By Mimi Alphonsus   Who will win the presidential election? Pollsters, relative newcomers to Sri Lanka’s election scene, claim to have the answer. Companies, organisations, and social media accounts have taken to conducting pre-election surveys, but it isn’t clear that their methods, and indeed, their results, are likely to be accurate. The Election Commission (EC) has [...]

News

Election Commission weighs in on opinion polls as they gain traction

View(s):

By Mimi Alphonsus  

Who will win the presidential election?

Pollsters, relative newcomers to Sri Lanka’s election scene, claim to have the answer. Companies, organisations, and social media accounts have taken to conducting pre-election surveys, but it isn’t clear that their methods, and indeed, their results, are likely to be accurate. The Election Commission (EC) has made its displeasure about intent polls well-known and is now at loggerheads with the pollsters.

There are multiple pre-election polls—many by random accounts on social media and some politically affiliated. But Helakuru and the Institute for Health Policy (IHP) appear to dominate the scene both in terms of respondent numbers and in terms of traction.

Helakuru is a popular Sinhala-language application headed by founder and CEO Dhanika Perera. It predicts an 80% victory for Anura Kumara Dissanayake. The Institute for Health Policy (IHP), a research organisation headed by Dr. Ravi Rannan-Eliya in July, had predicted a 37% victory for Dissanayake, closely followed by Sajith Premadasa at 36%.

But critics have pointed to sample biases and an inability to account for last-minute political shifts as reasons to dismiss the results. Helakuru relies on a large sample, which last week had over 30,000 respondents. Their respondents are almost entirely Sinhalese, and because they rely on users of their app to fill out the survey, the data skews towards young people who can afford smartphones and internet.

IHP relies on a combination of data from existing participants of their health surveys as well as random dialling of telephone numbers. They use data from nearly 20,000 interviews conducted since 2021, of which 701 were conducted in July this year, ahead of their 37% prediction for Dissanayake. The data is analysed using a Multilevel Regression with Poststratification (MRP) model to correct differences in their sample populations over the long period of time when data was collected. MRP has been successfully
used to get results at a localised level from large national samples, but IHP has adapted the method to narrow by time rather than region, raising questions.

Despite some scepticism, the predictions have gained traction. IHP’s poll in particular has been broadly considered reputable and has been reported in Ada Derana, Mawbima, Thinakural, and several other popular news sources.

The Election Commission (EC) is not pleased with the trend. Election Commissioner General Saman Shri Rathnayake questioned the agendas behind the “corrupt” and “illegal” practice and claimed that the Presidential Elections Act indirectly prohibits such activity. He worries about the sample bias in the polls deceiving people and fears that they will sway voters. For now, the EC has directed the media to refrain from publishing the results of election surveys without a description of the pollsters, their funding, and their methodology. “Elections are not a product to be given to the market for research,” Mr. Rathnayake said, adding, “These polls are an artificial intervention.”

However, Dr. Rannan-Eliya and Mr. Perera disagree. Dr. Rannan-Eliya says that there is no consistent evidence to show that polls sway election results and that in reality, polls, especially those that ask about voting priorities like the IHP polls, provide a valuable service. “By indicating a close election, we are providing a service to the Election Commission too,” said Dr. Rannan Eliya. “They should be focussing on educating people about preferential voting because not many people know how to do it, and second preference votes are going to matter in this election.”

He also argues that polls are a way of showing politicians “what people really want” and ensuring they align with the public’s interest. This sentiment is shared by Mr. Perera, who plans to use Helakuru’s polling platform to collect citizens’ opinions on policy decisions after the new government is formed.
“It is a resource for people to express their opinions,” he explained, “and by seeing how the public has responded, it’s a way to develop empathy for opposite
points of view.”

Political interest alone does not draw people to opinion polls, business interest does too. The Colombo Stock Market is already suffering from election uncertainty as shareholders speculate on what is likely to happen. Stockbrokers told the Sunday Times that in addition to news reports and the release of manifestos, the market is also reacting to polling data.

Whether the Election Commission likes it or not, voting intent polls seem to have cemented themselves in
Sri Lanka’s election landscape. However, hiccups on the international scene provide a cautionary tale against over-reliance on these polls. None of the major polling firms in India correctly predicted the recent Lok Sabha elections this year, and French pollsters didn’t anticipate the hung parliament and ultimately leftwing
alliance victory.

Share This Post

WhatsappDeliciousDiggGoogleStumbleuponRedditTechnoratiYahooBloggerMyspaceRSS

The best way to say that you found the home of your dreams is by finding it on Hitad.lk. We have listings for apartments for sale or rent in Sri Lanka, no matter what locale you're looking for! Whether you live in Colombo, Galle, Kandy, Matara, Jaffna and more - we've got them all!

Advertising Rates

Please contact the advertising office on 011 - 2479521 for the advertising rates.