Continuing our series of Monkey Cage Election Reports, the following is a pre-election report on Tuesday’s tripartite election in Malawi.
We’re submitting this post from Zomba, Malawi on the eve of the country’s fifth election since the institution of multiparty democracy in 1994. The polls will open on Tuesday morning at 6 a.m. local time and will close at 6 p.m. in the evening, or until the last voter on the queue at 6 p.m. casts his or her vote. There are a total of 4,445 polling stations across the country and voters will cast three ballots: for local government councilor, Member of Parliament and president.
In the run-up to the election, there was one major episode of violence prior to the official campaign period (already discussed in our earlier post), and Monday there was an unconfirmed report of tension in the city’s capital, where some citizens burned official election materials (excluding ballots, we are told) that were being held at Lilongwe City Council. With these exceptions, the campaign has been mostly peaceful, and at political rallies, mostly festive. Campaigns closed on the morning of May 18, in accordance with the electoral rules barring campaigning in the 48 hours prior to the opening of the polls.
Most polls leading up to the election draw from non-representative samples and are dubious in predicting poll outcomes. For example, a Sunday Times poll from September 2013 reported the contest was too close to call. The poll placed Lazarus Chakwera of the Malawi Congress Party (MCP) marginally ahead with 27.21 percent of the voters, followed closely by Peter Mutharika of the Democratic Progressive Party (DPP) with 27.06 percent of the total votes polled. In third and fourth place were Atupele Muluzi and Joyce Banda of the United Democratic Front (UDF) and People’s Party (PP) respectively. However, the poll’s sample relied on self-selected respondents who sent text messages to a premium number or voted online.
A second poll by Research Tech Consultants predicted victory for President Banda with 42 percent of the vote, followed by MCP’s Chakwera at 23 percent. Mutharika and Muluzi were placed third and fourth with 22 percent and 10 percent respectively. The pollsters claimed to have interviewed 3,883 respondents across the country. However, the poll’s credibility was widely questioned, especially since there was no record of the polling firm or any of the leading members of the team having ever done a survey of this kind in Malawi or anywhere in the world.
The best available data were collected from March 22 to April 5, 2014 by the Afrobarometer, an independent, non-partisan, African-based network of researchers that specialize in measuring citizens’ attitudes toward democracy, governance, leadership, identity, and other related issues. The 2014 Malawi Afrobarometer survey interviewed a random sample of 2,400 adult citizens across the country and had a margin of error of +/- 2 percent.
An Afrobarometer Dispatch concluded that the election is “too close to call,” despite the fact that the survey results showed a plurality of voters (27 percent) said they favored Mutharika of the DPP, followed by Chakwera (MCP, 21 percent), Banda (PP, 19 percent), and Muluzi (UDF, 14 percent). Of particular note was the fact that 15 percent of respondents said they had either not made up their minds about who they would vote for or were unwilling to share with the survey team their voting intentions.
It is important to keep in mind that though the Afrobarometer pegged younger voters as giving greater support to Peter Mutharika, this group has historically reported lower rates of voting. For example, in the 2012 Afrobarometer Malawi survey, 87 percent of eligible respondents reported to have turned out to vote in the 2009 election, but only 77 percent of those under 25 years old reported turning out in 2009. Furthermore, age is but one potential identity that can predict how people will vote.
Much of the political science research on voting in Africa has examined the salience of ethnicity in politics (particularly in Malawi). Political scientist John Ishiyama argues ethnic bloc voting in Africa is more common where ethnic groups are geographically concentrated. There are reasons why ethnic bloc voting can’t be easily tested in Tuesday’srrow’s election. The four front-runners hail from three ethnic groups: Mutharika is Lomwe, Chakwera is Chewa, and both Banda and Muluzi are Yao. Ethnic bloc voting would attribute variation in political support based on candidates’ ethnicity (e.g., Lomwes would support Mutharika, Chewas would support Chakwera, and Yaos would be split between Banda and Muluzi).
We expect there will be regional patterns to voting that casual observers might attribute to ethnic voting. However, even if we were to extend the ethnic bloc voting logic to include support for candidates from non-co-ethnics living in the vicinity of co-ethnics, none of the four leading candidates hail from groups predominant in Malawi’s northern region. According to 2008 census data, most districts are predominantly populated by ethnic groups from which no front-runner presidential candidate hails.
Based on our analysis of media reports, the current and earlier waves of the Afrobarometer, and previous electoral data on vote choice and turnout, along with discussions with campaign personnel, we agree with Afrobarometer’s conclusion that the race is too close to call. We share below a map of predictions of which of the four leading candidates will likely achieve a plurality of the votes in each district.
Check back in the coming weeks on The Monkey Cage for our post-election report. We will also be posting regular updates on Twitter on election day and throughout the vote count until a winner is declared.
Kim Yi Dionne is Five College Assistant Professor of Government at Smith College. She studies identity, public opinion, political behavior, and policy aimed at improving the human condition, with a focus on African countries. Follow her on Twitter at @dadakim.
Boniface Dulani is Lecturer of Political and Administrative Studies at Chancellor College, University of Malawi’s flagship arts and sciences institution. He studies presidential politics, public opinion, and political behavior. Follow him on Twitter at @BoniDulani.