1
Kito Kidogo in Kajiado, Kenya
A descriptive case-study about the self-containing corruption in the
public transport sector in Kenya
By
Amber van de Klippe
A thesis written for the Master of Science (MSc) in Political Science: Political Economy
Student number: 11260564
Supervisor: dhr. dr. Seiki Tanaka
Second reader: dhr. dr. Michael Onyebuchi Eze
23-06-2017
2
Doing this research gave me the opportunity to spend six weeks in the amazing country of Kenya. I have met so many beautiful people who were inspiring because of their strength and positivism (and a bit because of their complete craziness). Many came out of their comfort zone to talk to me, some even felt that they put themselves in danger, just because they wanted to help me. Most of them will never read this, but I would like to thank you all so much.
Despite that this experience was wonderful and incredibly educational it was also confronting and challenging. Without my dear friends and family to support me in my pessimistic
moments this thesis would not have been here. Thanks mum, Thala, Emma and my
Hangmatjes for sharing the dark side of Africa with me, and thanks Saul for putting this all in perspective.
My personal goal for this thesis was to gain more knowledge about corruption so I could understand the complex matter better. I have partially succeeded. I definitely gained more knowledge about corruption but I am not sure that I understand the concept better, as during the research process even more questions were raised. This journey has not ended for me yet.
Although much research about its causes and consequences has been done, the complex nature of corruption is still far from being unravelled. I believe that a better understanding of what corruption exactly is will enhance the current anti-corruption strategies and I sincerely hope that my thesis can contribute to that.
3 Content Introduction ... 4 Literature ... 9 Theoretical Framework ... 13 Theory ... 13 Relevance ... 14 Methodology ... 17 Case selection ... 17 Methods ... 19 Results ... 21
Empirical evidence - interviews ... 21
Empirical evidence – surveys ... 26
Analysis of the evidence ... 30
Conclusion ... 34
4
Introduction
Corruption is a major problem in Africa. It hampers economic growth because it, among other reasons, hinders investments (Justusen 2014). It lowers the legitimacy and pressing powers of the government, for the reasons that the trust of the citizens in the politicians is low and that the law can easily be bypassed (Aluko 2002). It deepens inequality and creates instability and insecurity, because the rich and powerful people benefit more from corruption, which puts the poor in an insecure position (Hope 2014, Obala 2014).
Despite numerous efforts, up until now, no fully effective anti-corruption strategies have been developed for African countries. There are some success stories in other countries, such as Hong Kong’s Independent Commission Against Corruption (ICAC). This is often mentioned as a poster child for effective national anticorruption institution. It is characterized by its threefold strategy: investigation, prevention, and education (Huberts 2000), but implementing similar strategies in countries in Africa has not yet been fruitful.
To understand why it is necessary to investigate a strategy for the African context, it is important to take one step back. Besides the contextual and cultural differences between Asia and Africa that make the Hong Kong strategy not directly applicable to Africa (Huberts 2000), the conceptualisation of corruption has to be examined. The strategies based on the current conceptualisation of the corruption problem in Africa are not effective because of systematic differences in corruption and therefore we need to rethink our approach. According to the current literature corruption can be incidental, systematic or systemic. This will be elaborated later on, but essentially it means that corruption can occur occasionally and mostly with low-ranked officials (incidental), in several or almost all layers of society (systematic) or that it has become the norm instead of the exception (systemic).
In most African countries with a high level of corruption, it has a systemic character. In a limited number of cases it is also referred to as institutionalised corruption: ‘when corruption becomes institutionalised in a society, it infiltrates into the value-system, it becomes a norm, part and parcel of culture and subsequently goes into the realm of behaviour’ (Aluko 2002, p. 395). The characteristics of systemic (or institutionalised) corruption are so different from corruption that is just incidental or systemical that it also needs another approach when one aims to decrease it. To develop effective anti-corruption strategies for African countries, with corruption that is so systemic that is has become the norm, a better understanding of this type of corruption is necessary.
5
However alternatives are rising, corruption is still most often described within the theoretical framework of the ‘principle-agent problem’. This theory is based on the assumption that corruption is a system involving two unequal actors in which the high-ranked is exploiting the low-ranked one. The decision whether to be corrupt or not is expected to be made based on an individual cost-benefit analysis. However, within a the few years a new narrative is being presented, arguing to consider corruption as more of a collective action problem, instead of a principle-agent problem (Persson 2013). The core assumption of this theory is that the decision of being corrupt or not is not made based on individual considerations, but on group behaviour; everybody is corrupt because they expect everybody else to be corrupt as well.
It is not to be argued that this collective action framework has to replace the principle-agent framework completely. When incidental or systematic corruption is present, this framework is suitable. It is however not suitable to explain all the corrupt behaviour, especially not when the corruption has become systemic. The collective action theoretical conceptualisation of corruption is still in its infancy, but scholars argue that this becomes more applicable as the corruption in a country is more systemic than incidental (Persson 2013). For this reason this collective action approach will form the basis for the descriptive analysis of this research.
This systemic (or institutionalised) characterization refers to corruption on a country level, it defines the corruption system of a country as a whole. However the different corrupt institutions in a country are closely intertwined, not all the institutions develop their corruption in the same way. The variation is based on the legal framework, the human resources practices, information and communications technology (ICT) and the engagement of non-government actors (Mills 2012). When a country is considered to be systemic corrupt, all the institutions in the public administration are considered to be systemic corrupt as well.
However, describing these institutions as systemic might not be comprehensive enough. The most common definition for systemic corruption in the current literature is that corruption has become the norm instead of the exception (Arellano-Gault 2016; Johnston 1998). However in countries like Kenya, where corruption has been present in a systemic way for so long, this definition does not cover the full complexity of the corruption present in some of its institutions.
When corruption becomes normalized the choice of whether to be corrupt or not slowly merges from an individual one into a collective one, as argued in the collective action theory. The theory is that as a result of this the individuals performing the individual corrupt
6
actions become less important (Persson, 2013) and the system in which they operate can become self-containing.
To test whether self-containing corruption does exist, a case-study is conducted. When looking for self-containing containing corruption the focus is one the most corrupt institution, the police, in one of the most systemic corrupt countries, Kenya.
Kenya is ranked 145 out of 176 in the Corruption Perception Index by Transparency International in 2016. New Zealand and Denmark both received the 1st rank, as being the least corrupt countries, while Somalia is ranked 176 (Transparency International 2016). Many countries, both local to Kenya and further afield, can boost our understanding of corruption in political cultures. However, when compared with neighbours with similar (or even higher) corruption levels, such as (South-) Sudan, the DRC and Burundi, Kenya enjoys a higher level of security. Together with the fact that most people speak English, as this is taught in primary school, Kenya makes a safer and more practical research case.
Kenya also scores the highest score compared with other African countries on bribery according to the research conducted by Afrobarometer: 49% of the respondents has at least once paid a bribe (Peiffer and Rose 2014). Corruption is also still increasing in the country. In a press release in 2015, Afrobarometer (2015) stated that 64% of the Kenyans experience an increase of corruption in the past year. Their trust in the government is low, 70% of the Kenyans perceive the government as handling the fight against corruption ‘fairly badly’ or ‘very badly’.
Because this research is looking for high levels of corruption, the focus will be on the most corrupt institution within the country; the police. Not only in Kenya is the police described as the most corrupt institution, in the whole of Africa it is seen as highly vulnerable for corruption (Kivoi 2013, Peiffer and Rose 2014). To study its systemic character, we will look into the one of the sectors that is affected by the corrupt police the most; the public transport sector. The public transport sector in Kenya is almost fully carried out by ‘matatus’. These are small taxi busses, with a(n official) passenger capacity between 8 and 16 passengers which only leave when they are full.
Essentially, this research will focus on exploring this system of corruption between the matatu business and the police. The first question that has to be asked is: does self-containing corruption exist in Kenya? And if so, the research question will be: how to describe this self-containing corruption and its characteristics properly?
7
The detailed operationalisation of corruption will be elaborated in the concept section, however two elements are important to understand at this point. The first is that corruption is an overarching term; it includes several actions and actors. Especially in a country like Kenya where corruption is very endemic; it is present in all layers of society. This research narrows down to one of the most determinative actions: bribery. However corruption can include other actions - like embezzlement, fraud, extortion and blackmailing – in this research corruption is operationalised as bribery because this is the most common corrupt action. Bribery is often called ‘kito kidogo’ in Kenya, which is Kiswahili for ‘something small’.
Secondly the level of corruption that this research is focussing on. Corruption in the institutions in the public administration is defined as “petty”, administrative or low-level corruption, affecting small groups of people or individuals with small amounts of money (Mills 2012). This petty corruption is opposed to grand or high-level corruption, which is corruption involving high-level politicians and large sums of money (Ackerman 1997).
To answer the research question, several aspects have to be taken into account. These aspects form the basis characteristics for describing self-containing corruption. The first one is one is the extent to which the choice to be corrupt is an individual one. When corruption is incidental or systematical the choice whether to be corrupt or not depends on an personal cost-benefit calculation. When corruption becomes more systemic this individual calculation changes into a collective calculation, based on the likeliness that others will be corrupt. In addition to this the pressure to be corrupt when you have the change will be higher when corruption becomes more normalized, thereby decreasing the individual choice option.
The second aspect is the extent to which the corruption plays an organisational part in daily life. When occurring just sometimes, corruption disrupts the system of law. But when corruption is very systemic, the corruption often replaces certain elements of the organisational systems in society. The continuity of the corruption becomes therefore a necessity for the functioning of the current system (Arellano-Gault 2016). However, when corruption becomes self-containing I expect that it will not only be part of the organisational structure of Kenya, but that there has been established a structure to guarantee the continuation of the corruption.
The third element is the perception of the people themselves. To describe self-containing corruption properly, the local perception on the phenomenon is crucial for understanding. Based on these three aspects the sub-questions are:
1. How individual is the choice to be corrupt?
8
3. What is the perspective of the respondents on corruption?
The answers from the sub-questions will form the descriptions of the three most important characteristics of self-containing corruption. To answer the main- and sub questions, a combination of quantitative and qualitative methods will be used. The obtained data will be analysed on a descriptive basis. This will be elaborated on in the methods section.
In sum, I argue that when systemic corruption at country level exists for a long time a self-containing corruption system can evolve in one or more of the sectors in the country. One can speak of containing corruption when it is not only systemic (accepted as the norm) but also structured to such an extent that it has its own rules and guidelines. As a result, the choice whether to be corrupt or not can no longer be conceived of as solely an individual one. Additionally, systems are established to maintain the corruption, thus becoming embedded in the system and potentially indispensable for the functioning of the daily life. When corruption is truly self-containing, citizens can no longer imagine that sector or even the whole society without corruption; I expect to find these factors being all present in the corruption between the police and the matatu-business in Kenya.
This thesis contains the following. It starts with a literature review, providing the necessary background information and discussing the current findings about corruption. It summarizes the most important findings considering corruption of the last decades; its negative impact on development, the anti-corruption strategies and the conceptualisation of corruption. This literature review also demonstrates where the current literature fails to completely describe corruption; the present approach is not comprehensive enough. The second part, the theoretical framework, first reveals the theory of this thesis and further consists of the conceptualisation of the key terms, the research objectives and the relevance of this research. This is followed by the methodology section which includes the research approach, the case-selection and the methodology. After that the results will be discussed, followed by the analysis containing the answers on the sub-questions, ending with the conclusion.
9
Literature
In the West, corruption is inherently seen as unacceptable and dreadful. On the other hand it can also be considered to be an alternative way to pay government officials; a way of taxing (Olken and Pande 2012). The crucial conclusion that can be drawn here is that the concept of corruption, a payment practice, is not in itself necessarily bad but the consequences can be. These consequences are extensively debated in the literature. This will be summarized here shortly because it is not the main focus of this research, however it is still important to understand the context.
On macro level corruption affects the economy and the government. Corruption obstructs investments and subsequently economic growth (Justusen 2014, Mauro 1995). Besides this, corruption decreases the ability of a government to correct externalities. Externalities are the (unintentional) negative side-effects of an activity, for example a truck with cargo which is causing damage to the road surface. Within a democracy, the government can impose taxes and hand out fines to (de-) stimulate certain behaviour. But when people have the opportunity to undermine this system by paying corrupt officials, the government loses its ability to control its citizens (Olken 2009).
Hope (2014) expands this argument by stating that corruption corrodes the democratic state, by undermining its powers. The legitimacy of the government is affected and ‘democratic values such as trust and tolerance got tossed away and are replaced by ethnic violence’ (p. 501). Obala (2014) also argues that corruption is an important foundation for the emergence of conflicts. He uses the example of the unfair use of land rights, land grabbing. His study showed that corruption creates conflicts not only between individuals, but also between a landlord and its tenant because of this.
A striking example of the disturbance of government actions is given by Olken (2005) in his research about missing rice within government projects to support the poor. He found that the difference between the provided rice and the actual received rice was 18%. These losses can be big enough to decide that the welfare benefits are not enough to continue with the program. Not only can the losses prevent governments from supporting the poor, corruption can also minimalize the opportunities for reducing poverty (Salim 2014).
The literature gives plenty of reasons why fighting corruption should be high on the agenda. Serious attempts have been made by African politicians, non-governmental organisations and international institutions. As Riley (1998, p. 132) says: ‘Serious attempts to control corruption are often as old as corruption itself’. Broadly, there are four levels in which
10
anti-corruption strategies are developed: international; national; local; populist (Theobald, 1990). Regardless of the level, most of the anti-corruption strategies are based on influencing the incentives, supplemented with trainings for officials and knowledge enlargement in general (Riley 1998).
This is because these strategies have their roots in neo-classical economic theory, for example the influential work of Klitgaard (1998). He argued that the formula for the causes of corruption is: ‘Corruption equals monopoly plus discretion minus accountability’. While Klitgaard is correct that some corruption emerges because unaccountable officials have the possibility to make unrestricted decisions, framing the causes of corruption as a ‘formula’ (which implies that it is a fixed precedent) is imprecise and misleading.
Within the neo-classical economic theory a revisionist approach is often applied ‘which defined corruption in terms of divergence from a specific norm of accepted behaviour’ (Caiden 1977, p. 102). In the last decade corruption was mostly framed as a principle-agent problem (Klitgaard 1998, Persson 2013, Rose-Ackerman 1997). The first core element of this theory is that a goal conflict exists between the two actors, the principle and the agent. The principle represents the public interests and the agents have the possibility to use public means for their own benefits. The second element is that there is a knowledge disparity because the agents have more information than the principles, hence creating an information asymmetry (Persson 2013).
All these theories and approaches imply an individual contemplation whether to be corrupt or not. It is based on a rational actor approach; ‘whether an individual pays a bribe depends on his or her cost-benefit calculation of the cost of the bribe for a service in comparison to the utility of obtaining or doing without the service’ (Justesen and Bjornskov 2012, p. 2).
These theories form the basis of, for example, the World Bank strategies. Changing incentives and implementing more sanctions are two elements of its 10-piece list on how to fight corruption (World Bank 2015).
Anti-corruption agencies in Singapore and Hong Kong have been very successful following this way of theorizing corruption and fighting it. However, as pointed out in the introduction, no fully effective anti-corruption strategies have yet been developed for African countries. In a considerable amount of cases it has even be proved to be counter effective. According to Persson (2013) this is due to a mischaracterization of the problem of systemic corruption.
11
In the literature a distinction is made between grand corruption (which affects the government on a large scale) and petty corruption (affects a group of people on a small scale) and both types can be systemic or incidental (Fry 2015). Kenya is often mentioned as one of the most corrupt countries in Sub-Saharan Africa because it has both grand and petty corruption being systemic (Fry 2015, Hope 2014, Kivoi 2013). Incidental corruption is evolving into systemic corruption when it becomes part of the day-to-day life and part of its organisational structure. Systemic corruption is embedded in the core system of the society (Arellano‐Gault 2016).
The problem of the above-mentioned theories and approaches is that they basically ignore the different characteristics that systemic corruption has, relative to incidental corruption. ‘Corruption’ is a qualitatively different problem when systemic: it is embedded in political and economic systems in ways that both reflect its impact and help sustain its force’ (Johnston 1998, p. 88). When corruption becomes more systemic, other influences become more important than individual choice.
Articles that dare to specify systemic corruption more comprehensively than ‘when corruption becomes the rule rather than the exception’ (Johnston 1998), ‘a situation where wrong-doing has become the norm’ (Caiden 1977) or that corruption becomes a way of life (Myint 2000) are scarce. Instead of a full description, they rather mention some characteristics which will be discussed here.
O’Hara (2014) argues that systemic corruption is endogenously ingrained in institutions and Popov (2015) states that in a systemic corrupt country no institutions can escape from being corrupt: ‘in a corrupt society, it is never the case, for example, that the police are corrupt, while educators are not. Moreover, a corrupt policeman could eventually interact as a client with a corrupt educator, who in turn could be a client of a corrupt doctor’ (p. ?).
Arellano-Gault (2016) is one of the few who provides a definition of systemic corruption. According to him corruption has become systemic when it has become part of a coherent and deeply embedded system of understandings about social life, in which it is not seen as a disease. The corruption plays an important, or even crucial, role in the organisational aspects of the country and has become part of the culture. Riley (1998) defines systemic corruption by comparing it with incidental (small-scale, only few public officials) and systematic (has a small impact on the government and development) corruption. Systemic corruption is more influential compared with the latter two: every element of the system is corrupt and it has a huge developmental impact.
12
The neo-classical theories regarding corruption might be applicable for incidental and systematical corruption, but they are certainly insufficient to describe systemic corruption. Despite being restricted and minimal, the literature does provide descriptions for systemic corruption, just not for the situation in which the systemic corruption has become so structured that is has become self-containing. Three important elements, the ones founding the three sub-questions, are missing. This will be elaborated on in the theory section.
13
Theoretical Framework
This section will elaborate on the theory that corruption can become so structured that it grows a self-containing character. This will be followed by the research objectives, this part focuses on the goal of this research which is to establish a comprehensive description of self-containing corruption. The third part explains the relevance of a research like this.
Theory
To properly describe systemic corruption to find out whether or not the corruption has become self-containing, three elements are crucial. These elements form the main characteristics for describing possible self-containing corruption.
To begin with, the choice of whether to be corrupt or not is not an individual one in a systemic corrupt country. Therefore, framing systemic corruption as a problem of personal incentives with punishment as a solution is not correct, Previous research about systemic corruption has shown that individuals are influenced by the corrupt environment to an extent that makes it almost impossible for them to make an individual choice. Persson (2014) argues that not being corrupt when it is expected can lead towards social exclusion. People who have, but not use, the possibility ‘to eat’ are considered foolish, or ‘they are actually gonna curse you (Persson 2014, p. 458).
The choice of whether to be corrupt or not does not only become a collective one because of the outside pressure. Because the corruption becomes so normalized, people will feel that they have to be corrupt because everybody else is corrupt as well. Being corrupt has become the norm to such an extent that not being corrupt will make one disadvantaged compared to the others.
These phenomenons are a fundamental element of self-containing corruption, but they are also applicable to systemic corruption. However, I like to argue that when corruption becomes self-containing in a certain sector, the choice of whether to be corrupt or not is not only not individual anymore because of pressure (both from outside as well internally), but also because people are forced by the corrupt system itself. The corruption has become so normalized that a structure – which can have various forms – is developed to keep the corrupt system in place. Because the corruption is established in a structure, not being corrupt is no longer an option. Punishing individuals by removing them will only lead to their replacement by other individuals. The system is stronger than the participating individuals (Puckett 2010).
14
Subsequently, this ‘structure’ is the second element that characterises self-containing corruption. The term structure refers here to a framework of measures that have been evolved to guaranteeing the continuation of the corruption. Usually, when describing systemic corruption, it is argued that the corruption has become embedded in the social and organisational structures of society. Essentially this means that it has become a part of daily life (Arellano‐Gault 2016).
I would like to argue that when corruption becomes self-containing in a certain sector, the corruption does not only has a part in the daily organisation of society but also its own organisational structure. This structure can include various components, depending on the specific sector. This will be discussed more extensively in the theoretical part.
The third element for describing self-containing corruption is the local perception on the issue. Corruption is often, although this is a declining trend, described as an ethical problem. A lack of norms and values is considered to be one of the reasons that the corruption persists in systemic corrupt countries (Kim 2014). Even more often the problem is attributed to a lack of knowledge about the devastating consequences of corruption (Treisman 2007) or it is argued that they are aware of the consequences, but that they deny their personal role in this and explain away their own responsibility (Arellano‐Gault 2016).
I would like to argue that when corruption becomes self-containing, most people will have enough knowledge, they do realise what is wrong and what their own role is, but they have passed the point where they reflect on this. They will not expect any change in the future and therefore they have accepted the situation as it is.
Relevance
The relevance of this study is threefold. In the first place it will provide empirical support for the new theoretical framework to describe systemic corruption, as explained above. Secondly this research will contribute to the still limited amount of corruption data that is directly measured. Thirdly will it provide Kenya with proof about what happens in their public transport section.
Before fully effective anti-corruption strategies can be developed, causal mechanisms about corruption must be found. To conduct research that is capable of identifying these causal mechanisms a foundation of descriptive research is necessary. This research will contribute to the core knowledge about the concept of corruption: it will contribute to the understanding of what corruption actually is.
15
However the current tendency is to make a distinction between incidental, systematic and systemic corruption there is not yet enough attention for the extent in which corruption can be systemic. The new framework, to describe corruption that is so systemic that it has become institutionalised, will provide a better understanding of corruption. As a result of this, anti-corruption strategies can be improved. They will be better adjusted for the situation and thus more effective. Sufficient and effective anti-corruption strategies are highly necessary, which became apparent in the literature review. Corruption is often mentioned as one of the main causes of poverty, inequality and conflict in Africa (Justesen 2014, Hope 2014). This new category in the conceptualization of corruption is still in its infancy and this research is unable to cover all of its aspects but it will make a significant contribution to its development.
The second purpose of this thesis is to contribute to the data regarding corruption that is obtained directly instead of indirect. Big data macro level research is mostly based on the perception of corruption: how people perceive the magnitude of corruption in their environment. The issue with this indirect way of measuring is that it is impossible to know if this perception is actually a reflection of the reality. Previous research has shown that the perception and reality of corruption can be divergent. Personal biases - for example: men tend to report higher levels than women – might influence the corruption perception more than the actual levels of corruption. Also it is not reasonable to expect people to have complete and accurate knowledge about all of the corrupt actions in their environment, because it is an illegal activity of which the executors want it to remain unknown (Olken 2009, Olken and Pande 2012).
This research will question corruption, focussing on the action of bribery, directly: this will provide new data about the extent of bribery in this country. The reason that it is possible to study corruption directly in this case is that the stigma on bribery is not as big as on other corrupt actions (Olsen 2009).
The third contribution of this study is to provide Kenya with more scientific proof of the practices in its public transport sector. Traffic, and especially public transport, is a big cause of death, injuries and costs in developing countries. Poorer people especially are often victims of the unsafe public transport because they are (due to financial means) forced to use the cheapest and most unsafe travel methods (Nantulya 2002). This is most definitely applicable to Kenya which is having the highest road fatality rates in relation to vehicle ownership in the world. ‘Nearly 3,000 people are killed on Kenyan roads annually. This translates to approximately 68 deaths per 1,000 registered vehicles, which is 30-40 times greater than in highly motorized countries. Road traffic crashes are the third leading cause of death after
16
malaria and HIV/AIDS and present major public health problem in terms of morbidity, disability and associated health care costs’ (Odero, Khayesi and Heda 2003, p.53).
Despite this huge burden, road safety measures in place are ineffective, characterized by crack downs on motor vehicles following a tragic road crash”. However the low quality of public transport has many causes, and corruption is regarded as determinative because it makes it allows the vehicles to drive around with defects, drivers with not enough experience and a lack of general control (Nantulya 2002).
That corruption between the traffic and the police is extensively present in Kenya has been noticed before. Even recently, a matatu driver had the opportunity to speak to the Kenyan president about this (Peralta 2017). This study might provide actual proof of the systematic demand for money from the police towards the matatus. It will even be possible to make a rough estimate how much they pay the traffic police in Kajiado.
17
Methodology
Case selection
In 2016 Kenya was ranked 145 out of 176 in the International Corruption Index, published by Transparency International. Most countries in the region got even higher rankings (which means more corruption) like Sudan, Somalia and Burundi, but these countries are unstable and vulnerable for (or in the middle of) conflict. Kenya is ranked as being in a ‘medium’ state of peace (GPI 2016), making Kenya a sufficiently safe and responsible research field. Additionally its corruption is famous for its endemic and systemic character.
When looking for the most systemic corruption within Kenya, one institution in particular catches the eye: the police. The police is often mentioned as an institution that is highly vulnerable to corruption in African countries (Kivoi 2013). The Kenyan police force is notorious for its high levels of corruption, Peiffer and Rose (2014) even argue that it is the most corrupt institution in the country. ‘Kenya Police’s long and not so good history is characterized by inefficiency, corruption, disregard for rule of law, police brutality and blatant abuse of human rights’ (Kivoi 2013). These statements are supported by data from Afrobarometer (2015): Three-fourths of the respondents state that ‘most’ or even ‘all’ police officers are corrupt.
One of the elements in society that is influenced by the corrupt police is the public transport sector. Infrastructure is crucial for development. Leo, Ramachanddran and Morello (2015) noted that the need for infrastructure enhancements is a top level issue in almost every African country. Traffic, and especially public transport, is a big cause of death, injuries and costs in developing countries. Especially the poor people in society are affected by the bad quality of the transport, because they do not have the means to pay for a safer option (Nantulya 2002). It is estimated that road traffic accidents causes 1.3 million deaths and 20– 50 million injuries yearly. In Kenya road safety is a serious challenge with multi-faceted causes. The matatus are involved in a large part of the accidents (Raynor 2014).
The public transport sector in Kenya is almost entirely carried out by matatus. These small taxi busses have a(n official) passenger capacity between the 8 and the 16 passengers, leave only when they are full and are often equipped with impressive sound boxes. The voluntary merge of individually operating matatus into Savings and Credit Cooperatives (SACCOs) has resulted in improved service, regulatory compliance and technology adoption, however little is known about how the operate exactly (Behrens 2017). The drivers get paid
18
based on a target, instead of a fixed salary, which leads to over-speeding, long waiting before the vehicles are full, carrying excess passengers and irresponsible driving behaviour.
Behrens (2017) is arguing that because of the watchdog position of the SACCO’s the matatu business has become more organized which has improved the quality. Fixed salaries are coming up, all the vehicles have a speed controller and they have to be inspected every year. He states that ‘properly incentivized and monitored salaried drivers are key to curbing dangerous driving practices and excessive in-vehicle passenger waiting times, and depreciation costing is key to maintaining roadworthy vehicle fleets (Behrens 2017, p. 87). However, the traffic police and its corruption is a major problem, hampering the safety improvements (Raynor 2014). Raynor (2014, p. 245) states that ‘bribery has the become the cultural norm’.
Because of limited manpower, financial means and time it is not possible to study the relationship between the matatus and the police through the whole country. For this reason is the research design is based on a single case-study. This study is focused on the county of Kajiado. The area covers 21,900.9 square kilometers and the population of the county was estimated at 807.070 residents in 2012 (GCK 2016). The county has, like the other counties in Kenya, gained more independence since 2010. Kenya implemented a new system of governance which is more decentralized than before. The services have been brought closer to the people (TI Kenya 2014).
The reason to choose Kajiado is that this county has one clear matatu route. It starts it Kitengela,
goes through Isinya and via Kajiado to
Namanga. The three other towns with matatu services in the county, Ngong, Loitoktok and Magadi, are part of matatu routes which start and end outside the county.
19
Methods
This research has a descriptive approach. Descriptive studies are essential in social science, we cannot establish substantial causal explanations without a good description of the concept. ‘Description comes first; it is hard to develop explanations before we know something about the world and what needs to be explained on the basis of what characteristics’ (King 1998, p. 34). The goal is not to focus on the causes of corruption but rather to create a better understanding of the topic. As explained in the literature review, the characteristics have been uncritically copied from incidental corruption to systemic corruption for too long. Before causal explanations of systemic corruption are possible descriptive analysis of what systemic corruption is, is necessary.
The strategy that is chosen is a case-study. For the data collection quantitative and qualitative research methods are combined because both approaches have their advantages. Triangulation, the combination of the two methods, makes the results more comprehensive and in-depth. It increases the deeper understanding of the studied phenomenon and the validity of the data is higher because the different measurements should tell the same story (Hussein 2015).
Semi-structured interviews are used as qualitative method, and surveys as quantitative method. The interviews, 11 in total, contribute with knowledge about the local perception of the problem. They provide the background information and also details about the operation of the process. The respondents for the interviews included three SACCO supervisors, these men are responsible for arranging all the logistics on the site; the matatu stage. It also involves two interviews with SACCO managers, who are responsible for the SACCO organisation for the whole region. The other six interviews are in-depth interviews with experienced drivers, matatu-owners and one retired police officer.
This data is complemented with quantitative data from the surveys. This surveys, answered by 144 matatu drivers, provide more knowledge about their experience. They answered questions about all the actions and transactions between them and the police, as well as detailed questions about the exacts amounts of money that they pay and how and when they pay. And finally they answered questions about their opinion of this corrupt system and the possibility and probability of change.
The advantage of this choice of methods is that it creates an opportunity to validate the answers. The stories told by the managers have to match the stories of the supervisors and certain aspects (when asking about facts of this corruption) of the stories of the drivers have to
20
be similar. This direct measurement of corruption is possible because of there is a low (or non-existence) stigma on bribing (Olsen 2009).
21
Results
The first part describes the corrupt system between the traffic police and the matatu business, based on the information coming from the semi-structured interviews. The second part presents the statistics coming from the survey. This will be concluded with the analysis section, where the data from the interviews and the surveys is brought together to answer the sub-questions.
There might be a small repetition in the statistics after the situation description. The reason for this is that the situation description is based on the information coming from the interviews. The statistics part is solely displaying the results from the survey. The repetition is increasing the validity of the data, considering that the different data has to match to be veracious.
Empirical evidence - interviews
Based on the semi-structured interviews the corrupt system between the matatu business and the traffic police is described. This description is holistic, not yet zooming in on the three crucial elements for describing self-containing corruption, to portray the whole situation. First the different actors and their roles are discussed. These are the drivers, the SACCO, the traffic police and the superior of the traffic police. This is followed by a description of the system that they have developed together.
All the respondents are men, which is not surprising as it is a men’s business. Until a few years ago, the matatu drivers had a very bad reputation of ‘misogynistic language, their overcrowding, their playing of loud music, their speeding, and their rough-handling of passengers’ (Mutongi 2006, p. 549). Back in the days, matatu drivers often had dubious additional earnings being involved in criminal activities (Interview respondent C). Although their reputation has improved the drivers still have to deal with the image of doing ‘a dirty job’, they face stigmatization and discrimination (Ferrence 2016). One of the drivers said: ‘this is not a woman’s job, they shouldn’t do this kind of work’. However they are not forbidden to work in the business, the SACCO Supercoach has a female supervisor in Namanga. In a small talk conversation she said: ‘They (the drivers) see me as the big mama, that is my authority’ (Respondent F).
All the matatus have to be organised in SACCOs: Savings and Credit Cooperatives. This started as voluntary but has become an obligation in 2010 (McCormick 2015). The SACCOs consist of a board, the executive managers, the supervisors and the members. The board consists of a president, vice-president, secretary, treasurer and a member representative.
22
The respondents know that the board is responsible for decision-making, but they do not know what other responsibilities they might have. Every region has a manager and every matatu stage has a supervisor. As became clear in the interviews, the members of the SACCO are the owners of the matatus. Sometimes they also drive the vehicles themselves but more often they hire a driver. Matatu owners can have more than one vehicle in one SACCO but when they have the financial means for more they often spread them over different SACCOs. ‘Being a member of a SACCO gives you one vote, not matter how many vehicles you have. When you have more vehicles you distribute them over more SACCOs, which gives you more votes’ (Interview Respondent D). According to the matatu owners, the most important benefit for being a member in a SACCO is the possibility to get loans. One vote equals one loan, with an agreed fixed amount (determined per SACCO). Therefore having multiple matatus in one SACCO is not beneficial, this would only produces one loan, whereas multiple SACCO memberships guarantees multiple loans.
The drivers, unless they are also the owner of the vehicle, are not a part of the SACCO structure. They have no voice in its decisions and do not receive their salary from the SACCO. They only have contact with the supervisor on the stage, who registers which car arrives when, and which car is allowed to leave. Drivers gets a part of the daily profit, sometimes this is a percentage, sometimes an amount that is agreed on in advance with the owner. A number of drivers have said that they sometimes go home without any money at the end of the day, because they have to pay for gas, pay the owner, the conductor and the touts. The conductor and the touts can be seen as assistants of the matatu. The conductor drives along with the vehicle, helps people in and out the vehicle and collects the fares. The touts are stationed at the matatu stage, making sure that people get in the matatu of their SACCO instead of in one of the others (whether or not through some shouting and pushing of the customers).
The existence of the job of the touts is based on the fierce competition between the SACCOS and the individual vehicles. When asked about challenges in the matatu business this competition is mentioned as second biggest challenge, just after the corruption. There are too many matatus operating on the same route (or too few passengers), resulting in long waiting times – unbeneficial for both customers and crew – because the vehicles only leave when full.
It is not the SACCO but the owner of the vehicle that is responsible for the maintenance of the car. The government has implemented some requirements, ‘the Michuki rules’ – named after the then transport Minister John Michuki – in 2014. The vehicles have to
23
be equipped with speed governors and safety belts, the drivers have to wear uniforms and they need to carry identification cards and their driving license (McCormick 2015). The respondents also mentioned that the both the driver, the conductor and the car need a PSV. PSV stands for ‘public service vehicle’ and having a PSV for the driver, conductor and basically means that they are licensed to transport other people. The license of the car has to be visible on the front screen, together with the sticker that shows the presence of the speed governor. These requirements have to be renewed every year, thereby giving the government the opportunity to control the vehicles regularly and guarantee quality and safety. Well, that is at least the theory. In reality the compliance with the rules is quite low. The speed governor is a small device that registers the speed of the car and prevents it from driving faster than 80kph. However, I have observed a lot of cars drive around with non-working speedometers, putting the speed governors on the sidelines because they cannot register the speed anymore. A lot of drivers expressed their opinion that carrying your driving license with you when the police know who you are is an unnecessary measure. The stickers on the front window are often expired, which means that the vehicle has not been checked recently.
When trying to identify the culprit, the traffic police seems to be the most obvious. They force the matatu drivers – and the other road users who are not included in this research – to pay money on a weekly, daily and/or random basis. The disadvantages for them are low because there are little possibilities to sue them and the advantages are low. When asking the drivers about the rumours about how much the police earns because of the bribes, the estimates range from ‘they earn their monthly salary – which is around 30.000 KES1
– in one day because of the bribes’ (Interview Respondent A) to ‘they can increase their salary tenfold because of the bribes’ (Interview Respondent E). The drivers only base these estimates on their own experiences and the stories they hear, but it might be clear that the police make a lot of money out of bribing. There also seems to be a clear power inequality, the police officers can ask the drivers the amount they like. When the driver tries to be lawfully correct and does not accept the bribe, the police make up more charges to make sure that the driver does not go to court, and instead pays on the street.
But when looking at this more accurately, the situation is more complex and the police officers are not simply the only guilty ones. They have the obligation to amass enough money on the streets to provide their superiors with ‘something’ to eat. How much they are obliged to give away is not known by the drivers, but that the officers have no choice but to collect
24
enough money is commonly known. ‘When an officer would not ask for money on the streets, he cannot provide its supervisor with enough money. He will be exiled to an almost unpaid – not enough to support its family – office post’ (Interview Respondent A). Another driver said: ‘some of those officers are my friends. I do not blame them, we cannot refuse them like they cannot refuse their bosses, like they cannot refuse theirs. We are all in this together’ (Interview Respondent H). All the drivers describe a chain, everybody is forced by their superiors.
However, the chain is even more complex than only the drivers who are forced by the police officers, who are forced by their base-commander, who is forced by their superior and so forth. Among the actors in this corrupt actions there are also the executives of the SACCOs. The older drivers, with more than 15 years of experience, all stated that 10 years ago they only had to deal with the police officers themselves. There were no agreements about the regularity of the payments. The drivers only had to pay the officers 100KES- ‘yes always 100’ – sometimes when meeting them on the streets, an amount that was a bit higher when the vehicle had a mistake.
Nowadays this has become more complex. The drivers mention two reasons for this development: the increase of anti-corruption officers and the widespread introduction of the SACCOs. In the recent years, the Ethics and Anti-Corruption Commission (EACC) of Kenya has deployed undercover agents to discover corruption. With the 100KES money transfers that took place almost every day of the week, the probability of detection was quite high. That is why the police officers preferred to reduce the transfer moments, changing to weekly payments.
Arranging weekly payment agreements with all the individual matatu drivers would have been an almost impossible task. This is where the role of the SACCO comes into play. The executives of the SACCOs have negotiations with the base-commanders of the regions in which their matatus operate, to determine the amounts that have to be paid. How these negotiations exactly go, when they take place and which actors exactly are involved is not known by the matatu drivers. The result is an agreed fixed amount that the drivers have to pay to the police on a weekly basis. The next section with statistics will elaborate on the details of these payments.
These payments are referred to as ‘general money’, ‘routine money’ or ‘operating money’ by the drivers. This is because these payments are not made because their vehicle has faults they would like to cover, or to prevent so that they are not sent to court. This is money
25
paid simply because they operate in the area of those police officers. This will be referred to as routine money from now on.
In addition to this routine money, there is mistake money which has to be paid - by the driver to the police - to avoid being send to court when they have a mistake. When referring to the defects that their vehicles have the drivers speak of mistakes, that jargon is therefore also used in this thesis. The most mentioned mistakes are: not carrying a driving license or the driver’s picture that has to be in front of the car, not wearing a uniform, low tire pressure and expired stickers. The amounts that are requested for these various mistakes differ based on the mistake, the mentality of the police officer, the relationship that the officer and the driver have and the negotiation skills of the driver.
The motivation for the drivers to pay this money is twofold. The weekly payments reduces the contact moments with the police from daily to weekly (or even less, this will be elaborated later), thereby reducing the chance to be caught with mistakes. The second reason is that the officers are much milder when checking the vehicle. This clearly shows the role of the driver as an actor in this corruption structure. The possibility to bribe the police officers permits them to drive around with mistakes – that would be expensive to repair – without being sent to court and pay the much higher fines there. ‘When you have paid your weekly amount and the police next to the road tries to stop you, you just flash your lights and you can continue […] They check their records and if they see you did not pay they will come after you’. It is striking that the drivers, although they do like to complain about the traffic police, do not actively try to stop them. They even support the police by protecting them from the anti-corruption officers through their agreement of the ‘getting caught’ avoiding measures.
All the actors together – the traffic police, their superiors, the SACCO executives and the drivers – have created a system to keep the corruption in place. To maintain this system, a structure has been developed. There are undoubtedly more, but three elements to maintain this corrupt system have come to light. This first one is, as shown by the quote, that the police are keeping recordings of which vehicle has paid its weekly commitment. The second one is that the paying actor – the driver – has the possibility to complain. When the police officers go too far and request too much money from the drivers, they can go (after consulting the SACCO supervisor or manager) to the base-commander to complain. The base-commander is likely to agree with the complaining driver to avoid conflicts, provided that it does not happen too often. This because the base-commanders have so much interest in a continuation of the bribery that they will do everything to keep the system in place. This complaining possibility was only mentioned by three of the 11 respondents, therefore the validity of this statement
26
might be on the low side. However, no other respondents have said anything that is denying this element.
The third element is the implementation of the middleman. Some drivers pay their routine money to the police directly, when meeting them on the streets. Others pay to a middleman on the matatu stage. This man differs per week, it is usually a driver with good contact with the police officers of that town. He collects the money, writes down the number plates of the vehicles that have paid and hands over this money and registration list to the police. This measure lowers the chance for the police officers to be caught with corrupt actions. Most drivers have various payment methods: in some towns they pay on the streets and in others they work with a middleman. There is more about this discussed in the statistics part
Empirical evidence – surveys
Almost 90% of the respondents answered that they bribe on a regular basis; weekly or daily. Interestingly, none of the respondents answered that they pay bribes just occasionally; they either pay regularly or nothing at all. The drivers stated that this dichotomy was developed in recent years, before it was possible for the drivers to pay the traffic police occasionally; when they were stopped on the road. It was
always the same amount, 100KES2. Because of the establishment of anti-corruption officers (who are travelling undercover with the matatus sometimes), these payments became too risky. That is why the police switched to another method; the weekly payments. Because the money transfer is only weekly, the chance to be caught is lower. However, some drivers explained that they still pay daily; the weekly-payment system is still developing and expanding. The matatus are organised in SACCOS. This recent development (it is only since a few years that it is mandatory for a matatu to be in a SACCO) creates the possibility for collective negotiations between the SACCO owners and the traffic police. Once in a while the manager of the SACCO that operates in a certain area sits down with the base-commander of that region to negotiate about the amounts of money that have to be paid. The base-commander has the command of all the traffic police officers in one village and the surroundings. 2 100 KES = 0,96 USD 70,50% 11,4% 18,10% Do you bribe? Weekly Daily Never
27
The majority of the respondents did see this corruption as a problem. They feel that they pay unfairly, that it is a waste of money and that it decreases the quality of the transport because there is no money left to pay for fixing the defects. A large majority of drives claim that the vehicle in which they are driving has at least one mistake at the moment and that all the vehicles have that, even the new ones. They argue that, regardless of the lack of money, there is also no point in fixing these defects because the police will make you pay anyway. ‘A few years ago we did try that. We repaired all the mistakes, but then the police charged us for other things. Like drunk driving, even if I was completely sober. You cannot fight it’ (Survey Respondent 26). Some drivers claimed that they
sometimes try to do the good thing and not accept the bribing option. ‘I once said: okay send me to court because I do not wear my uniform. But the police won’t accept that. He said: yes of course I’ll write you an official receipt. No uniform, but I also see that your tires are flat, your breath smells like alcohol and that you have three excess passengers. I was sober as a child, and the other charges were as false’ (Survey Respondent 102). Because the police prefers being bribed, they make up extra charges if the driver is not willing to bribe. They do this because it benefits them, they make more money out of bribing than they do from their salary, but also because they are forced by their bosses. The base-commander will expect to receive a certain amount3 of his traffic police officers and if they do not meet this threshold they will be stationed in an office, where they cannot make enough money to maintain their families.
However, a striking 58,8% of the drivers would not prefer a change in the system if this would mean that they could not bribe themselves out of an official charge. They acknowledge the corruption as a problem with disadvantages for themselves and the society, but if they would be offered the choice between keeping this system or a corrupt free system
of rule of law, most of them prefer this system. This is because the fines that they have to pay in court, and the time that it costs them to go there, is too much. For example: an excess
3 The exact amounts are uncertain, because there was only one policeman among the respondents. The other
information about the police is coming from the matatu drivers and the SACCO managers.
67,70% 9,80%
29,50% 0
Has the vehicle mistakes?
Always at least one Sometimes No 58,80% 38,60% 2,60%
Do you want change?
no yes No answer 11,70% 68,90% 19,40% 0
Corruption system a problem?
yes no depends
28
passenger (one passenger more than seats in the car) will cost them 500KES on the streets, but in court between the 100.000 and 200.000KES and in addition at least two days in which they cannot drive (and therefore not make money) because their vehicle is repossessed. As a comparison: an average driver earns between the 700 and 1000KES a day.
Most drivers are not optimistic about the possibility of change. Only 20,1% is convinced that change will come or is already happening.
The majority believes that it will never change: ‘corruption is in the blood of the Kenyans, even a baby in the womb is already corrupt’. Some others argue that it might change within years, but none of them knows how this change could or should be achieved. The most important reason that is mentioned why this will never change is that too many people they benefit from the system.
To conclude this data section, one striking question has to be answered. How much money are we talking about? The majority of surveys (Afrobarometer, Transparency International 2016, Worldbank) dealing with corruption do not address this aspect because it is not expected that the respondents still remember the exact amounts, plus one cannot guarantee honesty. But, it is possible in this research because of the research methods and the structure of the corruption.
As shown in the methods section, drivers of different villages have been surveyed. They answered detailed questions about the amounts of money they have to pay and when and how these money transfers take place. If most drivers would make up their stories the testimonials would not match. The second reason why it is possible to question the actual amount of money paid is the structure that is established between the police and the matatu-business. By now this has become clear, the corruption is structured. I will elaborate on the structured character of this corruption in the analysis section.
As can be seen in the first pie chart, 7 out of 10 drivers (70,5%) have at least one weekly payment arrangement with the police. The exact agreements that they have depends on the route and regularity. Consequently the total amount of what the drivers pay differs and is diverse. The minimum that a driver claimed he had to pay every week is 300 KES4. This because he only operates in one village, Namanga, and therefore only has to make one weekly payment. He always pays on Monday, just like 65% of all the drivers that operate in
4 100 KES = 0,96 USD 37,30% 29,40% 20,10% 13,40% Is change possible? Never Maybe (later) Yes I don't know
29
Namanga. The maximum is 2400KES per week and the mode, the most common amount, is 2000KES per week. The average of all the drivers that have at least one weekly agreement is 1573KES per week.
11,4% of the drivers claimed to have no weekly payments, they only pay daily on a regular basis. The amount they have to pay is always the same, 100KES per day per village where they operate. However, the police is not always present on the road. ‘You better not travel on Sunday, everybody knows the traffic police is in church then’ (Survey Respondent 67), but also during weekdays they are not consistently there. When asked how much these ‘daily paying drivers’ met the police to pay, the answers ranged from three to six times, with four as average. The average they pay per week is therefore partly estimated, because the drivers had to guess how much they meet the police. They pay on average 1440KES per week.
When all the surveyed drivers are taken together, the average of what they have to pay is 1587KES a week. However, an important side note has to be made. In the in-depth interviews it became clear that there is a distinction between what the drivers call ‘mistake money’ and ‘routine money’. All the mentioned payments averages exclude the mistake money, because of two reasons. The first is that the validity of the answers would be low. The problems – discussed on the previous page - that usually occur when trying to calculate corruption apply here too: it is not expected that the respondents still remember the exact amounts, plus one cannot guarantee honesty. The second reason is that this research focusses on the structural aspects of the observed corruption. This ‘mistake money’ is an excellent example of how bribery how is seen and studied in most studies. Not including this mistake money means that the total amount of what the drivers have to pay is even higher than the following calculation.
More than 24.000 matatus operate in Kenya (KBS 2016). It cannot be assume that the matatus operate the same in Kajiado county as in the rest of Kenya, but to clarify the magnitude of the problem I will calculate how much money is paid to the police if the average amount of money paid to the traffic police in Kajiado would be the same everywhere in Kenya. Amount of matatus (24.000) x Average weekly payment (1587KES5) = 38.088.000KES per week. Per year is this: 38.088.000 x52 = 1.980.576.000 KES. That is 19.126.779 USD.
5
30
A former director of the former Kenya Anti-Corruption Commission (KACC, in 2011 renamed in Ethics and Anti-Corruption Commission (EACC)) stated that the total transport sector bribes the total police (so not only the traffic police) for an equivalent to US$22.5 million annually (Lumumba, 2011). The transport sector is much bigger than only the matatus: the trucks, ‘Lorries’, (who pay even more than they do, according to a few matatu drivers) the personal vehicles and the big busses are also included. And the police is much bigger than just the traffic police: there is the normal police, and the OCDP. Conclusion: we can expect that the total paid amount is much higher than 22,5 million dollar a year, if only the matatu drivers alone pay the equivalent of more than 19 million dollar to only the traffic police. Despite the attention for the problem is it still underestimated.
Analysis of the evidence
The answer on the question how self-containing the corruption between the police and the matatus is contains three elements. The first and most important one is the extent in which being corrupt or not is an individual choice. The second on is the structure that is developed to maintain the corruption. The third one is the local perception on corruption.
1. How individual is the choice to be corrupt?
All respondents emphasized the existence of ‘a chain of corruption’ in Kenya. The matatu drivers are forced by the traffic police to pay, which are forced by the base-commander, who is forced by the commander of the OCPD (Officer Commanding Police Division) and so forth, all the way up to the top. All the actors in the game are obliged to renounce money to their superiors, an obligation they are punished for if it is not fulfilled. The different actors are not only forced by their superiors, but also by their environment (Arellano‐Gault 2016). One respondent pointed out that almost all traffic police officers have more than one wife and several children which they cannot maintain without the money they receive via their corrupt acts.
When zooming in on the matatu drivers, they feel like they no choice but to be corrupt. The charges in court are so high that they are impossible to pay for a low-paid matatu driver. An additional factor is the police making up extra charges when a driver tries to refuse a bribe, which makes more difficult – basically impossible – to refuse the bribing option. Despite this harsh treatment the drivers themselves do not really blame the police officers, ‘they are also just humans’, but they do blame the system.
Even the amounts of money and the money transfer moments are not in their hands. The manager and/or the board of the SACCO has negotiations with the several
