By Maxwell Sibanda
In the aftermath of the announcement of the presidential results, independent electoral watchdog, the Zimbabwe Election Support Network (Zesn) publicised its sample based observation (SBO) results projection.
The SBO results were consistent with the official results as announced by the Zimbabwe Electoral Commission (Zec) with a margin of error of plus or minus two percent.
A lot of questions have been raised about the accuracy and effectiveness of the SBO.
The Daily News sat down with the Zesn Chairperson Andrew Makoni to discuss the SBO methodology and the projections of the SBO for the recently held 2018 Presidential elections and address some of the queries that have been aired by some members of the public.
Below are excerpts from the interview:
Q: Zesn released it 2018 Presidential election projections based on its Sample-Based Observation (SBO) can you explain what this entails?
A: Zesn trained and deployed over 6500 election observers to every ward and constituency in Zimbabwe. Out of these, 750 were Sample Based Observers.
By way of definition, SBO is an advanced and proven election monitoring methodology employed around the world that allows non-partisan citizen observers to systematically assess the quality of opening of polls, voting, closing and counting processes.
Using this methodology, an SBO can accurately project a range of the possible outcomes of the presidential elections, serving as an independent check on the official election result. SBOs can confirm the accuracy of official results, or expose manipulation during tabulation process if it occurs. SBOs use direct observations by trained observers and do not ask voters’ for opinions.
Q: How is an SBO different from traditional Election Day observation?
A: For the 2018 harmonised election, Zesn used both the SBO and general observation methodologies as part of election observation.
The major difference between the two methodologies is that under the SBO, observers are deployed to a nationally representative sample of polling stations carefully selected by expert statisticians.
Observers make use of mobile phone technology to transmit observation data quickly and completely to a central database enabling observers to provide the public with more accurate and comprehensive information about the conduct of an election for the entire country.
The findings of the SBO are then released based on sound and systematic data and not anecdotal information or opinions.
Q: Why did Zesn conduct an SBO?
A: Elections in Zimbabwe have since the early 2000s become highly contested with disputed electoral outcomes sometimes leading to injury and death of political supporters.
In this context, Zesn decided to conduct an SBO in order to remove uncertainty over presidential election results by independently validating or invalidating the official presidential election results.
The objective was to assess the results management process and provide an alternative view and thereby reduce any potential political turmoil that might have arisen.
Q: In your statement, you mentioned that there was a margin of error, can you explain what this is and to what extent can it compromise your data?
A: While statistics are a proven tool to draw accurate conclusions about a larger population, all sample-based statistics have a margin of error.
When a sample is drawn and stratified well, distributions should reflect the same distributions in the population at large.
The SBO sample demonstrates a sound distribution of the rural and urban divide and the distribution of polling stations and registered voters across the country. In simple terms, the margin of error can be thought of as a window on either side of the estimate in which there is uncertainty.
Q: What methodology did you use during SBO and the criteria used for selecting the polling stations?
A: Strictly applying random selection methods and by sampling with probability proportionate to population size (PPPS), 750 polling stations were chosen as a national representative of approximately 10 800 polling stations at random.
This approach removes bias since every polling station had the exact same chance of being selected.
The SBO sample is stratified so it covers every district, province and constituency in the country, and the distribution of observers is proportionate to the percentage of polling stations and registered voters in each province. This approach makes it representative.
Q: The neutrality and credibility of Zesn has been questioned following the SBO projections, what is your comment on those questioning your credibility and neutrality?
A: Zesn is a Network of non-governmental organisations with a long history of promoting peaceful democratic electoral processes in Zimbabwe. In terms of recruitment, there is a strict requirement for recruiting observers, including that they cannot be an official member of any political party nor publicly support one.
The observer must sign and adhere to a strict code of conduct and pledge of nonpartisanship which if not adhered to results in the observers being recalled. These observers are adequately trained to ensure that they accurately observe and report their findings.
Q: Some allege that Zesn endorsed the presidential election results, what’s your take on that?
A: Zesn as a non-partisan and impartial organization did not endorse any election result; in addition no voters were asked whom they voted for.
The SBO is not an exit poll and findings of the SBO simply revealed that the official results as announced by Zec are consistent with our projections, which some may call validation. Since the SBO is based on a sample of polling stations it produces estimated ranges within which the results announced by an electoral management body should fall.
SBO projections on the presidential results were misinterpreted as an endorsement of the Zec results and some took the SBO as counter to claims by some of the contesting candidates that the election results had been tampered with.
It is important to note that the Zesn SBO could not confirm whether or not there should have been a run-off and did not pass a judgement on other electoral processes as we are still observing the post-election period.
In addition, it is important in future for the media to acquaint themselves well with the numerical and sampling methodologies and seek expert views on these to enable them to report accurately.
Q: Where else have SBOs been done?
A: To date, SBOs have been used as an observation methodology in almost 50 countries worldwide. In Africa, SBOs have so far been used in 12 countries such as Cote d’Ivoire (2016), Malawi (2009 and 2014), Zambia (2008, 2011, 2015 and 2016), Ghana (2008, 2012 and 2016), Uganda (2011), and Nigeria (2011, 2012 and 2015) and Zimbabwe (2008 and 2013) among others. In all these instances, the SBO projections have validated or invalidated the official presidential results.
Q: Are there any instances where the SBO projections have been wrong?
A: We cannot say that there are instances where the SBO methodology has been wrong; the question should be whether there have been instances where there has been a mismatch.
In 2015, a mismatch occurred in Nigeria. The SBO detected inflated votes for Goodluck Jonathan in the south of the country.
The projected range for votes for Goodluck Jonathan was between 37 percent 41,4 percent and yet the electoral commission announced 45 percent, which meant that what was announced by the electoral commission showed slightly more votes for Goodluck Jonathan and slightly less votes for Buhari.
The group alerted the Commission of the discrepancy and they agreed that the results in the south had been tempered with.
However, since it didn’t ultimately change the outcome, the electoral commission let it slide. The SBO tends to validate or invalidate results and it has been successful in doing so.
Q: What are the shortcomings of the SBO and how did you overcome these?
A: SBO can only evaluate the process and information collected by the standardized SBO observer form on Election Day.
SBO cannot tell you a voter’s motivation for voting; hence it can only provide a nationwide assessment if a very high response rate is achieved.
Nationwide SBO has limited ability to assess sub-national trends; hence the data is only representative at national level. SBO cannot directly assess quality of pre-election day processes such as the legal framework, voter registration or the campaign.
Q: The Constitutional Court recently dismissed the electoral petition by the MDC Alliance, what is your reaction to this judgement?
A: The Constitutional Court is the final court of appeal for all matters or disputes, including a presidential election and its decisions are binding on all other courts in Zimbabwe.
The decisions of this court may please some people, while at the same time displeasing others.
As Zesn, we respect the rule of law and call upon the people of Zimbabwe to respect the court outcome – even when you may disagree with it – and continue to observe as well as remain united, politically tolerant of each other, peaceful and tranquil.
Q: From your general assessment, were the elections free, fair and credible?
A: As I mentioned earlier, Zesn and other election observer missions are still observing the post-election period which is also one of the key factors we use to ascertain if the elections have indeed been free and fair.
Prior to the holding of the 2018 harmonized elections, we noted that the pre-election period was generally peaceful with isolated cases of inter and intra-party violence, subtle intimidation of the electorate as well as free campaigns by all political parties in most parts of the country.
On Election Day, almost all polling stations opened and closed on time, they all had the polling materials and all presiding officers, electoral officers, observes and polling agents were present.
Before the announcement of results we noted disturbing events namely the death of six people and injuries. The heavy handed response by the security offices was widely condemned by election missions including Zesn.
On the other hand, Zec did not address concerns raised by stakeholders with regards to the design, printing, and dispatch of ballot papers and did not permit meaningful observation of the production of ballot papers or testing of indelible ink.
In addition, of concern is that once proclamation of the election date was made, the Zec failed to meet its obligation to make available an analyzable copy of the voters’ roll to electoral stakeholders before the sitting of the Nomination Court.
We also noted that the 2018 voters roll that was made available by Zec in June is an improvement from 2013; however urban registration lagged behind rural registration with 73 percent of people in urban areas registered to vote compared to 82 percent in rural areas.
State media and other State resources were abused throughout the pre-election period to the advantage the ruling party.
Zesn will issue an election report that will give a complete assessment of all key electoral processes for the 2018 harmonised election.
We remain committed to providing objective, credible and unbiased electoral information.