Why the Sierra Leone election was lost by APC after ten years of governance

Author: Yusuf Bangura: Sierra Leone Telegraph: 9 April 2018:

The March 31^st, 2018, presidential run-off election was an amazingly close race. The result has left the All People’s Congress party (APC) quite humbled. Julius Maada Bio of the Sierra Leone People’s Party (SLPP) emerged victorious with 51.8% of the votes,. His rival, Samura Kamara of the All People’s Congress (APC), scored 48.2%. Only 92,235 votes separated them.

A switch of 46,118 votes in the other direction would have produced a different result. No other election in Sierra Leone’s history has been this close, except perhaps the 1967 election, which was conducted under a parliamentary system of government.

However, if one compares the vote shares of the two parties between 2012 and 2018 (57.8:37.4; and 51.8:48.2 respectively), this was a massive swing of 10.5 percentage points—close to the other big swing of 12.3 percentage points (70:45.4) against the SLPP in 2007.

In most mature democracies, 10.5 percentage point swings would require at least two election cycles to overcome. The result is, therefore, a huge defeat for the APC and a great win for the SLPP.

How did the APC squander its 21 percentage point margin with the SLPP? And why did such a massive swing occur? The second question can be rephrased as ‘Why did the APC lose the 2018 election?’

A four-region swing

A bipolar ethno-regional cleavage underpins Sierra Leone’s electoral politics. This bipolarity is based on the numerical dominance of the two largest ethnic groups, the Themneh and Mende, which are roughly equal in size, and constitute slightly more than 60% of the population.

The Themne are located in North and have a large presence in the Western Area, and the Mende are largely found in the South and East.

The Mende are more hegemonic in the South-East, where they dominate six electoral districts, than the Themneh are in the North, which is more heteregeneous. Because of the North’s ethnic heterogeneity, Northern block voting, not ethnic group voting, historically informs voting behaviour in the region; whereas in the South and East, apart from Kono, Mende ethnic block voting seems to be the rule.

In the 1967 election, the population ratio between the North-Western Area and South-East was almost 50:50. However, by 2007, this ratio had changed to 55:45 in favour of the North-Western Area, and in 2015 to 56.5:43.5.

However, the ratio of registered voters has been more lopsided since 2012—it was 59.35:40.65 in 2012 and 60.5:39.5 in 2018.

The vote shares of the APC and SLPP in the last three elections reflected this ethno-regional divide. In 2012, for instance, the APC received 80% of its votes from the North and Western Area, and the SLPP  76% of its votes from the South and East.

In an ethnically bifurcated electorate in which voting is largely ethnic, elections are often won by maximising votes in one’s ethno-regional stronghold and making slight inroads in the ethno-regional stronghold of the opposing party.

In the case of the APC, for instance, prior to 2018, the strategy was to maximise its votes in the North and Western Area and penetrate Kono, which is the only non-Mende-speaking district in the East.

For the SLPP, the strategy has been to maximise its votes in the South-East and make inroads in Kambia and Koinadugu, Northern districts with substantial minority group presence.

Ernest Koroma needed a four-region strategy to avoid a run-off in 2012. He could not have won on the first ballot by relying only on the North, Western Area and Kono. He needed the votes of all four regions to get to the 55% victory score.

However, because of the lopsided nature of the electorate in ethno-regional terms in 2012, Koroma could have won a second-round ballot (which requires only 50%+1 votes), with only the 51.6% of the votes he received from the North, Western Area and Kono. This is a three-region strategy.

The 2018 results indicate that because of the lopsided distribution of registered voters in favour of the North-Western Area, Bio needed a four-region strategy to win the election in the second round of voting.

Relying on the South and East would have given him only 34.85% of the votes; and including the Western Area would have raised his vote share to 46.32%. It is only when his votes in the North are added that he is able to get to the 50%+1 mark.

The interesting point about Bio’s Northern votes is that reliance on only his votes in the districts with strong minority presence (Kambia, Koinadugu, Falaba and Karene) would have given him only 2.92 extra percentage points, which would have raised his overall vote share to 49.24%.

He needed his votes in the predominantly Themneh-speaking districts of Port Loko, Tonkolili and Bombali (which gave him 2.57 extra percentage points) to get him across the victory line[1].

The SLPP’s four-region strategy trumps the APC’s two-region strategy in this election, and it vividly explains how the APC lost the election. The APC may have been lured into a false sense of security by assuming that the ethnically lopsided nature of the electorate in favour of its regional strongholds gives it the option to ignore the South and East and focus largely on the North and Western Area.

How else can one explain the sacking of Sam Sumana (Photo), an elected vice president from Kono, and the alienation of the Kono electorate? Or the failure to choose a standard bearer or running mate from the South-East, even though a Southerner, Victor Foh, was vice president?

The choice of Samura Kamara, a Northerner, and Chernor Maju Bah, from the Western Area, as standard bearer and running mate respectively, may have sent a strong message to South-Eastern voters that they did not matter in the calculations of the APC.

The focus on the North and Western Area may explain  why the APC opportunistically tried to change the constitutional rule that requires 55% of the votes to avoid a run-off, to a simple majority of 50%+1 in the last days of the last parliament without public debate.

No party has ever won an election with a two-region strategy. The results do, indeed, indicate that even though both parties still draw most of their votes from their traditional ethno-regional strongholds, the APC has become much more regional than the SLPP.

89.2% of the APC’s votes are from the North and Western Area, whereas the South and East account for 67.3% of the SLPP’s votes.

The SLPP increased its vote share in every district, whereas the APC lost ground in all districts, including in Bombali where it obtained 90.7% of the votes, which is incredibly high, but is lower than the 93.2% it received in 2012.

The key to the SLPP’s victory was the maximisation of its votes in the Mende-speaking districts of Kailahun, Kenema, Bo, Pujehun, Bonthe, and to some extent Moyamba, to stratospheric levels; tapping into the anti-APC grievances in Kono, where it raised its vote share from 38% in 2012 to 72.6% in 2018; raising its votes from 25% to 39.5% in the Western Area; and making reasonable inroads in the North, where it increased its vote share from 6% in 2012 to 17.8% in 2018.

The SLPP’s votes in the Mende-speaking districts were, indeed, stratospheric – the party obtained 89% of the votes in those districts, with Bonthe, Pujehun and Kailahun each giving the party 90% or more of their votes.

The APC lost much ground in its traditional strongholds. For instance, whereas in 2012, it obtained 88% of the votes in the North, it could only get 82% of the votes in 2018.

The slide in the North is related to the challenge faced by the APC from the National Grand Coalition and other small parties, which took 19.5% of the votes in the first round.

The APC was only able to claw back 67% of those votes in the run-off, but this was not enough to prevent the SLPP from winning.

The SLPP was competitive in Kambia (it received 30% of the votes), Falaba (42.7% of the votes), and Koinadugu (32% of the votes). Kambia had the lowest voter participation rate in the run-off, suggesting a lack of interest after the NGC, the party of a plurality of the voters, failed to make it to the run-off (only 65% voted as opposed to a national average of 81%).

The APC’s harassment of the NGC and smaller parties in  the region would have made it difficult for the APC to win a higher percentage of these small party voters in the second round. Similarly, even though the APC obtained 72% of the votes in the Western Area in 2012, it received only 60.5% of the votes in 2018.

It was heavily trounced in Kono, where its vote share dropped from 58% in 2012 to 27.4% in 2018; and it failed to defend the gains it made in the Mende-speaking districts in 2012: its vote share declined from about 18% in 2012 to only 11% in 2018.

Debates on Sierra Leone’s electoral politics have often focused on the phenomena of ethno-regional strongholds and swing districts to determine the winning chances of parties.

The swing districts are assumed to be the Western Area districts and Kono district, with the idea that these districts have changed winning parties a few times in our last five party competitive elections.

The notion of swing districts gave rise to the view that no party can win an election without winning the Western Area or Freetown.

The historical record thus, indeed, indicate that when the APC won the elections in 1967, 2007 and 2012, it also won the Western Area districts; and when the SLPP won the elections in 1996 and 2002, it was also victorious in the Western Area. (Photo: Voters in Freetown).

The SLPP’s victory in 2018 without winning Freetown or the Western Area indicates that the idea of swing districts is unhelpful in understanding electoral politics in Sierra Leone.

The concept of swing districts or swing states makes sense in US presidential elections where the winner is not elected by the popular vote, but by an electoral college of voters representing the 50 states and Washington DC, with delegate vote share per state determined by the number of Senators and House respresentiatives for each state.

Under this system, a plurality of the popular votes in a district (apart from Maine and Nebraska) gives all the delegate votes of that district to the winner. It is logical, therefore, for American political scientists and commentators to focus on battle ground or swing states, which often determine the winner.

The notion of swing or ‘marginal’ constituencies (in UK parlance) also makes sense in first-past-the-post parliamentary elections, where a plurality of votes is needed in each constituency to determine winners.

However, where elections for the presidency are based on the popular vote, the appropriate focus should be on the percentage vote shares of parties in each district and vote swings between elections.

This approach indicates that analysts should focus less on parties flipping electoral districts and more on parties’s ability to improve vote shares. In other words, a party does not need to win a district in an opponent’s stronghold to win an election.

All it needs is to improve its vote share to a level that will get it to 55% or 50%+1 vote shares when the votes in all the 16 electoral districts are added. The APC and SLPP are still dominant in their respective ethno-regional strongholds of North-Western Area and South-East.

The only district that flipped in 2018 was Kono (and Kambia in the first round, which gave a plurality of its votes to the NGC). However, changes in vote shares in ethno-regional strongholds are what accounted for the election outcome.

Why the APC lost

Why did the APC lose the election? Some of the reasons are embedded in the first question discussed above on how the party lost ground in all four regions. In this section, I highlight four important reasons.

The first is the party’s arrogant sense of invincibility, which is informed by its monopolisation of power for 24 years (1968-92) and the lopsided ethno-regional distribution of voters.

It’s monopolisation of power from 1968-1992 is captured by the party’s infamous mantra of having 99 tactics of winning elections; and the ethnically lopsided electoral distribution of voters led to a false belief that it will never lose elections since voting is largely ethnic. (Photo: APC presidential candidate – Samura Kamara on the right,  after losing the first round voting).

These two factors created a dangerous, anti-democratic mindset of invincibility and fuelled the myth that ‘the APC does not lose elections it organises’. It partly explains why the SLPP’s base opted for a leader with a military record that ended the APC’s rule in 1992.

The logic of the SLPP’s choice is that the APC cannot be removed by democratic means alone—an idea that is equally dangerous for democratic politics. Elections cannot perform their basic job of accountability and mandate renewal if they are predictable or based on who has superior methods to fight or rig the outcome.

After the announcement of the final result, I sent a WhatsApp message to friends with just five words: ‘the end of an era’. The historian, Ibrahim Abdullah, and my insightful discussion partner on WhatsApp queried my use of the term ‘era’, since the APC’s defeat will not signal a clear break from the past.

However, the term very well signifies an important aspect of the evolution of our democratic politics: the slaying of the myth of APC’s invincibility through democratic means, without military intervention or civil war.

The APC tried many tricks to discredit the top management of NEC, especially its chairman, N’fa Alie Conteh; sought an injunction on the run-off; and accused the international observers, especially those from the European Union, the Commonwealth headed by Ghana’s ex-president John Mahama, and the British High Commissioner of orchestrating regime change.

Bio was forced to play rogue by calling for nation-wide protests if the elections were not held on the day they were scheduled before the injunction was granted, and asserted that he would stop to recognise Koroma’s legitimacy after 27 March.

This did not go down well with many people, and references to his impulse for militaristic or insurrectionary interventions re-emerged.

Sierra Leone has, however, changed in one significant respect: voters were tired of instability and war. The international observers did a great job of diffusing the tension, the NEC Chief was resolute in defending the independence of his institution, the injunction was lifted, Bio’s threats did not materialise, and the APC, in the words of Chairman Mao, turned out to be a ‘paper tiger’.

The humbling of the APC, in my view, is the single most important outcome of the 2018 election. It removes the idea of invincibility and assumed right to govern indefinitely in our politics.

The APC’s arrogant sense of invincibility gave rise to a second problem: the belief that it can govern without much respect for institutions or checks and balances. It’s huge mandate in 2012 (a vote margin of 21 percentage points) and control of more than 60% of parliamentary seats provided it with a buffer to rule without much accountability.

The party’s disrespect for institutions includes the illegal sacking of the elected vice president, Sam Sumana; selectively using the dual citizenship law to witch-hunt Kandeh Yumkella of the NGC, while simultaneously having a large number of ministers who are dual citizens; attempting at the 11th hour of the last parliament and in the middle of an election to change, without public debate, a constitutional rule that 55% votes are required to avoid a run-off in a presidential election; and elevating political parties over citizens’ democratic choices in determining when presidents and vice presidents can remain in office.

With the Supreme Court’s support, a party leader, who may not even contest elections in his party, is now more powerful than a president elected by millions of voters if they belong to the same party.

The party’s leader, Ernest Koroma, became all powerful and basked in the dubious title of Supreme Authority. He singlehandedly selected the party’s standard bearer and running mate—a first in the comity of democratic nations; and accepted the title of Life Chairman and Leader from the youth wing of the party—this was a throwback to the bad governance days of life presidents that ruined African countries in the 1970s.

The sense of invincibility and disrespect for institutions gave rise to a third problem: an inability to rein in  the country’s chronic corruption, despite platitudes of running Sierra Leone as a business and having zero tolerance for corruption.

For more than 10 years, the government failed to act on the Auditor General’s report, which contains numerous cases of misappropriated funds or expenditures that are not backed by proper documentation; it also failed to hold to account those who the Auditor General identified as responsible for the missing 14 million dollars meant to fight the Ebola virus disease.

The scourge of corruption reared its ugly head at State House during preparations for the Hajj pilgrimage in 2017. Officials at State House and other functionaries who pilfered pilgrims’ payments have still not been brought to book by the Anti-Corruption Commission.

Sadly, the government granted zero taxation to companies and duty waiver of imported goods, which only enriched mining companies and key government and party functionaries during the mining boom of 2012-13.

It has been estimated that the state lost $224 million in 2012 on tax concessions to only five mining companies and a cement company. The flip side of chronic corruption was a public service recruitment practice that was largely informed by party patronage and ethno-regional considerations.

The final grievance against the APC was its inability to transform the economy, improve the quality of education and health services, and lift living standards.

Throughout the campaign, the APC touted road construction as its key selling point for mandate renewal.

However, road construction, or infrastructure development, needs to be anchored in a jobs-generating growth strategy to improve wellbeing.

Even if voters are provided with the best roads in the world, they will react negatively during elections if they do not have jobs, good incomes and social services.

The APC failed to understand this logic throughout its 10 years in office. To compound the problem, the economy has been in dire straits since 2015 largely due to the collapse of global iron prices and the Ebola pandemic that halted many activities.

Despite a moderate recovery in 2016 and 2017, the economy remains undiversified; about 60% of youths do not have productive jobs, and 80% of them are in poverty despite the creation of multiple state organisations on youth affairs; inflation is about 20%; and there is often a lag in the payment of public sector salaries.

A combination of these grievances metamorphosed into a protest vote to kick the APC out of power. Political scientist Jimmy Kandeh was the first to recognise the results as a protest vote against the APC on his lively Facebook page. And in a WhatsApp exchange with me, he pinned down this protest vote to Kono and the Western Area.

The only distinctly vote-catching message of Bio’s campaign was his promise of free education, which surely would have earned him votes.

However, accusations of complicity in the extra-judicial killing of citizens and corruption when he was a top member of the National Provisional Ruling Council continued to dog his campaign. And even though his campaign was effective, it is difficult to recognise major policy promises that set it apart from the major parties and are compelling enough to deliver victory.

If anything, Kandeh Yumkella and the NGC were the breath of fresh air—they had a more comelling message on the economy, inclusion and governance. However, the ethno-regional barriers proved resilient, and many voters who were tired of the APC calculated that Bio’s SLPP stood a better chance than the other parties of getting rid of the APC.

This protest vote was similar to what happened in 2007 when the electorate got tired of Ahmad Tejan-Kabbah’s SLPP and voted for Koroma’s APC, which did not advance anything substantively attractive in the campaign and struggled with it’s terrible history of violence, corruption, economic mismanagement and repression when it governed between 1968 and 1992.  (Photo: President Koroma selling his Agenda for Prosperity that remained unachieved).

Let me return to the four-region swing that I highlighted as responsible for the SLPP’s victory to make the following point: even though the 2018 protest vote was intense in Kono and the Western Area, the defeat of the APC was a nation-wide or four-region protest, including in the Mende-speaking districts that gave 89% of their votes to the SLPP.

The APC, to recall, lost votes in all the four regions. The Mende-speaking districts gave 18% of their votes to the APC in 2012. Indeed, 149,021 voters in those districts voted APC in 2012; in 2018, only 11% or 90,346 voters did.

What accounted for the 3.5 percentage point swing against the party or loss of almost 60,000 voters? I submit that ethnic factors were intertwined with the national protest wave against the APC to produce the stratospheric numbers for the SLPP in those districts.

The vote swing against the APC in the North was 3 percentage points (88% in 2012 to 82% in 2018). There, ethno-regionalism helped to reduce the impact of the protest wave against the APC. The Western Area vote swing against the APC was 5.8 percentage points (72% in 2012 to 60.5% in 2018).

The biggest vote swing against the APC was in Kono: a massive 15.3 percentage points. Indeed, if 46,118 of the 91,823 Kono people who voted for the SLPP had switched their votes to the APC, the latter would have won the election.

In this sense, the Kono voters can rightly claim that they gave the presidency to Bio and had the last laugh on Koroma and the APC. In other words, the illegal sacking of the vice president, who hails from Kono, cost the APC the election.

Conclusion: prospects for democratic politics

The demolition of the myth of APC invincibility bodes well for Sierra Leone’s democratic politics. No democracy is viable if one party believes it has a natural right to rule indefinitely.

Since we seem to be stuck with the APC and SLPP, despite a clear alternative in the National Grand Coalition in 2018, the only way voters can ensure some kind change and hold parties to account is if they have a genuine chance of throwing incumbent parties out for poor performance.

Party alternation in power may be a blunt instrument of accountability, but it is better than having single party dominance in the governance arena.

Koroma apologised to the nation in 2007 for the APC’s 24 years of reckless rule, and promised to transform the party into what he called ‘A New APC’. One of his greatest failures was his refusal to modernise his party.

Modernisation would have required making the party more democratic, merit-driven, inclusive and not beholden to patronage networks. Instead, he encouraged a culture of sycophancy to become deeply entrenched in the party, allowed the party’s non-democratic rule of selection to continue to determine how representatives are chosen, and became more powerful than the party itself.

Party members indulged him in his quest for total supremacy because of the misplaced belief that he was an asset in winning and retaining power. They even canvassed the public to give him a third term or an extended stay in office—moves that he refused to stamp out until public opposition forced his hand.

A state-supported university was named after him and there was a reported initiative by the Central Bank to have his image on one of the countries’ bank notes, even though such actions are wrong for a sitting president in a democracy.

The low point of this abuse of institutions was when he publicly stated that if he had wanted a third term, no one one would have stopped him. Not even the constitution? Rather than a new APC, the party elite’s mind-set of domination remained the same as that of the old APC.

The party needs a fundamental reform of its constitution, mind-set and practices, if it is to win the confidence of voters in future elections.

The defeat of the APC holds promise for building a culture of autonomy in Sierra Leone’s key election management institutions: NEC, the police and the judiciary. The top leadership of NEC performed well in the second round of voting, and refused to yield to incumbent party intimidation.

It still needs to clean house at the field level, where many of its officials collaborated with the two parties to rig the first round of elections.

Amazingly, according to NEC’s data, overvoting occurred in each of the 16 electoral districts in the first round, which suggests a widespread attempt to stifle the people’s franchise.

For institutions like the police and the judiciary, this second alternation of power between parties by democratic means teaches a crucial lesson: political parties do not have permanent tenure in government—the primary interest of officials of such institutions should be to serve the state and not the party in government.

What are the prospects for democratic politics in the current dispensation? Bio’s SLPP government faces two kinds of constraints that can check the invincibility impulse and disrespect for institutions witnessed under Koroma’s APC.

The first is the very small vote margin the party enjoyed in the election. The next election cycle needs only 1.8 p[i]ercentage point swing against the SLPP to send it out of State House. As we have seen, that is only 45,118 votes.

Voters’ remorse in the North, Western Area or Kono for any number of reasons could be catastrophic for the party. This suggests that the party is likely to eschew the APC’s arrogance of unchecked power if it seriously wants voters to give it a second term in office.

The 2018 election has sent a strong message that no party can use undemocratic methods to prolong its life; the leash the electorate has on the SLPP is very short.

The second constraint is the balance of power in parliament. The results indicate that the SLPP will not have a majority or emerge as the largest party in the legislative body. This is not the first time that a president’s party will not have a majority in parliament.

In 1996, Tejan-Kabbah’s SLPP government had only 39.7% of  the parliamentary seats—its alliance with the People’s Democratic Party gave it a comfortable majority of 57%. In the new parliament, the APC will have a very small majority of 68 out of the 132 seats[2].

This is the first time in our history that the governing party will not be the largest party in parliament—it will not be able, therefore, to provide the Speaker and Majority Leader and drive the law-making process[3].

The seat distribution indicates that Bio’s government will not enjoy Koroma’s freedom to do as he pleases, without parliamentary contestation and bargaining.

Key appointments, bills and the budget may have to be negotiated and not imposed or rushed through as in the previous parliament. Smaller parties like the NGC and the Coalition for Change will have the power to punch above their weight.

These two constraints can, however, also produce negative outcomes or fail to discipline the government. Sierra Leone’s history is marked by both hope and disaster—any analysis will need to recognise, therefore, both positive and negative outcomes.

The possibility of gridlock of the type that often occurs in the US when Congress is not controlled by the president’s party can be one outcome. This is likely to occur if the APC uses its majority power to block the president’s intiatives because of executive actions the party vehemently opposes.

One likely area of contention is the replacement of people in top positions in the public sector that owed their loyalty to the APC. Bio’s government will be under tremendous pressure from its base, especially the Paopa faction, to move swiftly on this issue—and they will point to the far-reaching changes in top level personnel when the APC came to power in 2007.

Such struggles, if not properly managed, could provoke the government to govern in a non-accountable and confrontational way in order to demonstrate executive power.

A second problem is limitations on the bargaining process across parties for majority outcomes. In a situation where the governing party does not have a majority, such bargaining has one significant drawback: the lack of autonomy of individual MPs in our parliamentary system. Articles 77 (k) and 77 (l) of the constitution state that an MP can lose his/her seat in parliament if he/she ceases to be a member of the party on whose ticket the election was won; and the Speaker, in consultation with the leader of an MP’s party, can cause the expulsion of that MP if he/she sits and votes with a party other than the one on whose ticket the election was won.

This may limit the ability of the president to make deals with individual MPs without the backing of the MPs’ party leaders. If the APC provides the Speaker and Majority Leader, the government will find it difficult to make deals with individual APC MPs without the consent of the APC leadership.

Experiences under Tejan-Kabbah’s SLPP and Koroma’s APC on this issue are not encouraging. In 1996, John Karefa-Smart of the United National People’s Party (UNPP) accused his parliamentary party members of disloyalty and collusion with the SLPP government. He dismissed 80% of them, but the affected MPs joined hands with SLPP MPs to have Karefa-Smart suspended from parliament.

This incident strained relations between the UNPP and the SLPP government, making it difficult for Karefa-Smart’s faction of the UNPP to work with Tejan-Kabbah in opposing the military coup of 1997.

In the case of Koroma, even though he did not really need the votes of the SLPP in parliament, SLPP partisans accused him of interfering in the protracted struggles among SLPP MPs for control of the party’s parliamentary agenda. The SLPP failed to hold the Koroma government to account in parliament.

Bio’s government and opposition parliamentary parties should study these and other cases around the world and work out schemes that will lead to healthy cooperation for democratic politics and development.

[1] There is a large Themneh presence in Kambia district, which it shares with the Soso, Madingo and Limba; and some Themneh presence in Karene, where the Limba are the largest group.

[2] The result for one seat in Falaba district is still to be declared by NEC.

[3] It is not clear what the status of the 14 paramount chiefs who were elected by non-popular vote will be in determining the leadership of the House.