Tuesday, June 4, 2013

Size, Number and Character of Devolved Units in Kenya

The new county government system was preceded by two main institutions – the provincial administration and the local governments (Local Authorities or LAs). The devolved system of government at independence with 7 regional governments and an eighth one for the city of Nairobi (famously known as the majimbo system), It was short lived as barely a year later it was abolished.It was replaced by the provincial administration (PA). The geographical units comprising the regional governments became the 8 provinces – Nyanza, Central, Rift Valley, North Eastern, Eastern, Coast, Western and Nairobi. Below the provinces were established districts, further divided into divisions, then locations, then sub-locations and villages. At each of these units an officer was appointed to represent the interests of the central government. This system was akin to the colonial system that had been used to control the populace. The political elite saw it as a key means of consolidating power and getting up to date intelligence on any matter.

In later days the provincial administration officers were to play a key role in the running of the ruling party – KANU. At one point it was even difficult to distinguish the boundary between KANU officials and government officials as each played the others role. PA officers became lords in their areas and dictated on every matter. Anyone opposing them was seen as an enemy of the state and acting in direct defiance of the president and was dealt with by brute force and without any fair trial. It is this that has bred distrust and negativity among the Kenyan people for the institution. However in practice the administrative functions they provided have been critical especially in the resolving of social wrongs, certification in matters of registration of persons and property, and as agents of the government on matter of law and order.

Also subdued under the central government were the LAs which were put under the ministry of local government. Before the formation of county governments there were 175 LAs. In theory the LAs were expected to be where political, administrative and fiscal decentralization was exercised. However, in practice they had weak system that saw the councillors elected while their executive arm appointed by the central government thus undermining their independence and ability to account to the public. Their limited capacity, controlled resources and rampant corruption with impunity further undermined them and worsened their performance. It was the abuse and failure of these two systems that led to great agitation for a fully devolved system of government.

Through the Constitution of Kenya Review Commission) CKRC) draft constitution, and deliberations at the National Constitutional Conference  (Bomas) and later of the Committee of Experts (CoE), devolution emerged as a key principle of whatever system was arrived at. Kenyans wanted to ensure that they had a say in their local affairs and that the leaders they elected had resources and space to make and effect their decisions. They also wanted to be able to have key services delivered by institutions close to them. Kenyans wanted a government that is close and effective, a government that would enable them to resolve long endured injustices.

This desire raised the question: what kind of structure would best deliver this? How many levels of government and how many units at each level? And what would be the powers at each level? What would be administratively and economically sensible so that Kenyans would not shoulder too heavy a burden?

According to the CoE, the key factors to consider in determining the levels, number and size of units included the geographical features of areas in relation to the services to be delivered; means of communication or accessibility for effective governance; density of population; resources, including human and physical infrastructure; social feasibility in terms of accommodating into the administrative unit; the historical and cultural ties of communities; minority interests; and the views of the people. While CKRC had recommended five levels of devolved government, the CoE through its harmonized draft constitution released to the public in November 2009 recommended three levels: National, Regional and County governments. The counties were to be the basic unit of devolution and there would be 79 of them based on the districts agreed in the Bomas Draft.

The regional governments, though not the basic units of devolution, were seen as important as they would be large enough geographical units with substantial populations and would  accommodate ethnic and cultural diversity and contribute to nation building; they would facilitate coordination of county governments and planning for services that cut across county boundaries; and they would form a productive linkage to the national government especially for equitable allocation of resources and the protection of the interests of devolved governments (CoE Final Report, 2010). Upon receipt of the views from the public it was deemed better to have only two levels largely on argument cost and simplicity of the system. There was also criticism that the CoE had given no very clear role to the regions. Thus the regional government level was dropped. The districts as enacted in 1992 under the provinces and districts act were adopted as the basis of the county governments.

What are current structures like – number and size?

The Constitution of Kenya, 2010 establishes two levels of government – national and county. The national government comprises the executive, a bi-cameral parliament and judiciary. 47 County governments are established each with an executive and legislative arm. Two main reasons seem to have been key in adopting the 47 units. First was the need for units that were easy to manage in terms of costs and size. One option was to have the 79 districts adopted at Bomas as the basis of counties. These were seen as too many and economically unviable. The other option was to go for larger units (say 25) that would mean the merging of some districts. This while making economic and administrative sense sounded to be politically unviable given the interests of various political groupings. Secondly was the delicate balance between having units that are small enough to ensure effective participation but also large enough to maximize on economies of scale of delivering key services. As a compromise, the 47 districts, existing in 1992, before more creation of districts of dubious constitutional validity, were adopted. This did not require a change of names or boundaries.

The Constitution does provide a mechanism for changing boundaries (Article 188), but this would require the recommendation of an independent commission, and then the support of two-thirds of all the members (not just those voting) of the National Assembly and of two-thirds of the Senate delegations (which basically means of two-thirds of the Senators). This will be probably hard to achieve. No change was allowed before the first elections, which took place this year.

The 47 counties vary in every sense of the word. The largest is Marsabit (70,961 square kilometres) and it’s also the one with the least population density – of 4 persons per square kilometre. Mombasa is the smallest county (219 square kilometres) after Nairobi and Vihiga which have a size of 695 and 531 square kilometers respectively and has a density of 4292 persons per square kilometre.. The table below shows the 5 largest and 5 smallest counties.  An interesting observation is that the size of the counties also has a close correlation with population density and poverty levels. Larger counties have high poverty incidence save for Garissa which is just below the 50 per cent mark and they are also sparsely populated. They are also remote and among those that have a low County Development Index and will thus receive a share of the equalization fund (CRA, 2013). This already indicates that the cost of delivering services there will be higher especially in terms of infrastructure than in smaller and urban counties.

Counties by geographic size in relation to population and poverty indicators (First and Last 5)












Size (km2)






















Density (persons per km2)











Poverty Incidence (%)











Source: CRA County Factsheets, 2012

Key characteristics of the devolved units

The 47 districts were not very different from those created, and named in colonial times, when the rationale was to divide and rule Kenyans. People were balkanized into ethnic blocks and pitted against each other. Others – as with the Kalenjin, Luhya and Meru – were put together as closely related yet they have significant dialect and sub tribe groupings. It is thus of interest that we have adopted the same districts with an objective of fostering national unity by recognizing diversity. With the exception of Nairobi, Mombasa, Nakuru and Eldoret, most of the counties have homogenous majority communities, and they consider the county “theirs”. Even the names of some counties correspond to the name of the majority tribe such as Turkana, Samburu, Pokot, Nandi, Kisii, Embu, Tharaka and Meru. This has raised concerns among minorities at county level. These are persons from communities with smaller numbers in each county and who fear being marginalized.

Given that political organization in Kenya has largely been on ethnic lines, we have started seeing possibilities of certain counties having an assembly and executive with members from only one community. This is why certain county assemblies will be without a single opposition member, as all persons elected are from the dominant political party. Examples are Mombasa with all MCAs from ODM, Kirinyaga with only one MCA from outside TNA out of 20, Kisumu  and Elgeyo Marakwet where all MCAs save three are from URP.. This presents a unique challenge for democracy in Kenya. The constitution makes provision for law to ensure that the county government reflects the community and cultural diversity and protection of minorities (Art. 197). In counties like this, diversity and minority protection may only be achieved through the party list members (often called “nominated”), for marginalized groups and gender balance, but it is unclear how far this has actually been achieved.

Does it matter if a county is homogenous or not? In theory, it is argued that homogenous sub-national units would have an easier time in agreeing on policy priorities. On this basis, one would expect counties with greater homogeneity to progress faster. This assumes the fact that homogeneity is also reflected in policy and political ideology. This would be interesting to test in Kenya as – while counties may be largely homogenous, at least ethnically, and even may have voted for persons in the same political party – they may not hold the same policy positions. This is because during elections, candidates do not necessarily campaign on an agreed set of party policies but on personally identified issues. Thus once in office, they are keen to have the promises they made to the people accomplished even if that may be at variance with what their party stands for. It is of course worth noting that the current devolved units have also meant that there are no more “home” regions for the larger ethnic communities. Good examples are the Kikuyu and the Luo who have been split into 5 counties. The inter-county competition may mean that there is less focus on negative ethnicity. This has the potential of creating an equalizing effect with smaller communities having a greater say in the county.

Here in Kenya, the functionality of counties seems to have overridden concerns about ethnic inclusivity, the latter being seen as an outcome of other factors such as resource allocation. It has been argued that the reason Kenya has experienced ethnic related conflicts is because one community (or group of communities) has been denied its rightful share of the national cake by another. Thus, goes the argument, if all matters of inequality – especially economic and political – are addressed then the ethnic antagonism would resolve itself. It is also why many a community want their own person to take the helm of power at the national level so as to be able to get greater opportunities for their areas. This has been a key argument for devolution of power to make and distribute wealth.

The devolved units also differ greatly in terms of natural resource endowments. While past focus has been on agriculture production of the earlier marked highlands, recent focus is on natural resources exploitation including precious metals, minerals, oil and gas.  Already there is a rush to acquire land in areas thought to harbour such resources. Other natural endowments are based on flora and fauna such as parks, mountains, valleys, lakes, forests and jungles. These endowments and how they are utilized will determine the progress of counties. For instance the great wildlife potential in parts of Eastern and North Eastern of Kenya remains largely untapped with tourism focusing on the west, central and coast of Kenya. Agricultural potential must however not be undermined as it still holds the key to food security and large scale employment creation. Human resource potential will also be key differentiating factor. This is largely as a result of progress in education access and attainment. This of course assumes that the well educated professionals will be willing to take up jobs and invest in counties.

 Assessing the impact of the current structure

The main concern is that the 47 devolved units are far too many for the economy of our size. Each comes complete with an executive and assembly functions modelled against the national ones. The cost is especially in relation to the wage bill and multiple overheads of running government. This would even be worse should corruption take root in the counties. As to whether counties need more money or better investment on key priorities may also determine the costs. More money is not necessarily equal to more development.

 There has been criticism of the scheme of building devolution on the basis of sharing national revenue because it creates an impression that there are unlimited resources held in some place that need sharing. It is emerging that with counties wanting to control more and more resources, counties with limited resource endowment may find it difficult to run. This is because the national government will have less and less revenue to share out. And those likely to be affected the most are the larger, and poorer, counties although that may be debatable too, especially with discovery of oil in Turkana and gas in Marsabit, as well as the great livestock potential of these areas. A shift of focus to internal generation of wealth through available opportunities would seem a better approach to address the economic challenge of counties.

As for development, the 47 units present more opportunities than challenges. The opportunities are in harnessing the competitive advantage of the counties and especially the human and natural resources.  There is a further opportunity in counties forming joint forums/authorities through which they can pursue cross-cutting development goals. A challenge will be in consolidating the support of all actors of county and national government including institutions such as state owned enterprises and state corporations. Private sector and civil society organizations have also built great capacity in dealing with some key issues and hence counties would greatly benefit in engaging them. All said, the differentiating factor between counties development is likely to be based on human effort, investment priorities and resource endowment and not so much the size of the county and the number of counties.

The writer is a policy analyst specializing in devolution and works at the Institute of Economic Affairs where he heads the Futures Program. This article is part of a series published every saturday by the star newspaper and coordinated by the Katiba Institute. It was published on May 18 2013. For other articles check http://katibainstitute.org/devolution-issues 

Some thoughts on future of Universities and Scholarship

It was my pleasure to give input on this topic at the just concluded DAAD Young Scholars in Africa Conference held in Nairobi, Kenya. As one...