‘Poll survey was professional’

The Zomba-based Institute of Public Opinion and Research (Ipor) has released findings of a 2020 Pre-election and Governance Survey which predicts an outright victory for Tonse Alliance candidate Lazarus Chakwera and his running mate Saulos Chilima. The poll finds that Tonse Alliance has an 18-point lead over the DPP-UDF partnership of President Peter Mutharika and Atupele Muluzi. In this interview our News Analyst LUCKY MKANDAWIRE engages Ipor director of research and operations Boniface Dulani on this and more. Excerpts:

What was this study about and what were the objectives?

This research in brief was about the state of governance ahead of the forthcoming fresh presidential election where we basically sought to gather the views of ordinary citizens about the state of preparedness of various electoral stakeholders including the voters themselves, the Malawi Electoral Commission (MEC) and also to hear their intentions on how they intend to vote in the forthcoming election. It was conducted between May 25 and June 3 2020 and the sample was national representative, disaggregated by regions using the 2018 Malawi Population and Housing Census. We randomly selected 150 enumeration areas allocated by location whether rural and urban so 18 percent of our sample was urban and then 82 percent in rural Malawi.

The study reached out to 1 346 respondents, how would you state the sampling method was a national representation of Malawians’ perceptions?

This is actually following international standards in random sampling and the sample is actually drawn by experts in sampling. In our case, some people might think the 1 346 respondents we spoke to is an insignificant number but in research you don’t necessarily need to speak to everyone to know what they are thinking and that’s the essence and science behind sampling.

I should also add that even in countries like the United States of America (USA) with almost 400 million people sometimes they can tell what their citizens are thinking by just sampling 600 people and they can still get very fair representation of what the wider citizenry is thinking about. So the 1 346 respondents we had satisfies the minimum standards of sampling. The rigour of the research is in how the sample was selected and not the number of people that we spoke to.

So what are some of the key findings from this research as regards to people’s intentions on voting?

Perhaps, before we talk about the voting intentions, it is also important to point out that we did ask respondent, how they voted in the previous elections and we see there is a quite significant proportion of voters who say they will switch their votes from the candidates they voted for in the previous elections.

For example, 18 percent of voters who voted for APM say they will now vote for Chakwera. On the other hand, there is about two percent of voters that voted for Chakwera now say they will vote for Mutharika. Of course, there are also switches across the board in terms of the political parties.

We also found that our voting trends still follow some regional patterns much that the Northern and Central Regions are predominantly interested in voting for the Tonse Alliance candidate whilst in the Southern Region about two thirds say they will vote for Mutharika although the Tonse Alliance candidate is fairing slightly better in the region than he did in last year’s elections. The major driving force could possibly be due to the alliance partnership that MCP and UTM have which is appealing to people.

So in a nutshell we find that 51 percent of voters say they would vote for Chakwera, 33 percent would vote for Mutharika and about 14 percent were undecided or they did not want to disclose who they would vote for where as two percent said they would not vote.

Again, in our numbers we have also tried to breakdown how we think the people who said they were undecided and those that refused to tell us their choices are likely to break and when we take those numbers into account we are projecting that at least Chakwera’s share of votes can go up to 58 percent while that of Mutharika could go up to 38 percent and either way we think the results most likely favour the Tonse Alliance candidate especially when using the 50 percent-plus-one majority threshold.

On the governance perceptive, what are the study highlights?

In terms of economic governance, many Malawians, large majorities actually, tell us that the country is headed in the wrong direction and that the state of national economy is also in a very poor shape.

Many Malawians also say that even though we have the Covid-19 pandemic elections should still go ahead and that even the campaigning should be allowed in the process.

We also hear Malawians saying as voters they feel that they are ready to vote and that they support the decisions of the Constitutional Court as well as the Supreme Court of Appeal to cancel last year’s elections.

Of course, Malawians have some concerns on the state of preparedness of MEC. They feel it is not as prepared, although they feel that opposition parties in particular are more prepared than the ruling party and its alliance partner the UDF.

What can you say about the amount of random sampling error in your study findings and how accurate have been previous Ipor surveys?

A survey with a sample size of 1 346 such as this one gives us a margin of error of plus or minus three percent at the 95 percent confidence interval and what that effectively means is that the numbers could be either high by three percent or less by three percent. In our previous surveys, we found that our elections tended to be very close to the point that it was always difficult for us to say we can project a clear winner but this time around, we see that the difference is so large that the more likelihood at the moment if these numbers were to hold would be that Chakwera would win.

But I should also add by saying that this is a survey and the real vote is on the election day. So, if this is what we are hearing from Malawians but whether Malawians will now go out and vote as they should because that’s their constitutional right then I think we might see whether we will be vindicated or not.

But we have always been vindicated in previous surveys so we are confident of our numbers especially given the rigor behind the sampling that goes into these studies.