.90

With interparty parliamentary group reforms of 1997, opposition parties could nominate
commissioners, expanding the composition of the Electoral Commission. In theory, this should
have made the ECK more independent. But there were two problems. First, the opposition, like
the party of government, appointed reliable political operatives in the expectation that they would
protect its interests in the commission. Second, once the commissioners were in place, they realised
they were independent of their appointing parties and that they had unlimited opportunities to
‘sell’ their discretion and judgment to the party of government. The result is that since 1997, the
diversion of funds and fraudulent spending at the electoral management body has ballooned, not
subsided.

89 Burbidge, D., (2015) The Shadow of Kenyan Democracy: Widespread Expectations of Widespread Corruption, Ashgate Publishing.
90 A work ticket is assigned to a government vehicle and has a vehicle number. All the movements are recorded including fuel intake, service etc. The officer allocated
the vehicle and the driver are recorded on the work ticket. Every trip must be authorised by the officer in charge of the vehicle. At the end of the month the work ticket
is audited by the officer in charge of transport and the internal auditor to ensure that all movements were work related and that no fraud has been committed. A special
police department can stop any government vehicle to ask if the movement is official. It is the only unit that can impound a GoK vehicle. There must be a security
reason for a GoK vehicle to have private number plates, but it still must retain the work ticket.

27
As an AfriCOG study91 shows, between 1991 and 2007, the ECK received Kshs 15.8 billion to run
elections. Of this amount, 1.9 billion was paid out to commissioners in irregular payments and
allowances, unaccountable vehicle hire, unsupported and wasteful expenditure, and imprests not
accounted for. Yet huge as those amounts are, they are nothing compared to the wastefulness of
the Interim Independent Electoral Commission (IIEC) and the Independent Electoral and Boundaries
Commission (IEBC) since 2008. If before 2007, corruption in the ECK entailed trimming and larding
expenditure heads within the budget, the period since has been characterised by open and rapacious
greed that is proportionately matched by a sharp deterioration in the quality of elections.

That claim can be shortly demonstrated. Following the post-election violence in 2008, the
Independent Review Commission (IREC), better known as the Kriegler Commission, recommended
wide ranging reforms to address what it described as institutionalised impunity. Yet, within
months of Kriegler’s recommendations, the successor to the ECK, the IIEC, had reverted to type,
getting embroiled in corruption on a scale that the ECK had not touched. From that moment on,
the EMB would not rely on illicit payments from the government. Commission staff would rig the
commission’s procurement processes to enrich themselves, knowing well that they would not be
prosecuted or called to account if they helped the party of government in the process.

In this first procurement scam, senior officials of the EMB were paid handsome kickbacks by Smith
and Ouzman, a UK-based security printer from whom they had contracted to buy electoral materials.
In a subsequent UK criminal trial for corruption, it emerged that the officials of the company had
paid up to £349,057 in bribes (over Kshs 45 million today), (christened as ‘chicken’ to IIEC officials
and commissioners)92 to secure the contract, which entailed printing materials for the by-elections
from the 2008 election and the 2010 referendum. In return for these payments, IIEC gave Ouzman
information on rival bids to allow the company to inflate printing costs. Many IIEC officials – including
the chair, Isaack Hassan – were lavishly entertained by Ouzman on visits to the UK.

But ‘chickengate’ was nothing compared to the wanton procurement corruption perpetrated by
the new electoral commission, the IEBC, in 2013. Every item bought for that election was bought
corruptly. The principal procurement, for the electronic voter identification devices (EVID), was so
tainted that the Public Procurement Administrative Review Board, (PPARB) would have cancelled
the contract were the election not so close. The Board was giving its decision in Avante International
Technology Inc. and 2 others v. The IEBC. The case had come before the Board on the main ground
that the IEBC had ignored professional advice and awarded a tender worth US$16,651,139.13
(Kshs 1,397,724,925.51) to Face Technologies, a South African company. To do this, the IEBC had
unlawfully revised an unresponsive bid by Face Technologies to make it legal.93 In the words of
the PPARB, the IEBC had been inexplicably “magnanimous in interpreting its tender documents” in
favour of Face Technologies. The IEBC had not only acted with impunity, it had from the very first
been “bent on awarding the [EVID] tender to Face Technologies”.94 In the Board’s view, the IEBC was,
“waving the card of public interest as its defence in the various breaches of the procurement law”.95

91 AfriCOG, (2009) Free for all?: Misuse of Public Funds at the Electoral Commission of Kenya https://africog.org/reports/free-for-all-misuse-of-public-funds-at-the-
electoral-commission-of-kenya/
92 The indictment also included corrupt payments made to officials in Ghana, Mauritania and Somaliland.
93 These actions are corruption offences under sections 45 and 46 of the Anti-Corruption and Economic Crimes Act.
94 See p. 43 of the Public Procurement Administrative Review Board, (PPARB) decision in Review No. 59/2012 of 19th November, No. 61/2012 of the 20th of
November and No. 62/2012 of the 21st of November 2012. http://www.ppoa.go.ke/images/downloads/arb-decisions/2012-decisions/DECISION%2059,61%20
&%2062-2012%20BETWEEN%20AVANTE%20INTERNATIONAL%20TECHNOLOGY%20INC,BIDVEST%20PAPERPLUS%20LTD(LITHOTECH%20
EXPORTS)%20AND%20SMARTMATIC%20INTERNATIONAL%20HOLDING.pdf
95 ibid p. 63.

28
Under different circumstances, the Board said it “would have [had] no hesitation [annulling] this
tender”. However, it would not do so here because that would “certainly jeopardize the holding of
the forthcoming general elections”. The IEBC’s conduct was so egregious, the Board recommended
that the “Director General of the Public Procurement Oversight Authority carry out investigations
pursuant to powers conferred by section 102 of the ACECA and take appropriate action”.96 A
special audit on the procurement of electronic voting devices for the 2013 general election by the
IEBC, ordered by Parliament would later prove that what the PPARB had found in the pre-election
litigation was just the tip of a monstrous iceberg. It turned out that all electronics purchased for the
2013 election had been misprocured.97

The audit found that biometric voter registration kits had also been bought irregularly: though
the Treasury had appropriated money for this procurement, the IEBC had inexplicably borrowed
commercially to buy the kits. This unusual method, which echoed some of the elements of the
Anglo Leasing scandal, meant that the tax payer would pay fees and interests that ought not to
have been paid. More illegalities were committed in procuring the results transmission system. The
system was never inspected on delivery, leaving its functionality doubtful on election day.

On receiving the audit report, the Public Accounts Committee was so outraged it recommended
an anti-corruption audit and criminal investigation of all IEBC commissioners, the committees,
and of the CEO James Oswago who, in addition, they said should not only be barred from holding
public office but also surcharged for paying out Kshs 258 million to Face Technologies without a
contract. None of these recommendations were implemented, although Oswago was replaced in
2015 by Ezra Chiloba. Most scandalous however, were the ‘hefty’ undisclosed amounts that the
IEBC commissioners were paid at the end of 2016 for ‘agreeing’ to retire early to pave way for
reforms ahead of the 2017 election. This sweetheart deal, put together by a bipartisan committee
of Parliament, signalled that impunity would be rewarded rather than punished. Though the sums
were not made public, the commissioners had argued that they were entitled to all their forward
pay if they were going to leave before the end of their terms.

That deal set the tone for the behaviour of the IEBC in 2017. Their attitude is already foreshadowed
by their response to the recommendation of the Ethics and Anti-Corruption Commission (EACC),
that the electoral management body bar 106 candidates – for governor, MPs, and members of
county assemblies – from contesting the August 8 elections as unfit to hold office.98 None of the
106 was barred and 60 percent of them were eventually elected.

96 ibid p. 64; Section 102 provides that:

(1) The Director-General may order an investigation of procurement proceedings for the purpose of determining whether there has been a breach of this Act, the
regulations or any directions of the Authority.
(2) An investigation shall be conducted by an investigator appointed for the purpose by the Director- General.
97 See https://oagkenya.go.ke/index.php/reports/cat_view/2-reports/72-special-audit-reports
98 See “Corrupt leaders were cleared by IEBC for polls, says EACC”, 22 September 2017, https://www.standardmedia.co.ke/article/2001255237/corrupt-leaders-were-
cleared-by-iebc-for-polls-says-eacc

29
No external audit has been done on the 2017 procurement, but an IEBC August 2018 internal audit
of the 2017 general election provides damning evidence of how corrupt the electoral processes
were. The internal audit reviewed 31 contracts, worth Kshs 6.2 billion, that the commission had
signed. The auditors feared that taxpayers had not got value for money in ten contracts worth
Kshs 4.6 billion. As with previous corrupt dealings at the IEBC, the culprits were CEO Ezra Chiloba
– suspended in 2018 – the directorates of finance, ICT, supply chain management and legal and
public affairs. The proposal to carry out the audit had generated serious internal conflict, forcing
the commission to send its CEO, Mr Chiloba, on compulsory leave to allow for “a comprehensive
audit of all major procurements relating to the 2017 general and fresh presidential elections”.
Shortly thereafter, three commissioners – Consolata Maina, Paul Kurgat and Margaret Mwachanya
– announced that they had no confidence in Mr Chebukati, the chair, and were resigning from the
IEBC. They would later rescind their resignations.

As with the Smith Ouzman and Face Technologies cases, the IEBC seemed hell-bent on contracting
particular firms. For example, the audit showed that the IEBC had awarded Safran Identity & Security
a Kshs 2.5 billion contract to supply election technology for the repeat presidential election of
26 October 2017 on a Kshs 423.6 million performance guarantee that had expired two months
earlier. At issue was not just the additional Kshs 2.5 billion contract that Safran Morpho got for the
October 26 rerun but also a further contract to reconfigure the 40,883 Kenya integrated elections
management system kits it had supplied for the August election. Not surprisingly, Safran made hay
while the irregularities sun shone. The IEBC paid Kshs 2.5 billion for the Safran system: two thirds
of what it had spent on the six elections involved in the August general election. Safran charged
the IEBC Kshs 443.8 million for election day support, nearly double the Kshs 242.5 million it had
paid for the same support in the general election. The internal audit concluded that a sum of Kshs
384.6 million that IEBC paid Safran for ‘programme and project management’ was unnecessary and
therefore wasteful.

As in the 2013 election, many aspects of technology acquisition were corrupt and highly irregular.
Airtel was contracted to supply 1,553 units of Thuraya IP SIMs loaded with data bundles for the
results transmission system in the areas without 3G and 4G network – 11,115 polling stations in
all. The company could only supply 1,000 by election day. The additional 553 units were supplied
after the election. Oracle Technology Systems (Kenya) Ltd provided database and security solutions
at Kshs 273.6 million without a signed contract. Scanad Kenya Ltd got the contract for the IEBC’s
‘strategic communication and integrated media campaign consultancy services’ even though its
price was more than twice the Kshs 350 million budget earmarked by the IEBC. Africa Neurotech
was contracted to install IEBC data centre equipment but at a cost of Kshs 249.3 million, an amount
almost double the IEBC budget of Kshs 130 million. The data centre equipment was not ready on
election day.

Further details about the extent of impunity in the IEBC come from the lawsuits filed against the
IEBC on procurement of electoral equipment and materials just before the elections. In early 2017,
the IEBC single sourced Safran Morpho to provide election equipment, the same controversial
French company with whom the IEBC had negotiated a tripartite agreement to buy biometric voter
registration (BVR) kits for the 2013 election. As in 2013, the IEBC argued that its single sourcing
decision was necessitated by the limited time left to comply with the election timetable, a problem
they said had been compounded by interminable litigation. Safran Morpho has a chequered history

30
and due diligence might have ruled them out. In the USA, its subsidiary has been accused of
misrepresenting the firm’s track record. In 2013, Safran was fined US$630,000 by a French court on
being found guilty of bribing public officials in Nigeria to win a Kshs 17 billion identity cards tender.

Summarising these experiences, the inevitable question is: why are electoral management bodies
in Kenya allowed to be this unaccountable? Who benefits from a criminalised EMB? The answer lies
in the ability of the EMBs to give ‘state capture’ formal legitimacy.

So long as the country goes through the formalities of an election that

international observers can say ‘broadly reflects’ the will of the people – whatever
that means – electoral management bodies can be rapacious in cannibalising their
budgets and the governments they help put in power can be trusted to look the
other way.

Box 1. Safran
The Safran group, a French company, specialises in propulsion and aerospace equipment. It has been
in existence since 2005, when the Société nationale d’étude et de construction de moteurs d’aviation
(SNECMA), specialising in aircraft and rocket engines, merged with the Société d’Applications Générales
de l’Électricité et de la Mécanique, SAGEM, a company known for security.

Morpho S.A.S. is a French company, specialising in security and identity solutions. Until 2016, Morpho
S.A.S. was a part of the Safran group. It was thus sometimes referred to as Safran Morpho.

In 2017, Morpho S.A.S. merged with Oberthur Technologies, a French digital security company. The
newly merged entity is now known as Idemia.

2013
After a lengthy procurement process, which was marked by violation of multiple rules and the
disbandment of the initial tender committee, the Government of Canada and the Government of
Kenya entered into an agreement through which a Canadian company known as Morpho Canada
would supply the BVR kits to Kenya. It is worth noting, however, that Morpho Canada was a fully
owned subsidiary of the French-registered Safran. Furthermore, Morpho Canada was not registered
until February 2013, one month before Election Day, and after the government to government
procurement process had already started. According to the Auditor General of Kenya, Morpho
Canada was registered to “conveniently create a link” between Canada and Safran.

Although the Memorandum of Understanding was between the governments of Canada and Kenya,
the Agreement for Sale and Purchase of Hardware and License for BVR kits was based on Safran
Morpho’s quotation of 58,551,240 Euros (later negotiated down to 56,209,189 Euros). Indeed,
despite the signed contracts between the governments of Canada and Kenya, it was the French-
registered company, Safran Morpho, that provided the kits. No contract existed between these
entities. Kenya’s Treasury then proceeded to pay 100 percent of the cost before delivery. At the
same time, the Treasury requested and received a commercial loan from the Standard Chartered
Bank of London, raising the cost of the BVR kits by 42 percent of the negotiated price. In the end, as
a result of alterations to specifications and a request for additional kits, the Government of Kenya
paid a total of 65,152,629 Euros for the kits. The Auditor General concluded that Kenya did not
receive value for money with regard to the BVR kits.

31
 2017
In 2017, Safran (now referred to as Safran Identity & Security) was back on the scene in Kenya. Yet
again, a competitive tender process for the acquisition of an integrated digital system – which would
provide both voter registration and voter identification in one kit – was replaced with single sourcing.
In fact, a decision to award the bid to Gemalto was terminated, but a report by the Public Accounts
Committee determined that the cancellation was not done in good faith. Upon cancellation, the
award was given to Safran for Ksh. 4.19 billion. It is noteworthy that Safran was chosen despite the
fact that it had been disqualified at the tender’s technical evaluation stage. Indeed, the requirements
were changed during single sourcing so that Safran would qualify. The evaluation report concluded
that “the outcome was already predetermined in favor of Safran Identity & Security.”

2018-Present
The Kenyan Ministry of Interior and Coordination contracted Idemia in 2018 to provide 31,500 mobile
biometric kits to be used in the country’s national population registration. After it was revealed that
Idemia had violated Kenyan law by failing to register in Kenya, however, the National Assembly voted
to recommend criminal investigations against the firm. After MPs recommended that the company
be banned from conducting business in Kenya, Idemia sued the National Assembly. The company is
arguing that it did not have a chance to defend itself.

Sources:
Kamau Muthoni. 8 May 2019. “Huduma namba firm moves to court to salvage its stay in Kenya.”
Standard Digital. Available at <https://www.standardmedia.co.ke/mobile/amp/article/2001324652/
huduma-namba-firm-sues-house>.

Idemia. 2019. Available at <https://www.idemia.com/>.

Office of the Auditor General. 6 June 2014. “Special Audit on Procurement of Electronic Voting
Devices for the 2013 General Election by the Independent Electoral and Boundaries Commission.”

The Public Accounts Committee. February 2019. “Report on the Examination of the Report of
the Auditor-General on the Financial Statements for the Independent Electoral and Boundaries
Commission.”

Safran. 2019. Available at <https://www.safran-group.com/>.

2. Capture technique 2: undermine law enforcement

Effective law enforcement institutions – especially an effective and honest police service, a
functional, independent and accountable judiciary, and a professionalised office of the DPP – are
central to fighting corruption effectively. None of these institutions is wholly accountable or fully
functional. The police remain unreconstructed and if the evidence revealed by the police vetting is
anything to go by, also unrepentant.

32
The office of the DPP has been haphazard in prosecuting, picking out cases for prosecution for
reasons that appear patently political, fumbling cases involving the big fish and generally being
assiduous on those that involve small fry. (See Box 2 The DPP sets ‘springes to catch woodcocks’99.)
The judiciary seemed on the mend after 2010 when the new constitution came into force but since
then things have gone awry: judicial vetting failed to fully root out corruption; bribery has crept
back and though some of the excesses of the past have not returned, a clean judiciary is still a long
way off. On the whole, the government attitude to law enforcement is consistent with the logic of
state capture: control and compromise the police and the DPP and weaken the harder-to-control
judiciary.

a. The police: a vertically organised criminal syndicate

It serves state capture if politicians are wilfully blind to police corruption. In Kenya, police ‘palm
greasing’ at traffic stops is so routinised that drivers arrive with the bribe already folded, ready
to be slapped onto the palm, or slipped into the pocket of the traffic cop. Presidents are often
excused from the predations of the police but their responsibility and that of their government was
explained many years ago by William of Pagula: “For, one who permits anything to take place that
he is able to impede, even though he has not done it himself, has virtually done the act if he allows
it.”100

“Recruitment into the police

service is a grant of a
preloaded cash machine”

99 From Hamlet (by W. Shakespeare) Act 1, Sc 3. This phrase draws attention to both perpetrator and victim. It suggests that the traps (springes) are easily laid and victims
are easily caught.
100 Chayes, S., (2015) Thieves of State: Why Corruption Threatens Global Security at p. 41. New York. W. W. Norton & Company

33
The stability of state capture rests on uniting the interests and fates of low level operatives with
those of their bosses. In the case of the police, recruitment into the police service is a grant of
a preloaded cash machine. This, in part, is what the recent police vetting revealed. The vetting
proved that what Sarah Chayes observed of Afghanistan is true of Kenya: namely, the conventional
wisdom that corruption involves doling patronage downwards to juniors is wrong-headed. Instead,
it is subordinate officials who pay off the top in return for “unfettered permission to extract
resources for personal gain, and second, protection from repercussions”.101 The critical point is that
this whole system depends on “faithful discharge, by senior officials, of their duty to protect their
subordinates” and this implicit contract holds, “much as it does within the mafia, no matter how
inconsequential the subordinate might be.”102

Box 2. The DPP ‘sets springes to catch woodcocks’: only small fry get caught
by anti-corruption traps
The Pattni and Anglo Leasing cases would set the trend for the future: successful corruption would
invariably involve small fish. The big fish have rarely been successfully prosecuted. Many cases
are terminated by a nolle-prosequi by the DPP, another lot by acquittal – principally for shoddy
investigations – and a third lot by discharge under the criminal procedure code. And when indictment
and conviction have happened, as they did in the case of Ketan Somaia, former owner of Dolphin
Group of Companies and Margaret Gachara, former director of the National Aids Control Council,
the sentences have been either lenient at the point of conviction or dramatically reduced on appeal.

Gachara, for example, was convicted for abuse of office and corruption involving the loss of Kshs 21
million. She was sentenced to serve only one year in prison in August 2004 but was released on a
presidential pardon on 11 December 2004, with barely a quarter of the sentence served. Somaia’s
charge related to a 1990 scandal in which he scammed the government out of Kshs 112 million to
import 500 new ‘London-look’ taxis but delivered, instead, only 200 second-hand ones. He, too, was
convicted in 2004 and sentenced to two years in Kamiti Maximum Security Prison. For most of the
time that he was incarcerated, he was, in fact, ensconced in the private wing of Kenyatta National
Hospital. His conviction was quashed on appeal in April 2005. Another ill-fated high-level indictment
involved Dr Mohammed Abdi Isahakia, former director general of the National Museums of Kenya.
He was charged with theft but the case collapsed on the grounds that three crucial witnesses had
left the country and Dr Isahakia had offered to refund the stolen money, Kshs 2.5 million.

The mechanics of police corruption can be seen in the police vetting exercise undertaken by the
National Police Service Commission (NPSC). As one newspaper account sarcastically noted, top cops
often seem hard pressed to cite a major crime burst but the state of their bank accounts showed
them to be men of great business acumen, a fact that alone would “put Kiganjo Police Training
College at par with the region’s top business schools in producing entrepreneurs of note”.103

Officers’ bank records show deposits of hundreds of thousands of shillings monthly, mostly
from “businesses” that they and their spouses own or from ‘convenient’ sales of assets that they
previously owned. Many were jacks of all trades, running businesses that run the gamut from

101 ibid p. 60.

102 ibid p. 60.
103 See Daily Nation 13 April 2015 https://www.nation.co.ke/news/Police-Vetting-Wealth-Businesses/-/1056/2684994/-/qci7fu/-/index.html

34
chicken farming, residential and commercial rentals and fish farming. A sampling of their evidence
before the vetting panels is revealing: an OCPD (Officer Commanding Police Division) from Nandi
Central had an official salary of Kshs 26,000 as of 2011, but he transacted up to Kshs 440,000 in
a day through his M-Pesa account. According to him, these monies came from his wife and from
‘consultation’ duties. For example, on 19 April 2014, he made two deposits, an initial Kshs 159,000
and an additional deposit of Kshs 500,000 later in the day. On 12 January 2012 he had made two
large deposits, a first of Kshs 50,000 and a later one of Kshs 150,000. This money, he explained,
came from his engineering consultancy .104

Another officer had a hard time explaining where annual deposits of Kshs 3 million above his official
salary came from, recording the source as “miscellaneous and rental payments”. The bank records
of a former director of police reforms showed regular deposits of millions of shillings from junior
police officers, among them world athletics champion David Rudisha, who had twice sent him Kshs
900,000. He explained that Rudisha had given him this money to buy “some farm inputs”. Most
alarming, though he had once led police reforms, he showed little idea of basic police reforms105.

A former deputy police spokesman bragged that he was “the only solution to the problems affecting
the police service” and asked the vetting team not to be surprised by the ‘millions’ in his account.
In a short while, there would be millions more in that account: he was just about to harvest 20,000
tilapia from his fish farm and would sell these at Kshs 250 apiece. He told the vetting panel that he
had honed his money-making skills from unspecified “agricultural practices” after he realised that
he could not live on his salary alone. Nonetheless, he agreed with the vetting panel that “corruption
was rampant in the police service” but that this merely reflected “the situation in the society”106.

That no major shakeup of the police has followed from this much publicised vetting shows that this
high profile ‘stagecraft’ was cynical ‘busywork’, both Chayes’ words, to manage the expectations of
a disillusioned public. So in the end, a hapless public finds itself caught between an abusive police
service acting as accomplices of a predatory government.107

The police vetting process eventually petered out and left as much confusion as the interest it
had piqued. Many of these entrepreneurial officers are still in the police force, still pursuing their
sprawling business interests. Who benefits from a compromised and corrupt police force?

b. Counter-corruption commissions are a waste of public money

Since President Kenyatta announced his new anti-corruption drive, the EACC has become busy.
However, neither the EACC nor its forerunners have demonstrated the will to fight corruption.
Since it was established under the 2010 Constitution, the EACC has never exercised the authority
that Kenyans would like to see it exercise. Some of the problems of its ineffectiveness have to do
with its own sloppiness: poor investigations, underhand investigatory methods and a penchant for
the dramatic gesture. That has also been compounded by lack of political support and internecine
conflict between the commission and the office of the DPP.

104 ibid
105 ibid
106 ibid
107 cf. Christopher Gitari, (2017), Failure to Reform: A Critique of Police Vetting in Kenya, International Centre for Transitional Justice (ICTJ)

35
The combination of its own weaknesses and lack of political support means that the EACC is rarely
taken seriously. As already noted, its 2017 recommendation that the IEBC bar 106 candidates
was ignored.108 The EACC blames the IEBC for this debacle; in truth both commissions have been
ineffectual and are often implicated in corruption themselves.

The EACC’s conduct of the criminal investigation against the former PS in the Ministry of Foreign
Affairs, Mwangi Thuita, and Allan Mburu, former First Counsellor in Japan, for corruption in the
purchase of the Chancery and ambassador’s residence for Kenya’s Embassy in Japan is a typical
example of the commission’s methods. At issue was nearly Kshs 2 billion of tax payer money.
The prosecution was terminated on a ‘no case to answer’ basis, meaning that the EACC had not put
before the court sufficient material to put the accused on their defence. The DPP’s appeal against
that decision was thrown out as spurious. The usual response of the authorities is to say that such
decisions from the court are proof of judicial connivance with the corrupt. In fact, it is easy, at
least in this case, to show that the real problem here was a combination of shoddy investigation
and vindictiveness. The EACC had constructed its entire case on three primary, but weak sources:
an investigation by the Parliamentary Committee on Defence, Security and Foreign Affairs; two
investigatory trips to Japan by the EACC and a special audit by the Auditor General. It turned out,
on the strength of a letter written to the Speaker, that the chair of the parliamentary committee,
Adnan Keynan, had tried to extort money both from the Minister for Foreign Affairs and from Thuita
Mwangi to slow down the investigation. The indictment detailed the corrupt actions as: a) conspiracy
to commit an offence of corruption, namely breach of trust by approving the purchase of the Tokyo
property; b) use of office to confer a benefit on the embassy’s former landlord, Nobuo Kuriyama, from
whom land and buildings were bought; and c) failing to follow the procurement law and regulations.

The case collapsed because the EACC had ignored exculpatory evidence: the accused were said
to have personally benefited from the purchase of the embassy but there were no bank records
to prove that presumption; the accused were said to have bypassed the ambassador to Japan but
there were photos showing the ambassador as team leader when the purchase decision was made;
the EACC could not explain why it had treated the variation between the price the property was
bought at and the government’s own valuation of the property as proof of corruption, given the
many other instances – both in Kenya and abroad – in which property valuation and market price
had differed. The investigation and prosecution relied on the initial queries raised in a special audit
by the Auditor General but ignored the fact that many of the auditor’s queries had been answered
by the ministry. In summary, then, both the DPP and the EACC ignored everything that a diligent
and professional investigator and prosecutor would have done to ensure a solid prosecution.

The modus operandi of the EACC, which has damaged its credibility and undermined its ability to
lead the fight against corruption, is to launch an investigation in the glare of publicity and make
sweeping claims that it is generally unable to substantiate later. The question is why successive
anti-corruption commissions have been allowed to continue operating in this manner? The answer
is another question: who benefits from this?

108 See “Corrupt Leaders were Cleared by IEBC for polls, says EACC” 22 September 2017 The Standard at https://www.standardmedia.co.ke/article/2001255237/
corrupt-leaders-were-cleared-by-iebc-for-polls-says-eacc

36
c. The judiciary: partly reformed and easy to sway for state capture purposes
As the ill-fated indictment and prosecution of Kamlesh Pattni in the 1990s shows, the judiciary was
a central pillar of the repressive and corrupt dispensation replaced by the 2010 Constitution. By
the year 2000 the judiciary was universally condemned as both corrupt and incompetent. The few
honest judges faced myriad problems: lack of research support; poor record keeping occasioned
by insufficient stenographers and electronic recording devices; a huge and ever-growing backlog
of cases; biased and politicised allotment of benefits to judicial officers, especially cars and houses;
and highly politicised appointments and promotions awarded by the President, nominally with the
advice of the Judicial Service Commission, but in practice, at his own plenary discretion.

When the Constitution of Kenya Review Commission (CKRC) travelled the country to collect
public views on constitutional change, it faced a real dilemma: what to do with incumbent
judges and magistrates. Many people said that they should all be sent home. The CKRC judges
discovered that judges and magistrates had “been appointed for the wrong reasons” and many
had “demonstrated neither competence nor integrity”.109 The commission balked at wholesale
‘dismissal’ only because it worried that this might be seen “as a grave interference with the
independence of the judiciary”. Judicial reform would eventually evolve in two steps: the first
phase, 2003-2006 would be the measures implemented by the Kibaki administration under radical
surgery, and the second would be judicial vetting under the 2010 Constitution. Both would prove
not merely inconclusive but also too poorly designed to achieve deep reforms. The result would
be a partly reformed judiciary that combines in equal measure many judges of the old corrupt
order and new, more professional judges committed to the ethos of the new constitution.

Shortly after the election of President Kibaki in 2003, the then Minister for Justice, Kiraitu Murungi,
launched what he termed “radical surgery”, a high profile process of identifying and purging corrupt
judges and magistrates from the judiciary. Soon it proved neither radical nor surgical. The reforms
were based on an investigation carried out by a former law partner of the justice minister. Thus,
congenitally politicised, radical forgery failed to mollify critics who saw it as a Kibaki plot to remove
President Moi’s judicial cronies to make room for friends of the new government, rather than a
root and branch reform of Kenya’s decrepit judiciary. This criticism was overdone. However, the
minister’s best intentions had no base in statute and without this, radical surgery merely mortgaged
judicial reforms to the factionalism that was then tearing apart the ruling coalition.

Eventually, over 80 magistrates and 23 judges were removed for corruption reasons, but by 2006 “not
a single judge had been found guilty”110 by any of the many tribunals established to investigate them.
In one particularly notorious case involving Justice Philip Waki, who would later lead the investigation
into the 2007 post-election violence, the tribunal was scathing about the methods that Justice Aaron
Ringera had used: unsafe reliance on clearly unreliable witnesses; failure to talk to the affected judge
and overlooking innocent explanations related to the claims made. More important, radical surgery
left many judges in place who would be later dismissed as unfit to hold office.

109 See The Final Report of the Constitution of Kenya Review Commission (2005) http://www.katibainstitute.org/Archives/images/CKRC%20Final%20Report.pdf
110 See AfriCOG, Shattered Dreams: An Audit of the Kibaki Government’s Anti-Corruption Drive, 2003-2007.

37
The result of this unsatisfactory purge was that there was still much left to do when the 2010
Constitution came into force. The constitution adopted a two-pronged approach to dealing with
corruption and judicial failure. The first was institutional design and the second was vetting of
incumbent judges and magistrates. The institutional reforms reorganised the powers, functions
and composition of the judiciary, strengthened the Judicial Service Commission and created a
Supreme Court. Vetting was based on Article 23 of the Sixth Schedule to the Constitution and the
Vetting of Judges and Magistrates Act. The aim of vetting was to clean up the judiciary, partly to
create a more accountable judiciary by removing the bad apples and partly to provide a transitional
justice111 mechanism that would eliminate institutionalised impunity.

Once it got going the vetting proved – as the police vetting was to prove subsequently – both
ineffectual and inadequate; ineffectual, because the vetting mechanism was congenitally defective
in that some aspects of vetting were challenged in court and heard by judges who were themselves
yet to be vetted. It is not surprising then, that the effect of this litigation was to constrict the wide
mandate and discretion initially given to the Vetting Board. In addition, the timetable for vetting
was hopelessly optimistic and had to be extended now and again through amendments to the law.
The resulting legislative delays slowly punctured the Vetting Board as did preparations for 2013
elections.

In theory, the vetting helped to clean out the courts, but it is hard to know what to make of statistics.
It seems like there were, in effect, two vetting processes: vetting as understood by the Vetting
Board and vetting as interpreted by the courts. One out of every three first instance decisions made
by the Board were reversed on review in both the High Court and the Court of Appeal. Among the
magistrates about 45 percent of the Vetting Board’s first instance decisions of unsuitability were
reversed.

The Vetting Board would eventually run into more serious problems: the ‘scope of vetting’ was
dramatically reduced by the Supreme Court. The Supreme Court said that the Board could only
vet judges and magistrates for conduct that occurred between the date of appointment up to 27
August 2010, the day the 2010 Constitution came into force. This had far-reaching consequences.
All the decisions of the Vetting Board that found judges and magistrates unsuitable on account of
things they had done before they were appointed, or for things done after 27 August 2010, were
effectively nullified.

The board was now obliged to send all cases related to the post-August 2010 period to the Judicial
Service Commission (JSC). In the end, judicial vetting never met its twin objectives of cleaning up
the judiciary and fully restoring public confidence in the courts. That it failed to clean the judiciary
explains the persistence of judicial corruption and resistance to reforms, periodically explained as
“cartels fighting back”. That vetting failed to restore public confidence explains why the judiciary
gets lukewarm public support when it is dismissed by politicians as ‘activist’ or ‘captured by NGO or
opposition interests’. Why did vetting fail to achieve its purposes?
111 Pablo de Greiff, Research Unit Director for the ICTJ writes,
‘At the highest level of generality, “transitional justice” refers to measures that are implemented in order to redress the legacies of massive serious crimes under
international law. Despite abiding disagreements about the outside boundaries of the concept of transitional justice, consensus has been achieved about a set of core
elements that a transitional justice policy minimally must include. These elements include prosecutions, truth-telling measures, reparations for victims, and some
initiatives tending towards institutional reform, particularly the vetting of security sector personnel. Other elements frequently said to be parts of transitional justice
include memorialization efforts as well as local justice initiatives. As in all attempts to translate concepts into practice, there are also plenty of debates about the best
way to implement these measures.’
“Transitional Justice, Security and Development: Security and Justice Thematic Paper,” World Development Report, 2011, World Bank.

38
To begin with, the vetting law was too restrictive. In retrospect, the law could have been more
robust than it was. One of its main weaknesses was that it gave no immunity to people who had ever
given or been asked for a bribe by a judicial officer. Without immunity from prosecution, witnesses
had effectively been denied the means with which to prove corruption. That left the board with the
unenviable task of inferring bribery from unexplained deposits in the judges and magistrates’ bank
accounts, very much like the police vetting.

Secondly, the Supreme Court’s formalistic reading of the vetting act to limit the relevant period
of acts committed between appointment and the constitution’s effective date, undermined the
broad purpose of the statute that was to remove undesirable individuals from the judiciary. That
decision created more problems than it solved. One, it allowed judicial officers known to be unfit
to continue in office, which was itself a serious blow to public confidence. Two, it introduced
unnecessary unevenness – some might even say discrimination – into the vetting process for those
who had already been vetted. Three, it saddled the JSC with what in effect were vetting decisions,
thereby mixing transitional justice issues – which is what vetting was – with the core mandate of the
JSC, which was more prospective.

39
Thirdly, it was wrong in principle that many aspects of the vetting process were litigated before
judges who had not themselves been vetted; that is, before judges with a personal stake in limiting
how deep and wide vetting went. This eroded public confidence in the integrity of the vetting
although the reason for vetting in the first place was the need to restore public confidence in the
courts.

Fourthly, the Vetting Board compounded its own problems. The board decided that in deference
to the seniority of judges of the Court of Appeal it would sit en banc, that is, as a full bench rather
in small panels of three, when it came to scrutiny of the judges of that court. The result was bizarre:
the board would sit in one capacity to vet a particular judge and if that judge was dissatisfied with
that decision, he would then seek a review, which would be heard by the same full panel of the
Vetting Board. Some lawyers saw no problem in this, arguing that the board’s review power was
no different from the power of an apex court to review its own decisions. Yet again, the question
was whether the public would appreciate what seems on the face of it to be a rather otiose legal
argument and whether this did anything for the public’s confidence in vetting.

The result of these partial measures is that the judiciary retains many of its old bad ways, which
are now being used to undermine ‘the good guys’. The government now blames corruption in the
judiciary as the greatest barrier to anti-corruption reforms, which leaves out of account the shoddy,
compromised investigations often carried out by the police, and the ineffectual and often politically
targeted prosecutions by the State Law Office. As with the other aspects of law enforcement, the
question is, who benefits when the courts are perceived as compromised or untrustworthy?

Kenya’s anti-corruption efforts - motion without movement

Given this analysis, it is clear that the latest anti-corruption efforts can be summarised as the ‘tried,
tested and known-to-be ineffective’ approaches of the past. This means that although it is good to
have an energetic prosecutor in office and that the EACC has bestirred itself, this won’t be enough.
The lynchpin of the government’s approach to fighting corruption is, like the Goldenberg scandal,
high profile arrests followed by quick indictments. Some people are impressed that some big
names have already been scalped: former Sports Cabinet Secretary, Hassan Wario, former Principal
Secretaries Lillian Omollo, Richard Ekai and Richard Lesiyampe, present and former Kenya Power
bosses Ken Tarus and Ben Chumo, Kenya Railways boss, Atanas Maina, chairman of the National
Land Commission, Mohammed Swazuri and senior managers at the National Cereals and Produce
Board. These arrests have generated much excitement, but the detailed history elsewhere in this
study suggests the excitement is premature. Kenya’s prosecution-driven anti-corruption strategy
has always been rather benign; it is ‘capture-mark-release’, a little like ecological methods of
estimating the population in an ecosystem. It is never meant to harm the corrupt.

Much the same thing must be said about the much-vaunted effort to repatriate cash from abroad.
This too seems like a good strategy but many of the people who should return money are in fact
part of state capture. This is why past efforts to repatriate money have failed. In 2003, the Kenya
Government hired Kroll Inc. to trace monies squirrelled abroad by the previous regime. Their final
report, the Kroll Investigative Report on Wealth of President Moi and Associates, located US$1
billion in looted taxpayer money hidden overseas. At the time, Andre Pienaar, the African office
of Kroll Inc., warned that tracing the money was easier than retrieving it, as other countries had
painfully discovered.

40
Pienaar pointed out that the DRC could not find any of the money supposedly stolen by Mobutu
Seso Seko, the dictator after his overthrow in 1997. Nigeria has not to date recovered monies
stolen by Sani Abacha, who died in 1998. In the Philippines, 20 years after Marcos was deposed, the
Presidential Commission on Good Government, (PCGG) had recovered only US$3.7 billion, believed
to be less than half of what Marcos had stashed abroad – that is after the PCGG had worked for
nearly three decades, employing 94 lawyers, researchers and administrators at an annual budget
of US$2.2 million.

It is not surprising that since the Kroll Report none of the traced money has been repatriated. Excerpts
of the suppressed Kroll audit published by the open information website, Wikileaks, indicated that
money from Kenya had been spirited to Liechtenstein, Liberia, Luxembourg, Australia, Belgium,
Brunei, Canada, Finland, Germany, Grand Cayman, Somalia, South Africa, Israel, Italy, Japan, Jersey,
Malawi, Namibia, the Netherlands, Puerto Rico, Russia, Sudan, Switzerland, the UAE, Uganda, the
United Kingdom, the United States and Zaire.

A second effort was initiated in 2010, with the British High Commissioner to Kenya, Rob Macaire,
promising that his government would seize and repatriate funds hidden in Britain. It did not happen.

Some of the problems in recovering money arise from international banking laws, the complexity
of the money laundering industry, the expense of recovery efforts and inter-jurisdictional legal
conflicts. The problem now, as in 2003, is lack of a strategy and an inability to slay sacred cows. The
Achilles heel of the post-Kroll investigation was an incoherent legal strategy: would it be seizure
(which would rely on co-operation and mutual legal assistance) or negotiated voluntary repatriation?
The moral force of the repatriation argument was also blunted by the government’s confused tip-
toeing around President Moi’s money. The authorities were too solicitous of the former President
and seemed too keen not to embarrass him, even though he enjoyed no legal immunity, then as
now. How far will the authorities be prepared to go if they come across the Kenyatta family money
out there?

Just how ineffectual recovery efforts have been can be gleaned from the fact that the problem
has got worse. The US National Bureau of Economic Research, estimates that Kenyans hold up to
Kshs 5 trillion (US$50 billion) in offshore accounts. In 2003, Kroll had estimated US$1 billion dollars.
The money stashed away now would be US$36.4 in 2003 prices. If the Kroll figures were all that
was then stashed abroad, the stealing has gone up 36.4 times (3,640%) in 15 years112. This time
round, voluntary repatriation has been tried and failed: the National Treasury Cabinet Secretary
Henry Rotich announced an amnesty that expired in July 2018, but no one took it. That suggests
that those affected either don’t think they will be pursued or else they believe that they can fight
off any attempted repatriation.Part 3 concludes on a grim note: if state capture is this entrenched
are there realistic reform options that can in anyway undermine it? That is what Part 4 of this study
lays out.

112 The Panama Papers identified 27 offshore entities linked to Kenyan interests registered in various tax havens. See https://offshoreleaks.icij.org/
search?utf8=%E2%9C%93&q=&c=KEN&j=&e=&commit=Search

41
Part 4

UNDOING STATE CAPTURE:

IN SEARCH OF OPPORTUNITIES
AND REFORMS
Democracy retains potential to undo state capture
Though this study has argued that corruption has escalated with the evolution of democracy in
Kenya, it is not meant to suggest that democracy be abandoned altogether. Deep democratisation
is a long-term project, a story, as Michael Walzer points out in a different context, “of victories and
defeats”113. Sometimes it involves two steps forward, one step back, and even one step forward and
two steps back.

The post-1990 democratic reforms were seen in linear, progressive terms: autocratic rupture,
followed by gradual transition, culminating in long-term consolidation. Practice has frustrated that
expectation and democratisation processes have generally not been linear:114 some countries that
transitioned to democracy in the post-1990 period “have ended up ‘getting stuck’ in transition,
or reverting to more or less authoritarian forms of rule”.115 These ‘ambiguous democracies’ have
put considerable empirical pressure on the early theory and generated a new lexicon: “choice-
less democracies”116, which are ‘societies that can vote but they cannot choose’; or descriptors
like “hybrid regimes” meaning states that “combine rhetorical acceptance of liberal democracy,
the existence of some formal democratic institutions and respect for a limited sphere of civil and
political liberties with essentially illiberal or even authoritarian traits”.117

Some of the literature speculates that such regimes occupy a “precarious middle ground between
outright authoritarianism and fully-fledged democracy”.118 These speculations are undermined by
two home-truths: they rest on the inarticulate premise of a linear progression from non-democracy
to democracy (no matter what their protestations) and, most important, they assume that what
they term the middle ground is unstable. In fact, state capture theory assumes just the opposite;
namely, that once the state has been captured it is possible for a transition to abort halfway to
democracy and acquire a stable, sub-optimal equilibrium with the façade of democracy, but not
its substance. This unconsummated democracy has three possible effects, all of which erode the
ability of democratic institutions to undo state capture.

113 Lecture by Michael Walzer, Freedom and Equality: Can the Two Stand Together? given at The Paris Institute of Political Studies on March 15, 2018, https://www.
youtube.com/watch?v=xZc2YOQG_0A
114 Menocal, A R and Fritz, V with Rakner, L (2008) Hybrid regimes and the challenges of deepening and sustaining democracy in developing countries, in The South
African Journal of International Affairs Vol. 15, No. 1, June 2008, pp. 29-40
115 ibid p. 30
116 Mkandawire, Thandika, (1997) Crisis management and the making of ‘choiceless democracies’ in Africa, in Richard Joseph (ed.), The State, Conflict and Democracy
in Africa. Lynne Rienner Publishers, Boulder, US.
117 Menocal, A R and Fritz, V ibid p. 30 quoting from Ottaway M, Democracy Challenged: The Rise of Semi-Authoritarianism. Washington, DC: Carnegie Endowment
for International Peace, 2003.
118 ibid p. 31

42
First, it can lead to a precipitous loss of faith in democratic politics, even though this may,
paradoxically, go hand in hand with sustained or growing interest in politics.119 Reybrouck argues
that this union of “distrust and passion” is deeply destabilising: He asks, “ … how much derision
can a system endure, especially now that everyone can share their deeply felt opinions online?”120
Taking this view, though early 21st century societies are honeycombed with disputes and conflicts,
democratic institutions have ossified and lost the resilience needed to mediate these disputes and
conflicts, principally because democracy itself no longer commands popular trust.

Second, ‘bogus democracies’ have frustrated moderates and pro-democracy forces and driven them
away from the democratic game. This widespread apathy at the centre of politics usually leads to
a lurch to political extremes, of both right and left. The underlying source of frustrations at the
centre is uncertain, but seems to be in two parts. First, the problem of ‘mandate ambiguity’– that is,
clarity about what politicians and bureaucrats are accountable for – as increasingly people have felt
that democracies are not serving their interests, even as government ever more loudly proclaims
the public interest. Second, more people have no levers of action to put failed politicians right.
That is to say, citizens everywhere feel unmanned by either of two factors: a lack of instruments
to sanction errant politicians and bureaucrats; or the prohibitive cost of using the instruments that
are, at least theoretically, available to them.121

Third, frustration with ‘ambiguous democracy’ can spawn a reform ‘fantasy’ that says what the
country needs is “a simple, brutal stroke that wins the day”122 and annihilates state capture. Such
fantasies can drive politics towards populisms of both the left and the right, either of which, once
embraced, can make deep democracy impossible. The central insight is that since the struggle never
ends, the continuing existence of political freedom is a necessary condition for the gradual success
of the struggle.

In short, there are grave dangers in abandoning democracy for some left or right wing
authoritarianism, or of allowing the country to be exploited by the state capture elite under their
own variant of ‘ambiguous democracy’. Even with these caveats, democracy can still undo state
capture. How might this happen?

1. How might the democratic process undo state capture?

Rupture
Rupture occurs when a corrupt and illegitimate regime experiences a terminal crisis that causes
it to collapse suddenly and unexpectedly. Most cases of rupture are driven by an economic
crisis. Since all regimes rely on some form of bargain with key actors in society, the inability of
the state to provide material benefits to key constituencies can lead to such constituencies
recalculating the costs of their support and defecting. In the Philippines in 1986 the military
recalculated their support for Marcos, defected to Acquino and brought the Marcos regime down.

119 David van Reybrouck, Why Elections are Bad for Democracy, The Guardian 29 June 2016 at https://www.theguardian.com/politics/2016/jun/29/why-elections-are-
bad-for-democracy
120 ibid
121 See a good early discussion of this in Fearon, J. (1999) Electoral Accountability and the Control of Politicians: Selecting Good Types versus Sanctioning Poor
Performance in A. Przeworski, S. Stokes, & B. Manin (eds) Democracy, Accountability, and Representation, Cambridge Studies in the Theory of Democracy, (pp.
55-97) Cambridge, Cambridge University Press.
122 Ibid, Michael Walzer,

43
An economic crisis in Peru, spurred in part by the 1979 oil crisis, strengthened the opposition and
undermined the “commitment of the military to the original authoritarian bargain”.123 There are
two potential causes of rupture in Kenya: the inevitable economic crisis likely to be spawned by
binge borrowing and the general immiseration that predation must invariably cause, even if Kenya
binge borrowing does not lead to crisis.

The political impact of a debt-fuelled economic crisis

The transition crises in Latin America in the early 1980s, especially in Uruguay and Brazil, show
how quickly rupture can occur. The recessions of the early 1980s sharpened intra-elite conflicts;
food protests and riots broke out; labour became more militant; the business community split and
moderate elements in the military defected from the regime, reigning in their hardline colleagues.
The immediate impact of a debt-fuelled crisis is the rapid shrinking of ‘state capture benefits’,
which is likely to provoke intra-elite conflict and convert some of the capture elite into pro-integrity
‘reformers’. An economic crisis would also mobilise the ‘losers’ in the wider public – workers, students,
and the urban poor. The combined effect of an impoverished, desperate public and a critical mass
of discontented elites ready to play the democratic game (imperfect though democracy is) would
undermine the ability of any one faction to tightly control the electoral process and, most critically,
could lead to splits within the security forces, thus cancelling out repression as an option.

Corruption becomes unsustainable by undermining long-term investment and growth

Even without a debt-fuelled crisis, it seems inevitable that the cannibalistic nature of state capture
corruption must eventually run its course. State capture creates a climate hostile to productive
activity. It does so by creating a ‘permissive ethos’ up and down the system. Superiors ‘steal and
cheat’ and a ‘me-too’ attitude cascades down the ranks of lesser officials with the same result. In
this environment, everyone knows that so much of the anti-corruption work targeting ‘procurement
officers’ and wealth declarations for civil servant is really whistling in the wind. Corruption “starts
at the top and slowly works its way through the arteries, veins and finally flows inexorably into
the capillaries of lower level individual action”.124 There is in such a set-up no part of government
committed to growth. It is inevitable that this must eventually shrink the benefits available for
capture. As in a debt-fuelled crisis, this too would immiserate the public, radicalise politics and
create opportunities for a new set of leaders that might include ‘clean-up’ autocrats and populists
of either the left or the right.

A powerful anti-corruption coalition overwhelms the capture elite

The excitement that President Kenyatta’s ‘show arrests and indictments’ have generated indicate
a groundswell of support for, and viability of an anti-corruption platform. The question is whether
such an anti-corruption coalition can emerge even with the electoral process captured (as described
in the analysis in this study). It is unlikely, but experience shows that the politically unexpected
happens more often than people believe: eastern Europe after 1989; Tunisia and Egypt after 2011;
and Ethiopia in 2018. All show the various ways in which politics can surprise.

123 Haggard, S and Kaufman, R R (1995) The Political Economy of Democratic Transitions p. 69. Princeton University Press, New Jersey, USA.
124 Rotberg, R. I., (2017) The Corruption Cure: How Citizens and Leaders Can Combat Graft at p. 195. Princeton, NJ: Princeton University Press p. 223

44
Attempts to control the 2022 elections could go wrong and a rank anti-corruption outsider could
be propelled to the Presidency. Most Kenyans increasingly recognise – as they did not before – that
civil society organisations (CSOs) have been right: “anti-corruption reforms rarely succeed without
exemplary efforts on the part of state leaderships and political elites”;125 that is, anti-corruption
reform has to start from the top. That means that the expectations that have characterised past
elections – that a population can elect corrupt leaders and then hold them to account once they are
in office – are delusional.

The cases of Singapore and Hong Kong both show that the support of the top leadership is
critical. Singapore is especially instructive. The country was hopelessly corrupt under the British:
“police doubled up as thieves and opium smugglers”126 and ran extortion and protection rackets
“collaboratively with Chinese gangs”.127 British efforts to rein in corruption had achieved little by
the time Singapore became independent in 1960. The Prime Minister, Lee Kuan Yew enacted a
strict Prevention of Corruption Act, professionalised the police services and revamped the Corrupt
Practices Investigation Bureau (CPIB), a moribund pre-independence agency. Lee then ramped up
the CPIB’s budget from 1 million Singapore dollars in 1979 to 34 million in 2011. By 2003, Rotberg
reports, “the CIPB was completing 99 percent of its investigations, usually within 90 days of
inception”.128 More important, “prosecutions took place in 85 percent of those”.129

Hong Kong confirms what the experience of Singapore shows: corruption can be reduced by leaders
determined to eradicate it and who create the right institutions to fight it. From its founding in
1842, Hong Kong leaders had fed off a long-tradition of unaccountable gifts and tributes; illicit
gambling was rife; illegal brothels flourished and the police was, as in Singapore, in on every racket
that this criminal city was famed for, until the early 1970s. In 1974, Governor Sir Murray MacLehose
established the Independent Commission Against Corruption, which reported directly to him. It
had sufficient financial resources to hire experienced investigators and wide powers to enforce
internal integrity without political interference. In the 30 years between 1982 to 2011 its budget
grew almost sevenfold, from US$14 million to US$105 million. It also recorded early success by
bringing down, in 1976/1977, a “drug-smuggling gang that had been paying police US$10,000 per
day for protection”.130

Using the opportunities created by the 2010 Constitution

The creation of 47 county governments under the 2010 Constitution, has, as the theory of federalism
predicted, created 47 opportunities for a governance experiment. The initial experience has been
disappointing because many counties imbibed the bad practices of the national government.
Nonetheless, counties are beginning to recognise that they and the national government are not
similarly situated, because: one, counties are geographically closer to the people and therefore bad
practices are easily ‘observable’; and two, some county governors are committed to doing the ‘right
thing’ and generating peer pressure encouraging others to emulate them.

125 ibid The discussion of the experiences of Singapore and Hong-Kong draws from Rotberg’s book.
126 ibid p. 111
127 ibid, p.111
128 ibid p. 112
129 Ibid p. 112
130 This account is summarised from Rotberg R. I., (2017) The Corruption Cure: How Citizens and Leaders Can Combat Graft, pp 115-117. Princeton, NJ: Princeton
University Press

45
Makueni County has designed Kenya’s most robust public participation system and demonstrated
that participatory budgeting is possible and that citizens know their own priorities. Nyeri has
instituted a human resource system that eliminates ‘ghost workers’. Laikipia has created an Economic
Development Board modelled on development boards in Singapore and Rwanda, to drive ‘Brand
Laikipia’. Makueni has created a universal health care system and Laikipia is working hard to do the
same.

These initiatives will generate further reforms. Effective service delivery demands strong disclosure
systems through open, accessible and reliable data. Counties that want to do the right thing will
leverage existing technological platforms, such as e-citizen, to support these governance initiatives,
remembering of course, that the abuse of the integrated financial management system in the
National Youth Service131 and the cash-gate heist in Malawi132 are cautionary tales on the limits of
technology. If a critical mass of counties implements such reforms, the ‘demonstrative effect’ would
undermine the hold that the state capture elite has on the national government at the centre.

2. What civil society organisations can do

Developing case files and archiving records for future prosecutions
The rapidly growing number of removals, impeachments, arrests, prosecutions and convictions of
former heads of state for corruption whilst in office since the mid-1990s has begun to change how
people understand accountability for wrongdoing. Leaders who frustrate efforts to hold them to
account in power do not necessarily escape liability if their wrongdoing is kept alive and the evidence
preserved for future action.133 In South Africa (2018), South Korea (2018), Peru (2009), Brazil (2018),
Venezuela (1993), Ecuador (1997), Indonesia (2001), Lithuania (2004), Paraguay (2012), Israel (2000)
and Guatemala (2015), former heads of state have been forced out of office or are serving long jail
terms for things they did when there were in office (see Box 3).

131 https://www.standardmedia.co.ke/article/2001282311/how-ifmis-works-and-why-it-has-claimed-many-suspects
132 https://www.economist.com/baobab/2014/02/27/the-32m-heist
133 See “List of world leaders ousted from office in recent history” in the New Straits Times of March 10, 2017 at
https://www.nst.com.my/news/2017/03/219330/list-world-leaders-ousted-office-recent-history

46
Box 3. World leaders forced out of office for actions while in office
Brazil: In Brazil corruption and financial mismanagement has been the quickest way to lose
power or to end up in prison. On December 29, 1992 Fernando Collor de Mello resigned from the
presidency before impeachment proceedings for corruption began before the Senate. Twenty-four
years later, on 31 August 2016 President Dilma Rousseff was impeached by the Senate for illegally
manipulating the national budget. On 12 July 2017, the popular Luiz Inácio Lula da Silva, Dilma
Roussef’s predecessor, was convicted of money laundering and corruption and sentenced to nine
years in prison. On 24 January 2018, an appellate court increased Lula’s sentence to 12 years. A
further appeal to the Supreme Court was dismissed on 5 April 2018. Lula declared for president from
prison but was disqualified under Brazil’s Clean Slate Law in spite of a plea by the United Nations
Human Rights Committee that the government allows him to exercise his political rights.

Ecuador: In 1997 and then again in 2005, Ecuador dismissed its presidents. The first, Abdala
Bucaram, was dismissed on 6 February 1997 for “physical and mental incapacity” barely six months
after his inauguration. The mental incapacity was never proven but Bucaram had presided over an
exceptionally corrupt administration. In April 2005, President Lucio Gutiérrez was dismissed for
packing the Ecuadorian Supreme Court with his associates. His administration collapsed when the
armed forces publicly stated that they had withdrawn support for Gutiérrez.

Guatemala: On 1 September 2015, the Guatemala Parliament stripped President Otto Perez of
his immunity following accusations that he was part of a corrupt ring that took bribes to allow
companies to import goods without paying import taxes. Faced with impeachment he resigned from
the presidency. His vice president Roxana Baldetti had also resigned in May 2015 for her involvement
in the same scam.

Indonesia: On 23 July 2001 President Abdurrahman Wahid was dismissed from office for
incompetence and corruption.

Israel: In July 2000, Israel’s president, Ezer Weizman resigned following allegations of tax fraud
and a corruption scandal and thus avoided inevitable impeachment. In June 2007, President Moshe
Katsav resigned following serious abuse of office charges, arising from accusations of rape and
sexual misconduct. In 2011 he was jailed for seven years but was freed in December, 2016.

Lithuania: On 6 April 2004, the President of Lithuania, Rolandas Paksas was impeached for granting
Lithuanian citizenship to a Russian businessman in exchange for a payout. He was subsequently
banned from standing for office in Lithuania.

Paraguay: On 22 June 2012, Paraguayan President Fernando Lugo was impeached by congress
in controversial circumstances after a crisis in which conflict over some land led to the death of
17 people. Lugo conceded the removal after the Supreme Court declared his impeachment
constitutional.

Peru: On 21 Nov 2000 Alberto Fujimori resigned as President of Peru by fax from Tokyo. Congress
rejected his resignation; voted to remove him from office; banned him from holding public office
for 10 years and had him extradited to stand trial for ordering the massacres of civilians and for
corruption. He was jailed for 25 years in 2009. On 24 December 2017, Fujimori was pardoned by
President Pedro Pablo Kuczynski on health grounds but that pardon was quashed by the Peruvian
Supreme Court in October 2018.

47
 South Korea: In October 2018, the former President of South Korea Lee Myung-bak was sentenced
to 15 years in prison and fined 13bn (£8.8m) for bribery and embezzlement. His successor Park Geun-
hye had earlier been forced from office amid protests before being jailed for 33 years for corruption.
The court found that she had taken bribes from some of South Korea’s largest companies, including
Samsung. In 2009 former President Roh Moo-hyun – president from 2003 to 2008 – committed
suicide after being questioned by prosecutors for alleged corruption. Earlier in 1996 two former
presidents, Chun Doo-hwan 1980-88, and Roh Tae-woo 1988-1993, were also convicted of bribery
but were pardoned in 1997.

Venezuela: In May 1993, the then President of Venezuela, Carlos Andres Perez was accused of
embezzlement and illegal enrichment. He was suspended in May 1993 and then dismissed by the
Congress on 31 August 1993.

These cases show that there is scope to hold corrupt leaders to account, especially after they leave
office. The danger is that if leaders know that the evil that they do will follow them into retirement,
they will be tempted to destroy evidence when they are in power.

CSOs can forestall efforts to destroy evidence by trawling through official records such as audit
reports, human rights investigations, and international monitoring mechanisms, and preserving
them for the future. But gathering and archiving evidence needs to go hand in hand with traditional
advocacy to keep the issues alive for future action. Though such advocacy and the sustained scrutiny
required to make it effective won’t necessarily get rid of corruption, it can eventually raise public
outrage and shame politicians, especially those not fully embedded in state capture networks, to
align themselves with CSOs.

As the Ipaidabribe movement in India134 shows, sustained advocacy can resonate with the public
and can “eventually morph into establishment of an effective political [movement]”.135 In the end,
although, “neither Prime Minister Manmohan Singh nor his ruling All-India Congress Party really
favoured” the deep reforms the Ipaidabribe movement wanted, the Indian parliament did pass a
law with some of the movement’s proposals. The movement’s real impact may lie in its ability to
keep the public mobilised on corruption, which is precisely the charged climate that would make it
possible for governments to take action in the future.

Advocacy in pursuit of foreign indictments and convictions

When domestic institutions are captured, the chances of effective action ever being taken
are vanishingly small. However, changing sentiment in the West, driven in part by a desire to
interdict funding for terrorism, has led to heightened scrutiny of illicit transfers, whatever their
provenance. This presents an opportunity for international advocacy, leading to indictment
and conviction of the corrupt, especially, those at the centre of capture. It is instructive that
most of what we now know about the Eurobond scandal in Mozambique has become public
principally because federal prosecutors in the USA have indicted the key players in that scandal.

134 https://theconversation.com/i-paid-a-bribe-how-some-citizens-are-fighting-corruption-from-the-bottom-up-48111
135 Rotberg R. I., (2017) at p. 273.

48
Examples from Kenya show that this is a viable route towards prosecution of the corrupt. The arrest
and extradition to the USA of Baktash and Ibrahim Akasha, sons of drug dealer Ibrahim Akasha, show
that pressure eventually works, especially in cases where criminals and the corrupt have used the
American financial system to make money transfers. The Akasha brothers are now facing long jail
terms. This is the culmination of many years of fruitless diplomatic pressure and Kenya government
pussy-footing.

Box 4. The unravelling of impunity

From the days before their father was gunned down in the red light district of Amsterdam, the
Akasha family plied their brutal trade with impunity. They were powerful, ruthless and untouchable:
they were hooked to the police, the judiciary, and local and national politicians, which afforded
them blanket immunity to make Mombasa an international drug supply hub, linking producers in
Asia with markets in Africa and the American east coast. They were eventually outed, according to
newspaper accounts136, by one Vijaygiri Anandgiri Goswami, an accomplice from India, also wanted
by the US federal agencies for narcotics charges. Ironically, when he was in Kenya, Goswami had
police protection at a time when the Inspector General of Police, Joseph Boinnet, had just withdrawn
security from governors from the opposition.
Like the Akashas, Goswami lived untroubled by Kenya’s security forces while he was in the country.
Goswami had in fact come to Kenya, (by his own account) on an investor’s visa, presumably because
he was going to set up a cement factory in Kilifi. Instead, what he set up, with the connivance of the
authorities, was a drug-manufacturing factory. That they have been finally brought down and that
they have named a long list of other Kenyan drug barons and their official protectors as part of an
extensive plea deal with USA agents, shows how quickly impunity can unravel.

Like the Akashas, Ketan Somaia, former owner of Marshalls (EA) Ltd and the Dolphin Group, was
another high roller whose illicit tentacles reached deep inside government and gave him carte
blanche to live a life of impunity. (See Box 5.). He was jailed for eight years by an English court in
2014 following a private prosecution by a former friend whom he had swindled.137

In a similar vein, the former managing director of Kenya Power and Lighting Company (KPLC) Samuel
Gichuru, and Chris Okemo, (then finance minister) have been battling, since 2011, an extradition
treaty to Jersey on a charge of money laundering. Both Gichuru and Okemo own a slew of foreign
companies that once supplied KPLC in consideration of paying “consultancy fees” that Mr Gichuru’s
Windward Trading Company subsequently split with Mr Okemo. The Jersey court has already seized
over half a billion Kenya shillings from Windward Trading and the company’s directors pleaded
guilty to money laundering charges years back. The intestinal transactions leading to the Jersey
extradition request were provoked, in part by Mr Gichuru’s divorce proceedings, when his former
wife blew the whistle on some of his vast wealth concealed in Jersey; and from a British House
of Commons enquiry about a Kshs 8.2 billion loan that was meant to fund a KPLC study on the
feasibility of the proposed Ewaso Nyiro hydro power plant, but part of which was diverted through
an intricate web of foreign companies to Mr Gichuru’s Jersey account as a kickback138.

136 See Daniel Wesangula, Enemy within who may have sent Akasha empire tumbling, The East African Standard, 11 November 2018 at https://www.standardmedia.
co.ke/article/2001302269/man-who-downed-akashas
137 See The Standard, 24 July 2014, “Fall of high-flying billionaire Ketan Somaia as UK hands eight-year jail term for fraud” at https://www.standardmedia.co.ke/
article/2000129256/fall-of-high-flying-billionaire-ketan-somaia-as-uk-hands-eight-year-jail-term-for-fraud
138 See Daily Nation, 17 July 2018, “How ex-Kenya Power boss Gichuru ‘escaped’ with millions” at https://www.nation.co.ke/news/How-Gichuru-got-away-with-
Kenya-Power-big--loot-/1056-4666764-158ttgez/index.html

49
 Box 5. Ketan Somaia’s ‘Smash and Grab’ ends in a UK Jail
In 2014, Ketan Somaia was found guilty of fraud and jailed for eight years by a London court following
a private prosecution, the largest ever, by former friend Murli Mirchandani. Mr Mirchandani had lost
nearly US$20 million to Somaia between 1999 and 2000. He and Dilip Shah, who had lost US$200,000,
are just two in a long list of investors who had been scammed over the years, by a seemingly immune
Somaia, on different continents. A small-time timber merchant in Kenya, he parleyed his way to global
prominence by working political connections with President Daniel arap Moi in Kenya and then the UK
Conservative party machinery, through Mark Thatcher, son of Margaret Thatcher, former British Prime
Minister. According to the Kroll investigation, Somaia started out as a minion of Hezekiah Oyugi, then
permanent secretary in the Office of the President. He also partnered Kamlesh Pattni of Goldenberg,
whom he later famously accused of defrauding him of his investment in the Dolphin Group (a
conglomerate invested in hotels, banks, casinos, a sugar miller and a motor dealership). His high-living,
and a penchant for reckless ventures, soon got him in trouble. His Dubai-based group collapsed in 2001
and bankruptcy proceedings and private debt claims plagued him across three continents. According
to Kenya’s Business Daily, the following are among the many frauds in which he has been named.
1. In mid-1990s he scammed Kshs 375 million (about US$8 million then) from the Government of
Kenya, to supply the police with communications equipment, but failed to deliver. He then fled to
London and refused multiple times to appear before Parliament to answer questions.
2. Still in the 1990s, he obtained Kshs 238 million (about US$5 million at 1995 rates) in a deal with
American hotel chain Starwood Hotels & Resorts Worldwide. Though Starwood sued and obtained
a judgment against Somaia in London they failed to enforce the judgment against the Somaia-
controlled Block hotel properties in Kenya.1
3. In 1997 he received US$2 million (then Kshs130 million) from his Dubai-based business partner
Surajit Sen and failed to repay the money.
4. In 1999/2000 he obtained nearly US$20 million (Kshs 1.5 billion) from Murli Mirchandani, a friend,
and US$200,000 from Dilip Shah, an in-law, as loans or investments. Somaia purported to sell
stakes in Delphis Bank (Mauritius and Tanzania), the Diamond Mining Corporation of Liberia and
other ventures. He paid Mr Shah a paltry US$7,000 through his accountant years later. In 2014 he
was convicted on nine counts of fraud by a British court and jailed for eight years after a private
prosecution begun by Mr Mirchandani in 2011.2
5. In 2001 he obtained US$15 million (then about Kshs1.2 billion) from a businessman referred to
as ‘Mr Bose’ in a court document, but subsequently only paid back US$2 million by handing over a
house in Dubai that had been pledged as security.
6. In 2002 he was arrested by Hertfordshire police in the UK for scamming a local entrepreneur of
UK£500,000 (then US$768,000). Released on bail, he fled to Kenya, which refused to extradite him
back to the UK. He was arrested in India in 2008, extradited to the UK but released because the
money had been repaid.
NB: Save for the lenient 2004 conviction and sentence–now quashed–Somaia has not paid for his
crimes and frauds in Kenya.
1 Daily Nation 4 June 2002: https://www.nation.co.ke/news/1056-314228-lbmh5kz/index.html
2 https://www.casemine.com/judgement/uk/5b2897fb2c94e06b9e19e985

What these cases show is that it is possible, but difficult, to get high level Kenyan crooks and
criminals who have used the international financial system, extradited to stand trial abroad when
Kenyan authorities won’t act.

50
Leveraging proposed constitutional reforms
There is one last opportunity that offers some slim chance for reform: the now seemingly inevitable
political reforms that both Mr Kenyatta and Mr Odinga seem keen to push through, via constitutional
amendment. It is unlikely that constitutional reforms designed to support ‘political deals’ between
politicians offer a real opportunity for overhaul of constitutional institutions. Nonetheless, if anti-
corruption forces mobilise they can get seemingly innocuous reforms – especially on public finance
management, reporting and budgeting for social economic rights, independent funding for the
judiciary, the Office of the Auditor General and the Office of the Controller of Budget – which in the
short-run may not raise the political temperature, but could prove extremely useful further down
the road.

Uhuru-Raila to change
the Consitution?

51
CONCLUSION
This study has not struck an optimistic note. Too much of past anti-corruption advice has been
unrealistic and based on ‘reform fantasies’ for which there is neither interest nor will. The result
of investing time and energy pursuing that which was clearly impossible from the very first policy
advocacy, has made unattainable that which was initially achievable. This study suggests that there
is a need to rest the advocacy approach along the lines suggested here. Reform opportunities often
come serendipitously and check list approaches, in which CSOs mark off one reform after another,
as they are implemented, have proven inappropriate in Kenya’s setting. Some of the proposals
made here are about seizing opportunities when they emerge and others – such as those around
international indictments and evidence gathering for future prosecutions – require that CSOs have
forensic skills and forge partnerships with auditors and others that have the data relevant for such
prosecutions. It is not an easy road but it is a realistic one.

52
AfriCOG would like to thank Wachira Maina for his work on this report.

Thanks are also due to the following who offered peer review and editorial support:
David Ndii, Seema Shah, and John Githongo.

Thanks to Waiyaki Otieno for comments on design and layout.

We are grateful to our team members for their dedication:

Njonjo Mue, Lucy Ngumo, Joshua Nzola, Naomi Shako, Dominic Mutuku, Eunice Mwende,
Fred Lusasi, Eunice Ombasa, Duncan Ochieng and Sam Kamande.

AfriCOG thanks its donors for their support, which makes our work possible:
The Open Society Initiative for Eastern Africa (OSIEA), the Royal Netherlands Embassy,
Amnesty International, TrustAfrica.

The opinions contained in this report are those of AfriCOG alone.

Nairobi, May 2019

Africa Centre for Open Governance (AfriCOG)

P.O. Box 18157 - 00100 GPO
Nairobi, Kenya
www.africog.org