REPORT ON THE EVENTS AND OPINIONS LEADING UP TO THE 2019 PRESIDENTIAL ELECTIONS

Introduction

Nigeria, the most populous country in Africa, will hold its general elections on the 16th of February, 2019 — four years after the last general elections that saw the main opposition candidate, Muhammadu Buhari defeat the then incumbent, Goodluck Jonathan.

Now President, Muhammadu Buhari will be vying for a second term in office. However, he has about 70 other candidates also challenging for the highest office in the land. Among those candidates is his main opposition and former Vice President of Nigeria 1999 -2007 — Alhaji Atiku Abubakar. Both men have a large following and either one of them is extremely likely to be President come May 29th, 2019.

Project Examine Nigeria (PEN) is a project initiated to understand the sentiments and opinions of Nigerians on Twitter leading up to the election date. This is in the area of sentiment analysis — an opinion mining methodology that has been used in past elections (such as the 2016 US Presidential elections) to predict the winner of elections based on the sentiments of a sample of the population (Twitter users). A more detailed description of the methodology and implementation for the PEN project can be found in the footnote below[1]. The main goal was to examine the opinions and sentiments of tweets centered mainly around the two candidates: basically, what were Nigerians saying about both of them on twitter? We did this for a period of 28 days (11th of January — 8th February) leading up to the election. Using the Twitter API[2], we collected tweets for each day and classified them as tweets with either a positive, neutral or negative connotation using rule-based heuristics and a machine learning classifier.

In this report, we present the results of the work done during the one-month period as well as extra analysis related to the 2019 elections. At the end of this report, we make predictions about the election based on the results of our analysis and a look back at patterns and trends from past elections.

Result Analysis

The data analyzed illustrates the percentage of positive, negative and neutral sentiments towards the two major presidential candidates for the 2019 Nigerian presidential elections to be held on Saturday, the 16th of February. Data collection and analysis were done over a 28-day period — 11th January to 7th February 2019.

NB: Sample size varies over the 28-day period.

WEEK 1 HIGHLIGHTS

11/01/2019: INTRODUCTION

From tweets analyzed, we found that there were mixed reactions towards both candidates, with the talk of the elections still in an infant stage. Over 30% of the tweets for both candidates were neutral — favoring neither Atiku nor Buhari. For Buhari, the following words contributed significantly to the sentiments of tweets analyzed: “Failed”, “Scam”, “Corruption”, “Fraud”, “Integrity”, “Hardship” & “Poverty”. Tweets with ‘Atiku’ as the subject mainly consisted of the following: “Criminal”, “Corrupt”, “Economy”, “Looters”, “Questionable”, “Alternative” & “Obasanjo”.

Buhari: Positive: 20%, Negative: 42%, Neutral: 38%

Atiku: Positive: 24%, Negative: 42%, Neutral: 34%

12/01/2019: PLAN TO SUSPEND WALTER ONNOGHEN — THE CHIEF JUSTICE OF NIGERIA

News of President Buhari’s plot to suspend Walter Onnoghen — the now suspended Chief Justice of Nigeria (CJN) was not embraced by the majority of Nigerians, with some viewing it as a desperate attempt to get the CJN ‘out of the way’ and appoint an acting CJN who he views as pliant and will therefore be made to superintend over election petitions in case his well laid out “rigging plan” fails. This took Buhari’s negative sentiments up to 46%. Meanwhile Atiku and PDP’s presidential rally in Plateau state accompanied with the negative opinions for Buhari took Atiku’s positive sentiments up to 29%.

13/01/2019: EFFECTS OF PLAN TO SUSPEND WALTER ONNOGHEN

With the agenda to suspend the Chief Justice of Nigeria still very much in the air, Buhari’s negative sentiments got even higher — 47%. Words like “Remove”, “Integrity”, “Corrupt”, “Sense” in tweets analyzed contributed significantly to this result.

Buhari: Positive: 19%, Negative: 47%, Neutral: 34%

Atiku: Positive: 27%, Negative: 36%, Neutral: 37%

16/01/2019: BUHARI ON #NGTHECANDIDATES

Buhari’s appearance on #NGTheCandidates with the ever vibrant Kadaria Ahmed took a toll on his numbers, with Nigerians particularly unimpressed with the manner in which he answered the questions thrown at him, some even going as far as labelling him “Clueless”. Negative sentiments after the debate reached a staggering 50%. Atiku recorded his best numbers as a result of this, positive sentiments went up to 30%

Buhari: Positive: 20%, Negative: 50%, Neutral: 30%

Atiku: Positive: 30%, Negative: 37%, Neutral: 33%

18/01/2019: ATIKU’S BANK PHB SCANDAL

News that Atiku was involved and benefited from the alleged slush funds that led to the eventual collapse of Platinum Habib Bank (Bank PHB) in August 2011 surfaced on the internet, which resulted in a lot of negative reactions from Nigerians on twitter, Atiku recorded his highest percentage of negative sentiments (42%) and lowest percentage of positive tweets (23%) as a result of these allegations. Some of the words used on twitter to express these sentiments include: “Corruption”, “Looted”, “AtikuNeverAgain”, “Disgraceful”. Meanwhile, Buhari and APC had their presidential rally in Kaduna which resulted in positive feedback on twitter and resulted in the first day that Buhari recorded a lesser percentage of negative sentiments than Atiku. Sai Baba?

Buhari: Positive: 23%, Negative: 38%, Neutral: 39%

Atiku: Positive: 23%, Negative: 42%, Neutral: 35%

WEEK 2 HIGHLIGHTS

19/01/2019: PRESIDENTIAL DEBATE FOR ALL CANDIDATES

Buhari and Atiku’s decision to boycott the presidential debate resulted in a massive backlash from Nigerians on twitter. Nigerians were not having any of that, Atiku’s issued statement explaining why he didn’t show up for the debate was not enough to calm the waters. Buhari’s ‘No Show’ and Atiku’s ‘Show Off’ eventually resulted in abysmal and identical numbers for both candidates, with the majority of tweets having negative sentiments.

Buhari: Positive: 17%, Negative: 51%, Neutral: 32%

Atiku: Positive: 17%, Negative: 49%, Neutral: 34%

20/01/2019: PRESIDENTIAL DEBATE FOR ALL CANDIDATES

Atiku’s excuse as to why he was absent from the debate seemed to anger Nigerians on twitter even more, as it surfaced that he was indeed present, but left at the last minute after finding out Buhari was not going to be participating in the debate. This caused a slight increase in negative sentiments from tweets analyzed on 19/01/2019–50%.

24/01/2019: ATIKU & APC’S PRESIDENTIAL RALLY IN KADUNA

Results for both candidates maintained a similar trend since the presidential debate held on 19/01/2019. However, Atiku’s presidential rally in Kaduna State came at just the right time and resulted in a slight drop in the percentage of negative sentiments and more neutral sentiments were recorded from tweets analyzed on this day. Buhari’s results maintained the same trend.

Buhari: Positive: 19%, Negative: 44%, Neutral: 37%

Atiku: Positive: 26%, Negative: 34%, Neutral: 40%

25/01/2019: BUHARI SUSPENDS THE CHIEF JUSTICE OF NIGERIA

Buhari’s decision to suspend Justice Walter Onnoghen after recommendations from the Code of Conduct Tribunal (CTT) pending the completion of his trial over allegations of wrongful declaration of assets caused a meltdown on twitter, majority of Nigerians who initially saw this as a desperate attempt to have a familiar face in the form of — Tanko Mohammed (Now acting Chief Justice General of Nigeria) in this position definitely did not have a different opinion. Emotions and opinions were made clear on twitter and this resulted in the most negative sentiments (53%) recorded for Buhari. “Tyrant” was definitely the go-to word for most twitter users.

Buhari: Positive: 14%, Negative: 53%, Neutral: 33%

Atiku: Positive: 30%, Negative: 31%, Neutral: 39%

WEEK 3 HIGHLIGHTS

30/01/2019: ATIKU ON #NGTHECANDIDATES

Atiku’s appearance on the demanding interview by Kadaria Ahmed received mixed reactions by Nigerians on twitter, some found the interview ‘Unfair’ and accused the host — Kadaria Ahmed of targeting PDP and asking more grueling questions compared to the ones face by President Buhari & Osinbajo on the 16th of January. The different opinions documented can be seen from the percentage of positive and negative sentiments of tweets analyzed for Atiku — 26% to 35% respectively.

31/01/2019: BUHARI AND PDP PRESIDENTIAL RALLY IN KANO/GANDUJE ALLEGATIONS

Tweets analyzed for Buhari have maintained the same trend since his decision to suspend Justice Walter Onnoghen on the 25th of January but during the routine APC presidential rally in Kano State, the Ganduje saga was bound to resurface and resurface it did. The 5 million votes promised to Buhari by Abdullahi Ganduje, leaked videos of the bribe collected by the governor and other related theories linking the bribe to the votes were discussed non-stop. Results collected for Buhari yet again, maintained the same trend.

Buhari: Positive: 24%, Negative: 42%, Neutral: 34%

Atiku: Positive: 26%, Negative: 35%, Neutral: 39%

WEEK 4 HIGHLIGHTS

In the final week of our analysis, there was an upward trend in positive sentiments for both candidates. However, Atiku’s positive numbers were consistently higher than the negative. On 4 days out of 7, his positive numbers outweighed the negative and were above 30% percent (when neutral tweets were included). We reported the weekly average of positive sentiments for both candidates (excluding the neutral tweets), and visualized the results.

Endorsements and their Effects

Political endorsements are quite common in the Nigerian electoral space. Different groups and major political players who have different political, religious and tribal interests pitch their tents with candidates who they think will meet their interests when they come into power. One political bigwig who has had a say in all elections since 2007 is former President, Olusegun Obasanjo. Every candidate he endorsed since 2007, has won the Presidential election including Muhammadu Buhari, who was a beneficiary of his endorsement in 2015.

This year, Obasanjo has pitched tent with his former 2nd in command, Atiku Abubakar, despite publicly criticizing him on previous occasions for issues related to corruption. In a closed door meeting on October 11 2018, Obasanjo said ‘he has forgiven Atiku’ and endorsed him publicly. Akin to that, Obasanjo held a press conference and granted an interview on BBC (January 20th and 21st respectively) where he heavily criticized the Buhari administration and the President himself for failure to secure the country and improve the lives of citizens.

Other socio-cultural groups including Ohaneze Ndigbo, Afenifere, PANDEF, Middle Belt Forum etc. also endorsed Atiku. There might not be much meaning to read into the endorsements by the socio-cultural groups as it’s difficult to gauge how much influence they have on voters, however it is interesting to note that Obasanjo’s endorsement created a lot of back and forth chatter in the Nigerian media (including social media) as well as the major political parties. It is safe to say he still holds some influence in the Nigerian electoral space.

Prof. Wole Soyinka, a 1986 Nobel Prize winner and respected poet/author, who has also been a strong critic of successive Nigerian governments especially during the military rule, endorsed Prof. Kingsley Moghalu — candidate of the Young Progressive Party thereby rejecting the presidential frontrunners and urging Nigerians to use their votes to say “Evil Spirit, Leave me Be!” in line with the meaning of the name ‘Moghalu’[3].

The 3rd Force Quagmire

At early stages of the election season, it would seem Nigerians were hoping for an alternative to the two major political parties, APC and PDP. There was a clamor for a 3rd force, one that will displace the ‘political establishment’ and usher in a new, fresh, young and vibrant style of leadership. Many new parties were formed, debut candidates began to emerge and it seemed many citizens were looking forward to vote for an alternative. The 3rd force plan seemed to be gaining momentum which cumulated to the emergence of PACT.

PACT which means Presidential Aspirants Coming Together, a coalition of political parties, was meant to produce a consensus candidate who will fly the flag on behalf of other aspirants and parties under PACT to challenge the main two parties. It was gathered that 18 presidential aspirants agreed to elect a consensus candidate, but some opted out of the process few hours before the election, leaving 11 aspirants to participate. The election was monitored and observed by Oby Ezekwezili, Nigeria’s former Minister of Education, who described the process as transparent and credible. Fela Durotoye, candidate of the Alliance for New Nigeria (ANN) emerged the winner of the elections. However, another candidate, Kingsley Moghalu rejected the outcome of the election and to sum it all up, the PACT arrangement failed. The observer of that PACT electoral process, Oby Ezekwesili later declared her intention to join the Presidential race on the 7th of October with the hope to pull down the stronghold of ‘two failed dominant political parties, APC and PDP’. Her declaration seemed to spark a lot of optimism and joy as many voters indicated support for her ambition.

Kingsley Moghalu, Fela Durotoye, Omoyele Sowore and Oby Ezekwesili failed to form a viable coalition but remained as strongest contenders likely capable of dismantling the stronghold of APC and PDP. On the 19th of January 2019, there was a Presidential debate in which the candidates of APC and PDP failed to attend. Unfortunately, Omoyele Sowore wasn’t invited for this debate which left open the possibility for the other 3 candidates to sell themselves to Nigerians. A result of the analysis done on that debate shows that Nigerians were highly disappointed that Buhari and Atiku didn’t turn up, but Kingsley Moghalu was the most preferred candidate out of the three.

On the 24th of January, Oby Ezekwesili pulled out of the race citing ‘differences in values and vision’ with the party she was running under — ACPN. This was received with a lot of disappointment from Oby’s supporters but backlash from many others who firmly believe no other party apart from APC or PDP have the structure and influence to win the 2019 Presidential elections.

It is safe to say the 3rd force didn’t fully materialize but the likes of Sowore, Moghalu and Durotoye and many others remain as viable alternatives for those who won’t vote for either a Buhari or Atiku ticket.

Analysis on Past Elections

A good starting point in predicting what may happen in the future is to look back at the past. A look back at trends and patterns from past elections could be useful to provide insight into the upcoming elections. This is what will be done here.

Voter Turnout (1999–2015)

Voter turnout has a considerable effect on the chances of a candidate to win an election. If voters don’t turn out to vote in key geographical areas where a candidate is hoping to amass many votes, it may reduce his/her chances of winning. Data shows that the voter turnout percentage has been declining since the 2003 elections.

84,004,084 people registered to vote for the 2019 elections. That is about 20 million more people than in 2015. Therefore, we expect that the percentage of voter turnout will increase this year compared to 2015. There are some interesting situations to consider however, one of them being voter apathy in the South East region of the country. With the resurgence of Nnamdi Kanu late last year and his recent order for south east voters to sit at home, coupled with ‘perceived’ disenfranchisement of the south east people, there may indeed be some of his supporters who will heed to the order. This will be detrimental to Atiku of the PDP, the party which has always won the South East region.

However, there is also a renewed zeal by the people of that region to get rid of President Buhari, who once made the infamous “97% versus 5%” comment in an interview in 2015 — referring to preferential treatment to those who gave him a large block of votes. To add to that, the Vice Presidential candidate of the PDP, Peter Obi is from the South East region. This may spur the people to come out to vote en masse.

States and Regions

Nigeria is divided into six geopolitical zones. The Northern part of Nigeria is predominantly Muslim and mostly inhabited by the Hausa and Fulani tribes. The Southern part is predominantly Christian and is inhabited by the Yoruba’s, Igbos, Ijaw’s and other tribes. Knowing how the states in these zones vote, gives an inclination as to how the votes will be represented in the coming election.

North Central

The North Central, also known as the Middle Belt is a tribally mixed zone where various tribes around the country can be found. It is also home for many Christians as well as many Muslims. In 2015, Buhari won 4 of the 7 states in the Middle Belt. He only lost out in the Federal Capital Territory, Plateau and Nassarawa.

In 2015, APC were in control of all the state governments in the North Central. This was a major advantage for them. However earlier in 2018, the political power houses in Kwara and Benue (including the Senate President Bukola Saraki, Kwara state governor and Benue governor Ortom) defected to PDP before the election campaign season. Saraki has a large cult following in Kwara which may work in PDP’s advantage. Beyond that, the past 3 years have seen a lot of deaths in the Middle Belt region due to the farmer/herdsmen clashes. For example, Benue state government conducted mass burials in January, March and May of 2018 and there have been many other reports of mass murder in Plateau and Nassarawa states. This may affect the way they vote this time as the failure of the government to protect their lives and property may still remain fresh as they go to the polls. We expect PDP to win in Benue, Kwara, Nassarawa and Plateau.

North West

The North West makes up about 24% of the total registered voters for the 2019 elections (20,158,100 voters). It is very important for whoever will turn out to be the winner of the elections. This is almost certainly where the then candidate Buhari won the 2015 elections. In 2015, the election was between a Northerner and Southerner, and there seemed to be a collective resolve by Northerners to deliver the Presidency to the North. Hence Buhari recorded large margins of victory in the North West especially. It’s different this time, it’s two Northerners (Fulani men) against themselves so there will be a lot of vote splitting.

Kano state is the second most populous state in Nigeria, hence also a very important state to keep eyes on. An interesting situation that has built up since the last election is that, Rabiu Kwankanso, a political godfather in Kano state who has a massive following, defected to the PDP in early 2018. This most likely will have an effect in the way things go in Kano, coupled with the bribery scandal surrounding the APC sitting governor, Abdullahi Ganduje. In Sokoto, Aminu Tambuwal the sitting governor, also defected to the PDP and hence that may also change the trajectory of things in Sokoto compared to 2015. Kaduna is another state where things may change drastically due to the mass murder of people in the past 3 years. The Zaria massacre in 2015 is a tragic event that may still be fresh in voters’ minds when they go to the polls. We expect PDP to win in at least 2 states in the zone.

North East

The North East has been plagued by terrorism even before the 2015 elections. The crisis still remains and voting will not take place in some local governments (in Borno particularly), instead a lot of voters will vote from internally displaced persons (IDP) camps. The PDP candidate, Atiku Abubakar is from Adamawa state, hence the statistics may change compared to 2015. Jonathan had 40% in Adamawa in 2015, Atiku will likely pull at least 55%.

Bauchi, Borno, Gombe and Yobe are strongholds for Buhari and it is difficult to see Atiku upsetting him in any of those states. However, Taraba is an outlier in the North east. It has a large Christian population and is also a pro PDP state which explains why PDP won there in 2015. PDP will win again in Taraba. Buhari however is most likely going to win the North east again but with a lesser margin than he won in 2015.

South East

The South East is a pro PDP zone, three out of the five governors in the State are PDP governors (1 from APGA, another from APC). Also, the South East states have consistently voted for PDP in the Presidential elections since 2003. In 2015, President Buhari only managed up to 19% of the votes in one state, Imo state. This number in Imo state may be due to the influence of the sitting governor, Rochas Okorocha who is part of the APC. However, there has been a crisis in APC Imo state chapter because Rochas’ preferred candidate, his son-in-law, didn’t win the APC governorship ticket. This may negatively affect the number of votes Buhari will get in that state but that is yet to be seen. In general, as long as all things remain equal not so much will change in the South East and PDP is expected to win all the states with similar margins. However, the voter percentage in the South East for the 2019 elections is only 11.97%.

South South

The South South is also a pro PDP zone. They’ve voted in favor of PDP since 2003. Jonathan won with large margins in 2015 except in Edo and this is due to the influence of then governor and present APC chairman, Adams Oshiomole. Not too much will change in the South South in which 5 states are PDP controlled except Edo state.

One state to watch out for is Rivers state where the former governor of the state, has some influence. He was suspended from PDP in 2013 and moved to APC then later went on to be appointed Minister of Transport. Despite Rotimi Amaechi’s influence, APC was able to gather only about 4% of votes. This time, Amaechi cried war songs in the APC campaign in Rivers state promising to somehow dislodge PDP (APC was barred by the Supreme court from participating in elections in Rivers state). Akwa Ibom is another state to watch out for. A state where Godswill Akpabio, who was former governor of the state and a staunch PDP member, defected to the APC in 2018. He also has some influence in the state. How much influence is yet to be known.

South West

The south west is another region to keep an eye on. It is occupied mostly by the Yoruba’s and in 2015 became an APC controlled region. All the six governors in the region are APC governors. However, in 2011 PDP won some of the states there. A major influence in the total shift in 2015 might have to do with the current vice President, Yemi Osinbajo who is also from the south west region.

Looking at the 2015 results, it is obvious the win margins weren’t that large and it seems that the same pattern will remain in 2019. Votes will be split almost evenly across both parties in all the states. If PDP can upset APC in states like Lagos and Oyo (the 1st and 3rd most populous states in Nigeria), then that will significantly increase their chances of winning. However, we expect that APC will win this region but with a very slim margin.

Predictions

We consider the results from the analysis we carried out on Twitter as well as the analysis on past elections to make predictions for the election on the 16th of February 2019.

At the end of the 28-day period, we have considered the overall average positive percentage for both candidates. The results show that Atiku had a 58% overall positive average compared to Buhari’s 42%. This indicates that for the time period leading up to the election, Atiku was the more favorable candidate. The reason is because Nigerians seemed to key into the Atiku agenda more than they did with the Buhari agenda. The past 3 years have been plagued with insecurity, loss of jobs, poverty and disunity. The Atiku plan and campaign was offering a better Nigeria with a more favorable economy (which was his key campaign promise), jobs, unity and security.

This result as we have it, is representative of the mood of a sample of voters leading up to the election. However, there are many other factors to consider when attempting to predict the results of elections such as voter turnout and patterns from past elections (discussed in previous sections).

Keeping all of that in mind (the mood of voters and past election trends), we make the following predictions;

● Both candidates will win 3 geopolitical zones each (in terms of states won). Buhari will take the North West, North East and South West. Atiku will win the North Central, South South and South East. There is a possibility for this to turn out differently in favor of Atiku especially if he could win key states like Lagos and Oyo in the South West and break Buhari’s strongholds in the North West.

● Atiku will win the 2019 Presidential elections with a smaller margin than Buhari won in 2015.

NB: Factors that may negate these predictions have to do with the following:

● Election insights unaccounted for.

● Sample size analyzed being too small to generalize.

● Other unforeseen or unexpected events/entities (such as vote rigging, power of a silent majority etc.).

● Social media doesn’t always reflect the reality during elections.

[1] https://medium.com/@senendu5/quick-summary-examining-sentiments-on-twitter-for-the-2019-nigerian-presidential-elections-81739873e71f

[2] https://developer.twitter.com/en/docs/tweets/search/api-reference.html

[3] http://saharareporters.com/2019/02/07/breaking-wole-soyinka%E2%80%99s-citizen-forum-endorses-moghalu-president

*Authors — Somto Enendu and David Anda