Fair representation in new boundaries review by IEBC

The Independent Election and Boundaries Commission will begin a boundary review exercise shortly. This will affect the 290 geographical constituencies in the National Assembly, and 1,450 county assembly wards, but not the 47 counties (each of which produces a woman county member of the National Assembly, a senator and a governor).

Periodic reviews of boundaries are normal. Population changes: Some communities have higher birth rates than others. Some communities attract incomers, and those incomers have left somewhere else. It is common to require adjustments after censuses. Our law requires boundary adjustment must use “enumerated national census figures”, meaning actual figures reported by the Census, not projections made after censuses. The IEBC plans to use results of the 2019 census for which the Bureau of Statistics is already preparing.

Another reason for review is that — especially with devolution — communities and communications will change. What previously made sense as an electoral unit may no longer do so.


County politicians may have views about whether certain communities are within the county because of their votes. And county governments might want certain towns to be within their boundaries because of their possible contribution to rates.

Sometimes communities themselves may want to be with “our people” in the same constituency.

When electing MPs or MCAs, boundaries can affect fair representation. Under the old Constitution, Embakasi, the largest constituency in Kenya, had 19 times the population of Lamu. The vote of a Lamu elector was equivalent to that of 19 Embakasi voters. In a parliamentary system, discrepancies such as those can affect how many seats a party wins, and thus which forms government. If we shift to such a system (as a new Bill proposes) boundaries would have greater significance. Even under our current system, where the President is directly elected, so does not depend on a parliamentary majority, it is not impossible that constituency boundary peculiarities might lead to his not having a majority of supporters in the National Assembly.


It is not practicable for every constituency to be exactly the same size in terms of the number of voters or population. But our Constitution attaches importance to fair representation, and provides that the IEBC must work progressively towards ensuring the number of inhabitants in each constituency is, as nearly as possible, the same. The same applies to wards. As a starting point, in urban areas, constituencies could be up to 40 per cent larger in population than the average, and in sparsely populated areas, up to 40 per cent smaller.

Also, the Constitution directs that other factors in fixing boundaries are geography, town and city boundaries, community ties and interests, economy and means of communication. Such cohesiveness makes it easier for the person elected to represent the constituency or ward. To accommodate these factors, the Constitution allowed even constituencies that are not heavily urban or sparsely populated to diverge by up to 30 per cent from the average.


Our figures are not too much out of line with those for other countries with similar electoral systems (single member constituencies). In the US, the constituencies are fewer. For their 320 million-plus people, they have only 435 seats in the national House of Representatives. But their states have much of the power of government — far more than our counties — so they have a smaller national parliament (Congress). They have fairly strict rules about equality of constituency size. On the other hand, countries with huge, sparsely populated areas, such as Canada and Australia, have very great variations in size of constituencies.


The previous boundaries commission, the Interim Independent Boundaries Review Commission of Kenya (The Ligale Commission), recommended some boundary changes, but their work had to be completed by the IEBC in time for the 2013 General Election. The IEBC used 2009 Census data, and reported early in 2012. A number of challenges were made in the High Court, which gave its judgment in July 2012. Under the Constitution and the law, reviews of constituency names and boundaries must take place periodically. This time, the final report by the IEBC must not be earlier than early 2020, nor later than early 2024.

The law sets out the procedure, which involves public consultation, and the input of the National Assembly, though the IEBC has the final say. Strangely, there is no role for the Senate, even though county assembly wards are to be reviewed.

If this process is not complete at least 12 months before the next General Election, the new boundaries will not take effect for that election. So, the final report this time must be ready no later than mid-2021.

Does this give the IEBC enough time for such a complex process? Last time, the first results of the Census were released one year after the count. A similar schedule this time would mean results being available in mid-2020, giving the IEBC just one year to complete its revision. In the UK, the most recent boundary review took two-and-a-half years (and was only about national constituencies). Another concern is whether the public participation window is enough — just 30 days. Inadequate public participation is likely to lead to more court challenges.


Because the Constitution fixes the number of constituencies at 290, this exercise cannot add to or subtract from this number. An amendment to the Constitution would be needed to achieve this. It would not need a referendum.

The previous boundary review could not abolish any constituency. This no longer applies, and two constituencies could, for example, be merged, or one split among other constituencies, provided there are still 290 constituencies overall.

The number of county wards is not fixed by the law, but by the IEBC itself, so the number can be changed.

The IEBC will not change county boundaries: This would require a special, independent commission set up by Parliament. That commission’s recommendations would have to be approved by two-thirds of all National Assembly members and two-thirds of all Senate county delegations.



Last time, most complaints about boundaries before the High Court were about historic and cultural ties, and administrative convenience. Though the Constitution refers just to these factors, and to equal size, the court pointed out that there are other relevant constitutional values, including inclusion and participation, and national unity, which should guide the IEBC. Boundaries can be rigged to affect the results of elections! It is called gerrymandering! Named after a US state governor and the salamander – because Governor Gerry signed into law an Act changing the boundaries of Massachusetts’s electoral districts to benefit the Democrats — and one of the districts resembled a salamander lizard. But gerrymandered constituencies need not be strange shapes.

Whether unfairness is deliberate or not, there is the question of equality in the sense of whether communities and parties have approximately equal chances to elect people of their choice. Whether inequality is intended or not, discrimination is still forbidden under Article 27.

There is a case from the state of Wisconsin awaiting judgment from the US Supreme Court. The case alleges the state government deliberately did two things: ‘Cracking’ and ‘packing’. The first means dividing a party’s supporters among many constituencies so that they never affect the election result in any, and ‘packing’ means drawing the boundaries so that a party’s supporters are concentrated in a few constituencies, whereas if the boundaries were differently drawn the same supporters could decide the outcome in several constituencies. This could happen here!

Achieving increasingly equal sizes of constituencies and wards, respecting geography, community ties and common interests, and the values of representation of the marginalised as well as rights of equality is no simple task, even when acting with the most honest intentions.

As the IEBC plans to embark on its complicated task, it should perhaps dust off its copy of the 2012 High Court judgment for some useful guidance.

And Kenyans can learn from the US about how to detect boundary rigging — and how to use the Constitution to challenge it.



Stay in the Know!

We respect your privacy.