How TMG quick count verified 2015 presidential polls results

SINCE the advent of Nigeria’s most recent democratic dispensation in 1999, one feature of the electoral terrain that has caught the attention of scholars and pro democracy activists is the volatile and disruptive nature of electoral competition. With the political elite always ready to go for the broke, not minding the health of the polity, there has been a tendency for political forces to do all within their powers to illegally subvert the due processes of electoral completion.

The result of such desperate pursuit for political power, are the flawed elections in which the voices of the electorate were entirely snuffed out of the democratic process. In turn, the governments that emerged out of those electoral victories procured by outright rigging and a wilful manipulation of the process, suffered a crisis of legitimacy that in turn impacted negatively on governance.

To find a way out of this quagmire, stakeholders explored the idea of putting an eye on the electoral process in such a manner that the processes and outcomes of elections could be independently verified. In other words, this intervention focused on addressing the trust deficit that usually characterizes the electoral environment. So by independently verifying the processes and outcomes through a robust observation effort, civil society groups could help deepen confidence in the process, detect and deter electoral fraud, and nip in the bud conflicts that are triggered by the contestations around elections.

This is exactly what the Transition Monitoring Group (TMG) has just achieved with its deployment of the globally tested Quick Count methodology for the observation of the keenly contested presidential election of March 28. The Quick Count methodology, also known as Parallel Vote Tabulation (PVT) was first used in the Philippines in 1986. That was the method the people of that Asian country used to get their voices heard in the electoral process. In that particular effort, they managed to stave off the burdensome dictatorship of Ferdinand Marcos, which was bent on manipulating the electoral process back then.

As a result of its efficacy in accurately and independently verifying the results of elections, Quick Counts have been variously used in testing the integrity of elections across the world. The Quick Count has thus been used in Ghana, Malawi, Zimbabwe, Zambia, Kenya and Tunisia. The most recent Quick Count is the one that was successfully used to verify the outcome of the March 28 Presidential election in Nigeria. It had earlier been used, also in Nigeria in 2011 to observe the Presidential election.

At the heart of a Quick Count is the need for highly trained non-partisan observers to deploy to a representative random sample of polling units in the electoral terrain. Observers, respond to a list of checklist questions on the various processes, as well as the official results as announced in the sample polling units. The data is moved to a National Information Centre via coded text messaging, where trained statisticians then analyze the data to produce estimates that are compared with the official results. If the official results fall within the margin of error of the Quick Count estimates, it implies they are accurate and should be accepted by all stakeholders in the electoral process.

However, if the official results veer of the Quick Count margin of error or its estimates, then something untoward has happened in the process.

On March 28 therefore TMG undertook the Quick Count to provide voters, candidates, political parties, and INEC with independent information about whether the official results for the presidential election truly reflect the ballots cast at polling units.

Critical to the success of a Quick Count is the profile and non-partisan disposition of the organization conducting it. TMG, which conducted the just-concluded Quick Count is Nigeria’s premier citizen observer group.  TMG observed the transition from military   rule and all subsequent national elections. As a coalition of over 400 civil society organizations based in all 36 states and the FCT, TMG looked the part in terms of a non partisan grassroots based coalition of Nigerian CSOs interested in ensuring the supremacy of the voices of the Nigerian people in the electoral process.

TMG prides itself as being driven by data, for all Nigerians, but beholden to none. TMG reports are based on data from our carefully selected, trained and accredited citizen observers. TMG volunteers are drawn from its membership and every corner of the country and include: men and women, the young and the elderly, Christians and Muslims, and Northerners and Southerners. TMG is an independent and strictly nonpartisan organization with no affiliation to any political party or candidate.

The Quick Count for the 2015 Presidential Election involved deploying citizen observers to a representative random sample of 1,507 polling units at which there are 849,460 registered voters. There were sampled polling units in every geopolitical zone, every state and 774 local government area (LGA) of the country. The sample is carefully constructed by a trained statistician to ensure that every geopolitical zone, state and LGA of the country is included proportionally.

The number of sampled polling units for each geopolitical zone, state and LGA is therefore based on the percentage of polling units and registered voters in that geopolitical zone, state or LGA. For example, South West geopolitical zone has 20.6 percent of all polling units in the country (24,683 of 119,979) and 20.6 percent of polling units in the Quick Count sample (310 of 1,507) are in South West.

Similarly North West has 25.6 percent of all registered voters (17,620,728 of 68,845,062) and 25.5 percent of the registered voters in the Quick Count sample (217,011 of 849,460) are in North West.

The observers deployed to sampled polling units at 7:00 am observed setup of polling units, accreditation of voters, voting, counting, announcement and posting of results.

Throughout Election Day TMG’s citizen observers sent in seven text messages providing structured data on the quality of the process of the election as well as the official results for their polling unit as announced by the poll official.

TMG’s Quick Count estimates and INEC’s official results clearly showed that Muhammadu Buhari of the All Progressives Congress (APC) received the most votes and Goodluck Jonathan of the People’s Democratic Party (PDP) received the second most votes and that all other candidates only received a very small number of votes. TMG’s Quick Count also confirmed that the APC and PDP presidential candidates attained “not less than one-quarter of the votes cast … in each of at least two-thirds of all the States in the federation and the Federal Capital Territory” as required by the Nigerian Constitution.

According to INEC, APC received 53.9 percent of the vote while TMG’s Quick Count estimates that APC received 59.4% (+/-2.3%) of the presidential vote. Similarly, INEC’s figures show PDP received 45.0% of the vote while TMG’s Count estimates that PDP received 39.2% (+/-2.2%) of the presidential vote. Therefore TMG confidently verifies the accuracy of the official results for the presidential election based on the independent collection of official results from sampled polling units.

However, INEC’s official results for APC and PDP did not fall within the margin of error of the TMG’s Quick Count estimates.

TMG Chairman, Ibrahim M. Zikirullahi explained this reality in the Verification Statement when he pointed out that “this does not invalidate the official results as announced by INEC as both INEC and TMG’s Quick Count show the same national voting pattern and both clearly show APC receiving the largest number of votes. In addition, for 31 of the 36 states and the FCT INEC’s official results are consistent with TMG’s Quick Count estimates.”

The Quick Count is not an exit poll. No voter is asked for whom he/she voted. The Quick Count uses the official results as announced and posted at polling units. INEC poll officials at polling units are the ones who count the ballot papers. TMG citizen observers do not count the ballot papers themselves. Rather they observe the counting process and report on the conduct of that process and the official result for polling units as announced by the poll officials.

Since the official results collectedby  TMG’s  observers  come  from  a  representative  random   sample of polling units, by adding those results together TMG is able to estimate the national results for the presidential election. The estimate is produced within a range, known as the margin of error, based on well-established statistical principles.  If INEC’s official results are consistent with TMG’s Quick Count estimate, then the official results reflect the ballots cast at polling units.

The point of conflict between the Quick Count estimates for the 2015 Presidential elections and the INEC official results is with the turn out in the South South. TMG’s Quick Count estimates strongly suggest that turnout was inflated during the collation process in the South South. The official turnout figures for five geopolitical zones are consistent with TMG’s Quick Count estimates, but the official turnout figure for South South is 55.9 percent while TMG’s estimated turnout for South South is 40.6 percent (+/-4.5) percent. Thus official turnout in South South was likely inflated during the collation process by at least 10.8 percent. TMG’s Quick Count show that in Akwa Ibom, Bayelsa, Delta, and Rivers states the official turnout figures were inflated. For Cross River and Edo the INEC official turnout figures are consistent with TMG’s Quick Count estimates.

These findings have important implications for the governorship elections in the States where conflicts exist between TMG Quick Count estimates and the official INEC result. While the message around these findings is not meant to taint the mood of celebration that has engulfed the country as a result of the triumph of democracy on March 28, it is meant to remind all stakeholders that there is a lot to be done in ensuring the votes count. It is also a call on INEC to look deeply into the collation process in the areas identified to ensure no electoral heist is perpetrated in future polls.

In the end too, the findings from the deployment of the Quick Count methodology would be useful in the area of reforming the electoral process. In its coming final report, which would catalogue all the areas that should be looked into, the Quick Count would prove instrumental in truly ensuring that the political actors and other stakeholders do their best to “let the votes count.”