The highest duty of the writer is to remain true to himself and let the chips fall where they may. In serving his vision of the truth the artist best serves his nation.
John F. Kennedy
University of Ghana (UG), for the second time getting close to Elections, has done well. Established in 1948 as University College of the Gold coast, UG has lead series of academic, political, economic research and studies spanning from several disciplines. In 2016, led by Isaac Owusu-Mensah, a lecturer, in a research, indicated that, the NPP (then opposition party) would record 49.9%; whereas the incumbent, then NDC, would record 39.9% in the 2016 General Elections. It however concluded that the NPP stood a chance of winning the 2016 General Elections. Contrary to their findings but true to their conclusions, though obvious, there was no run-off. The opposition won with a whooping 1million+ one touch margin.
A recent research undertaken by this same Political Science Department of the University of Ghana has this time around sent chills down the spines of over 260 Members of Parliament in Ghana.
The research is in part a wake up call especially for lawmakers who toy with the future and destiny of the people in their constituencies.
Nonetheless, this is a half-baked research. And the arguments underlying this paper caution researchers and professionals of the need to do a professional job void of bias that invariably leads to inconclusiveness in their findings.
In 2016, there were 15,712,499 people on Ghana’s electoral register. Out of this, 10,781,609 in 271 of 275 constituencies voted on Election Day December 7, 2016, representing 68.62% of registered voters.
According to the UG researchers, their findings disclose the fortunes/fate of the 275 Members of Parliament in the upcoming 2020 General Elections. Hypothetically, the choice of 10,781,609million (number of votes cast, 2016), all other things being equal, has been ascertained.
Their research methodology employs the simple random sampling using the HAT method, where different hats (red, green etc) perspectives, beliefs, reasoning, feelings are considered in making decisions and drawing conclusions. A wake up call to lawmakers to be on their toes–but, insufficient enough, looking at the parameters and modalities, to conclude as to whether a particular member of Parliament stands the chance to retain his/her seat.
The research used Electoral Areas (non-political setting) in deriving results for Constituencies (political setting). Ghana has 705+ Electoral Areas (non-political setting) but over 1,500+ (zonal areas)-political setting). The research selected 5 Electoral Areas (EAs) from each of the 275 constituencies with the exception of Ayawaso North and East, where 3 & 4 EAs were selected. The research engaged 100 Ghanaian respondents per constituency; 27,500 Ghanaian respondents from 1,375 EAs nationwide. That is, 20 respondents/interviewees from each of the selected 5 EAs. And 100 by 275 constituencies.
The researchers did not realize or take into account the confusion of mixing a political setting with a non-political setting (district and constituency as provided for by the Local Government Act). It failed to act on the right geo-political environment. The research failed in the way the SRS method was undertaken (merely identifying respondents). In that, not all respondents qualify, amongst other factors, to vote both in primaries or General Elections. The research based its findings on mere questions amongst other indicators.
Robert Krejcie and Daryle Morgan (1970) in their Sample size determination table (SSDT), suggests a sample size (s) of 10 for a population (N) of 10; 80 for 100; 132 for 200; 384 for a population size of 1,000,000 and above.
All other things being equal, the population (voter size) of a district is larger than a constituency. Example: Take Ga East (95,000+) as compared to Dome Kwabenya (995+), a constituency under Ga East. To have a clear picture of the situation on the ground, it is sufficient to target the 995+ instead of a random sampling in the 95,000+ where the researchers never bothered to target the 995+ kingmakers.
The UG research used an insignificant sample size of 27,500 (respondents) from the 275 constituencies nationwide.
We could choose the SSDT however this method in itself has been widely criticized and condemned. Example, according to the SSDT a sample size of 384 is needed for a population size of 1million or more. In other words, for a set of two different population of 19million and 3million. We still need a sample size of 384 for each of the population according to Krejcie and Morgan, regardIess of data strength, number, dynamics and deviation. In determining the sample size in factor analysis inter alia, scholars like (Arrindell & van der Ende, 1985, Nunnally, 1978, Everitt, 1975, David Garson, 2008; Everitt,1975) approve of the rule of 10. It would have been prudent for University of Ghana to have employed the rule of 10 or a simple 10% mark of each specific population under study.
For the rule of 10, we select 7,500 sample size (respondents) in each of the 258 MMDAs. A fair representation of the population, all other things being equal.
So in totality, we are looking at a sample size of not less than 1,900,000. That’s for a population size of 19,350,000 (258 district * 75,000) or a sample size of 100 (that is 100 delegates out of the 1,000 delegates) for each Constituency. Using Dome Kwabenya, the second most densely populated voter size as a benchmark.
The research failed to distinguish between set parameters and modalities for districts, municipalities and metropolitans but treated all MMDAs equally. According to the Electoral Commission, a district has a population size of 75,000; 95,000 for Municipality and 250,000+ for Metropolitan.
Ghana has over 216 + 38 MMDAs (Metropolitan, Municipal and Districts Assemblies). Let this figure represent 254+ districts, all other things being equal. Per definition, 75,000 * 254 = 19,050,000million (population of certified Ghanaians in the MMDAs), all other things being equal.
The research also took into consideration the newly created regions. What, the researchers failed to realize was that, the newly created regions currently have assumed different dynamics and, the researchers needed to have limited their study to the old ten. Such an inclusiveness destroys the fairness of a set of responses from a given population.
Below are the names that appear on the top 10 (Members of Parliament) in terms of performance (people’s perception)
1. Collins Dauda, Asutifi South, NDC
2. Alexander Afenyo-Markin, Efutu, NPP
3. Bryan Acheampong, Abetifi, NPP
4. Mrs Abena Osei-Asare, Atiwa East
5. Emmanuel Armah Kofi Buah, Ellembelle, NDC
6. Samuel Okudzeto Ablakwa, North Tongu, NDC
7. Mr Kwame Agbodza, Adaklu, NDC
8. Kennedy Osei Nyarko, Akim Swedru, NPP
9. Ms Joycelyn Tetteh, North Dayi, NDC
10. Fiifi Kwetey, Ketu South
1. Builsi South
2. Afram Plains South
5. Krachi Nchumuru
8. Offinso South
9. West Akim (Lower)
As already aforementioned, University of Ghana has given out a dossier to serve as a wake up call. However, it wouldn’t be prudent or fair to go by their findings. There is however insufficient data in their work to draw a 75%/95% conclusion on the fate of MPs in 2020. It is also truistic that, there will be surprises in the upcoming 2020 General Elections if lawmakers fail to change strategies.
Case Study: Adentan Municipality or Adenta Constituency. Every Constituency has polling station delegates. This figure is the sole determinant of the fate of the Member of Parliament in the most important stage (stage 1) referreed to as primaries.
*Do you want your MP to contest 2020?*
To get a perfect picture to the question one needs to ask the delegates. This is because there is no way members of the opposition party would respond in the affirmative, all other things being equal. The research was however not fair to the 275 lawmakers. It was not of protective of lawmakers who had showed some exemplary leadership skills yet are faced with unique dynamics.
There is a clear distinction between Adentan Municipality and Adenta Constituency. The former being a Municipal (non-political setting) and the latter being a Constituency (political setting).
Why University of Ghana Got it Wrong amongst other things include;
1. Poor Data Strength:
2. Missed Target
3. Missed Geo-political Setting
4. Insignificant Sample Size
5. Unrealistic hypotheses
6. Lack of fairness
Woe betides those who rejoice and pride themselves with a seeming good report borne out of mediocrity. The Lawmakers who feel that they have been graded creditably and therefore have a course to rejoice have nothing to rejoice over. Likewise the lawmakers who feel sad or indifferent have no reason to be sad. Who is free? No one is free. UG has done well but has got it all wrong with its half-baked research. Notwithstanding, the incumbent party, NPP, will retain the seat in both 2020 & 2024.
Dome Kwabenya Constituency
Economist & future Financier, NPP
Member, Nana One Touch
Member, National Communicators Cartel