How faulty census numbers influence sharing of revenue

What you need to know:

  • The allocation are based on census findings that were disowned by the Kenya National Bureau of Statistics, which had demanded a recount in eight districts.
  • Sharing of county funds follows a formula in which 49 per cent of the allocation is based on population.
  • Data from the statistics bureau, the CRA and the Constituency Development Fund Board show how counties such as Turkana, Wajir, Garissa, Marsabit and Baringo have been getting the largest share of resources allocated to devolved units, in some cases, on the strength of highly contentious population figures drawn from the 2009 census.
  • Population figures from the last five censuses show that whereas the average population growth between 1969 and 2009 was 3.71 per cent, counties in the northeast have registered abnormal growth, sometimes of up to 15 per cent.

The allocation of resources to counties is likely to come under more scrutiny, particularly because the formula used relies on faulty population data.

Traditionally sparsely populated areas, which are also marginalised and lagging behind in development, are now allocated double the amount per resident compared with richer counties.

A survey by the Nation shows that the devolution of resources is transforming the lives of the people, especially in rural areas, but fairness in the distribution is likely to become an issue.

The allocations are based on census findings that were disowned by the Kenya National Bureau of Statistics, which had demanded a recount in eight districts.

Sharing of county funds follows a formula in which 49 per cent of the allocation is based on population.

A county such as Mandera, whose population is recorded as having grown by 400 per cent in 10 years — a demographic impossibility — has one of the highest allocations.

The Committee of Experts, which wrote the Constitution, never intended to have the funds shared according to a static formula. Rather, it wanted an institution, the Commission on Revenue Allocation (CRA), to share the money on the basis of a qualitative analysis of each county’s needs, to ensure equity, to help poor counties catch up with the rest and to take into account the outcomes of previous investments.

GLARING INEQUALITIES

Data from the statistics bureau, the CRA and the Constituency Development Fund (CDF) Board show how counties such as Turkana, Wajir, Garissa, Marsabit and Baringo have been getting the largest share of resources allocated to devolved units, in some cases, on the strength of highly contentious population figures drawn from the 2009 census.

In contrast, counties in densely populated but highly productive regions continue to receive comparatively low allocations on the basis that they are more developed and do not, therefore, require affirmative action in the distribution of the national cake. 

An analysis of the 2013/2014 Budget for both CDF allocations and disbursements to county governments lays bare the glaring inequalities in resource distribution.

The figures reveal that the 15 counties that received the highest allocations per person based on the 2009 population figures were Isiolo (Sh16,912) per person, Lamu (Sh15,757), Marsabit (Sh13,972), Samburu (Sh12,525), Tana River (Sh10,956) and Taita-Taveta (Sh10,940).

Others were Turkana (Sh9,228), Wajir (Sh8,531), Elgeyo-Marakwet (Sh8,477), Garissa (Sh7,537), West Pokot (Sh7,007), Laikipia (Sh6,907), Tharaka-Nithi (Sh6,664) and Mandera (Sh6,610) per person.

The figures reveal that the total population in these counties is about 6.5 million and they enjoy a total of nearly Sh58 billion, translating into an average allocation of Sh8,780 per person.

CONTENTIOUS FIGURES

Contrast this with the fact that the counties that received the lowest allocations per person — an average of Sh4,217 during that financial year — have a total population of 19 million, almost half the country’s total population.

They are Nairobi (Sh3,153) per person, Bungoma (Sh3,647), Kiambu (Sh3,859), Meru (Sh4,060), Nakuru (Sh4,341), Kakamega (Sh4,429), Uasin Gishu (Sh4,548) and Murang’a (Sh4,585). Others are Mombasa (Sh4,628) per person, Siaya (Sh4,715), Kericho (Sh4,764), Trans Nzoia (Sh4,791), Narok (Sh4,872), Machakos (Sh4,982) and Kisumu (Sh5,022).

Whereas the CRA will readily argue that the allocations were derived from the revenue-sharing formula that considered five parameters  population (45 per cent), the basic equal share (25 per cent), poverty levels (20 per cent), land area (eight per cent) and the fiscal responsibilities of the county governments (two per cent) figures from the statistics office show that the sharing formula is based on contentious census figures.

Population figures from the last five censuses show that whereas the average population growth between 1969 and 2009 was 3.71 per cent, counties in the northeast have registered abnormal growth, sometimes of up to 15 per cent.

For instance, whereas the rest of the country registered an average growth of between 6.18 per cent (Nairobi) and 1.97 per cent during the 1999 census, the northeastern region's population rose from 371,391 people in 1989 to an alarming 962,143 people in 1999, a growth of 15.91 per cent.

ABNORMAL POPULATION GROWTH

Ten years later, census figures showed that the region’s population had expanded to 2,310,757 people, with then Planning Minister Wycliffe Oparanya stating that the figures for the northeast and Turkana were “higher than the population dynamics”.

“We raised this issue and we used scientific methods to correct the anomaly but unfortunately, people went to court and filed so many cases challenging the new population figures we had arrived at using the scientific methods,” says Mr Oparanya, now the Kakamega governor. “They all argued that the census figures should be upheld because you could not adjust census figures using scientific methods.”

The High Court ruled that the minister lacked the powers to cancel the census results, a ruling upheld by five judges.

The abnormal population growth in the northeast since 1989 is anecdotally attributed to the outbreak of the civil war in Somalia following the toppling of dictator Mohammed Siad Barre. This led to an influx of Somali refugees into the country who then proceeded to obtain Kenyan identity cards fraudulently.

An analysis of CDF funds also shows that constituencies in sparsely populated counties receive far higher allocations than their counterparts in densely populated ones.

The 2013/2014 CDF allocations show that constituencies in Lamu received an average of Sh1,252 per person, Taita-Taveta (Sh1,153), Marsabit (Sh1,028), Isiolo (1,014), Samburu (Sh995), Tana River (Sh805), Garissa (Sh726) and Turkana (Sh666).

REPRESENTATION IN PARLIAMENT

In contrast, constituencies in Nairobi received an average of Sh348 per person, Mombasa (Sh423), Bungoma (Sh440), Trans Nzoia (Sh478), Machakos (Sh479), Meru (Sh480), Kericho (Sh482), Uasin Gishu (Sh493), Nakuru (Sh494) and Kiambu (Sh505).

The situation persists in the allocations for the 2014/2015 financial year, with sparsely populated counties receiving an average of Sh1,152 per person while densely and highly productive ones receiving an average of Sh665 per person.

Statistics also show that the skewed population figures have also impacted representation in Parliament, with constituencies in the sparsely populated counties having higher representation per person than those in densely populated and urban areas.

For instance, Lamu County — with 101,539 people — is represented by three MPs, an average of one MP representing 33,846 people.

In contrast, Kakamega County — with 1,660,651 people — is represented by 13 MPs, translating into one MP for every 127,742 people. Nakuru, with a population of 1,603,325 is represented by 12 MPs, translating into one MP for every 133,610 people.

The statistics also show that the counties whose 2009 population census figures were disputed registered the lowest voter registration levels ahead of the 2013 General Election.

For instance, Mandera County, whose population was put at 1,025,756 in 2009, had only 121,005 voters, a mere 11.8 per cent of the total population, while Nairobi, with a population of 3,138,369, registered 1,778,903 voters, translating to 56.6 per cent of the total population.

Turkana County, with a total of 855,399 people in the 2009 census, registered only 120,345 voters or 14 per cent of the population. In contrast, Kiambu County, with 1,623,282 people, registered 860,716 voters, more than half of the population.