Interesting Statistical And Economical Inference From Data On The Regions Where The 30 Odd Legislators Who Defected To The Jubilee Alliance Party, Kenya In Kasarani Come From

Posted On // Leave a Comment
By Marcel Masaga 

Mathematics at its core is the science of pattern recognition. It is predicated upon reducing complex relationships into numerical form. And thereby making useful deductions on the correlations between diverse variables and sets of data, in order to come up with information that will help us to have a measure of predicting the future and thus making the necessary forward- looking plans. In a nutshell, math is the seeking out of patterns and using them to formulate new conjectures.

This being the case then, it is incumbent upon us to look beyond the obvious when confronted with new information but instead seek to understand the underlying relationships and patterns that connect them. Hence we must use pattern recognition to reduce big data into useful actionable information. And as sentient beings, blessed with 6 or more senses, use these sensors to collect, collate, assimilate and aggregate information for our current and future survival.

Pattern Recognition

Having said that, as a practitioner of applied mathematics (data visualization) and software engineering, I have been unable to stop thinking about the list of 30 odd legislators that the DP William Ruto recently announced at Kasarani Stadium, during the launch of the Jubilee Party, as having defected from the parties which sponsored them to parliament, be it County or National Assembly. And so my mind has been working in the background in a bid to understand where the pattern lies.

For indeed there are patterns in everything if you look hard enough. And pattern recognition is what has made us survive and be successful as a species (such as figuring out that a lion makes a meal of most people who have the ill fortune of meeting with one and thus avoiding such encounters). I listened keenly as the long list of defectors was called out by the DP, all 31, of them from areas as diverse as Marsabit County to Homabay County all the way to Turkana, Machakos and Kwale County and then all of a sudden it all began to click.

For the unenlightened, it is important to produce the aforementioned list of defectors for the record;

List of Defectors

Those drawn from ODM are;

1.       Gideon Mung’aro (Kilifi North)
2.       Zainab Chidzuga (Kwale County),
3.       Mustafa Iddi (Kilifi South),
4.       Peter Shehe (Ganze)
5.       Masoud Mwahima (Likoni),
6.       Mburi Apuri (Tigania East),
7.       Cyprian Iringo (Igembe Central)
8.       Joseph Lekuton( Laisamis)
9.       Samuel Arama (Nakuru West).
10.   Salim Mvurya  -(Governor Kwale
11.   Ukur Yattani – (Governor Marsabit)
12.   James Rege –( Karachuonyo)

Ford Kenya

13.   Khatib Mwashetani (Lunga Lunga) - Deputy Party leader
14.   Kisoi Munyao (Mbooni)
15.   Regina Ndambuki (Kilome)
Chama cha Uzalendo
16.   Vincent Musyoka (Mwala)
17.   Vincent Munyaka( Machakos Town).


18.   Samuel Moroto (Kapenguria)
19.   Moses Cheboi (Kuresoi South)
20.   Eric Keter( Belgut)
21.   Hellen Sambili (Mogotio)
22.   Simon Kachapin - West Pokot Governor

Narc Kenya

23.   Rachel Nyamai (Kitui South)
24.   Francis Mwangangi (Yatta) Muungano’s
25.   John Munuve (Mwingi Central)
26.   John Waluke (Sirisia),
27.   Protus Akuja (Loima),
28.   Stephen Kariuki (Mathare),
29.   Isaac Mwaura (nominated),
30.   Nicholas Ngikor (Turkana East) 
31.   John Munyes - Turkana Senator

Statistical Clustering

In order to make big data more useful the first step involves clustering, which is useful in organizing data points into groupings which share certain commonalities. This technique is also useful in separating the noise from the signal, and concentrating on the signal in order to make useful inferences which will stand the rigors of scientific scrutiny.

The long definition of clustering (the basis of machine learning and data mining) is that cluster analysis(clustering ) is the task of grouping a set of objects in such a way that objects in the same group (called a cluster) are more similar (in some sense or another) to each other than to those in other groups (clusters).Therefore, we will work to break down the regions that these legislators represent first by county and then by sub-regions (using the former original eight provinces of the Republic of Kenya as regions);

Eastern (Ukambani) - 7

1.       Kisoi Munyao (Mbooni)
2.       Regina Ndambuki (Kilome)
3.       Vincent Musyoka (Mwala)
4.       Vincent Munyaka( Machakos Town).
5.       Rachel Nyamai (Kitui South)
6.       Francis Mwangangi (Yatta) Muungano’s
7.       John Munuve (Mwingi Central)

Eastern (Meru) - 2

1.       Mburi Apuri (Tigania East),
2.       Cyprian Iringo (Igembe Central)

North Rift (Turkana)  - 4

1.       Protus Akuja (Loima),
2.       Nicholas Ngikor (Turkana East) 
3.       John Munyes - Turkana Senator
4.       Joseph Lekuton( Laisamis)

North Rift  (Pokot) - 3

5.       Samuel Moroto (Kapenguria)
6.       Hellen Sambili (Mogotio)
7.       Simon Kachapin - West Pokot Governor

North Eastern (Marsabit) -1

1.       Ukur Yattani – Governor Marsabit

Nyanza (Homabay) - 1

2.       James Rege (Karachuonyo)

Coast  - 7

1.       Salim Mvurya  -Governor Kwale
2.       Gideon Mung’aro (Kilifi North)
3.       Zainab Chidzuga (Kwale County),
4.       Mustafa Iddi (Kilifi South),
5.       Peter Shehe (Ganze)
6.       Masoud Mwahima (Likoni),
7.       Khatib Mwashetani (Lunga Lunga

Others or Outliers (Which In our case constitutes) Noise - 6

1.       Samuel Arama (Nakuru West).
2.       Moses Cheboi (Kuresoi South)
3.       Eric Keter( Belgut)

4.       John Waluke (Sirisia),
5.       Stephen Kariuki (Mathare),
6.       Isaac Mwaura (nominated),

Graphical Representation

For purposes of analysis we will omit the noise which in this set is made up of 6 data-points, since they add no value to our analysis, since they are random (as per our definition) and thus carry no useful information. Furthermore we are going to merge some regions as they share more or less the same underlying informational characteristics for this reason we will merge Ukambani and Meru into one mega Ukambani cluster.


·         Ukambani  – 9
·         Coast – 7
·         Turkana - 4
·         Pokot - 3
·         Marsabit -1
·         Nyanza – 1
The total data-points after cleaning and removing the noise (6) falls from 31 to 25 and they can be represented in tabular and graphical form thus.

Tabular Representation of Data


Graphical Representation (i).  Closed pie-chart.

Graphical Representation (ii).  Sliced pie-chart.


Now this is where all the interesting Sherlock Holmes detective work begins. As a good protégé of good ol’ Sherlock we want to find where the pattern lies in this grouping of data and therefore be in a good position to make inferences which are evidence-based. It is quite clear from the above data visualization that Jubilee Alliance Party reaped the most from Ukambani(36%), Coast(28%) and Turkana(16%), who in turn got the biggest pieces of the pie (pun intended)at 80%. The trillion shillings question therefore is, what does JAP want so bad from these regions that it focused so much energy in getting them to decamp? I daresay that it is their mineral wealth.

JAP is very much interested in these regions which have vast mineral resources and to quote POTUS 42. “It’s the economy stupid”. These regions collectively account for more than 80% of the mineral wealth of the country as they are the regions which have rich limestone, coal and iron ore deposits (Ukambani), Titanium and Rare-earth minerals (Coast) and the clincher, the black gold, natural oil, Turkana, Marsabit and Nyakach at the boundary of Kisumu and Homabay Counties (and yes there are proven economically-viable oil reserves in Kisumu and  Homa Bay County, Nyanza Region) additionally they are all rich in gemstones

Chinese Raw Material Interests

Here are the statistics to back up my claim and they are derived from no less a document than the Kenya Mining Investments Handbook 2015.

Kwale (Coast) – There are Titanium deposits in the Kwale Mineral Sands project totaling over 140 million tons of the mineral this excludes rutile, ilmenite and zircon which in total are worth at least 75.6 trillion Kenya shillings, 75.6 with 12 zeroes, at current prices.

The Mrima Hill located in Kwale County some 70 kilometers from Mombasa has Niobium and Rare Earth mineral deposits (5th largest in the world) worth an estimated 6.24 trillion Kenya shillings that is 6.24 with 12 zeroes.

For those who care to remember this Rare Earth concession was initially given to Cortec Mining Kenya Limited (later revoked) a company which was associated with the slain business man Jacob Juma. China controls 90% of the global trade in Niobium.

·         Turkana - The total economic value of the 600 million barrels of oil so far discovered in Turkana County will net the country of Kenya at least 6.4 Trillion Kenya shillings. These deposits are good for at least 23 years that is according to an Oxfam report.

·         Ukambani - Fengxi mining, a Chinese owned firm, are in the process of exploiting the 400 million tons of coal reserves worth 4.0 Trillion Kenya shillings. In addition to these this region is endowed with limestone with Africa’s richest man investing 40 Billion Kshs in setting up a factory there. The region is also richly endowed with gem and precious stones.

·         Pokot – The Chinese and Indians owned Cemtech Limited have invested 15.5 billion Kenya shillings to set up a cement factory thanks to the vast Limestone deposits in the country, with the total reserves estimated at 5.9 million tons of limestone .The factory expected to produce 24 million bags of cement per annum, which is a turnover of 16.8 Billion Kshs per annum. The region is endowed with more than 35 minerals including gold, gemstone and rubies.

·         Marsabit – There is ongoing exploration of oil and gas in Marsabit County which shares the same geology with Turkana which has confirmed and viable oil reserves.

·         Nyanza – The Kisumu oil Block 12B holds an estimated 22 million barrels of oil. The latest Tullow Oil update states that, “The prospect Ahero-A is estimated to have the potential to contain up to 22 million barrels of 2C (best estimates) Resources.” 22 million barrels is worth 98 billion Kenya shillings at current exchange rates.

The total value of the minerals in the six aforementioned regions conservatively comes to a whopping 100 Trillion shillings (USD 1 Trillion dollars). This is the real reason why JAP worked overtime to net legislators from these regions into the party. It was purely strategic and an attempt to have control over the leaders of these mineral rich regions and by proxy the wealth therein. Since the ratio of sharing royalties between the National, County governments and the people is 80%, 15% and 5% respectively. It is thus a smart idea for any government to want to control the local leadership of such resource-rich Counties.

Enter The Dragon - The Chinese Connection

The Leader Of The Communist Party Of China Delegation Mr Peng Qinghua, Addressing The Crowd During The Launch Of JAP At Kasarani Sports Centre On September 10, 2016.

Well, wresting power from the opposition and securing the governments interests in these Counties is not at all bad by any measure. The only problem is the entry of the dragon whose insatiable appetite for cheap raw materials knows no bounds and they, The China Communist Party, CCP, signed an MOU with the Jubilee Alliance Party (JAP) of Kenya and the agreement was signed by Meru senator Kiraitu Murungi and CCP’s Peng Hua in Nairobi.

Jubilee Alliance Party, China Communist Party Partnership

Mr. Shui was quoted saying, “The Communist Party of China would like to extend their message of congratulation to Jubilee Party and its leadership as they launch this new party,” read a statement from the party’s Central Committee. CBC is ready to work with Jubilee Party to realize change and boost experience in governance to promote China-Kenya partnership.”

The party further restated the milestone they had made during Kenyatta’s tenure and expressed willingness to better the relationship. “Under Jubilee coalition and the leadership of President Kenyatta, a lot of development has been realized.” Jubilee officials are also set to meet the Communist party officials in China on a bench-marking mission.

Map Of Kenya Mineral Occurrences

We can only hope that this partnership will not turn into an exercise, in which the country’s vast resources are mortgaged off to a supposed partner whose real interest is just to exploit our eagerness for easy money to bleed our resources dry. The Kwale governor has alluded to the same in an article which appeared on the Daily Nation where he admitted that the real battle in Kwale is for the minerals more specifically Niobium and Rare Earth minerals worth at least 10 Trillion Kenya Shillings.

Going from prior experience the Chinese are not averse to bribing officials in foreign countries in order to have their way and we hope that this was not the case with the defectors. The Chinese have perfected the Angola Model. What we can be sure off is that the likely scenario is the one of, “using money to make money”. After all there is never anything like a free lunch.

Map Of Oil Concessions(Blocks) in Kenya

In May of 2016 Kenya identified 17 new oil blocks which it plans to auction in 2017, they are updated in this map . According to Kenya's National Oil Corporation which is under the Ministry of Oil and Petroleum Kenya currently has 63 Oil Exploration Blocks. They are spread out over four basins.

  • Anza Basin
  • Mandera Basin
  • Tertiary Rift Basin
  • Lamu Basin

About The Author

  Marcel Masaga  is a Full Stack Engineer, Writer and Big Data Evangelist. He is also a farmer par excellence.