On October 9, Austria’s Chancellor Sebastian Kurz was forced to resign amid a unique scandal. He was accused of orchestrating false polls and bribing the media to present them as genuine opinion polls. It’s common for political leaders to get caught up in scandals, but this may be a first where the leader of a nation has had to step down over seemingly innocuous opinion polls. It is a reflection of the growing “weapon” of opinion polls in electoral democracies.
In the early days of the election polls that began with the US presidential elections of 1936, these polls were just a pastime for the curious. Then, frequent polls evolved to serve as a feedback loop for governments and politicians. However, in recent years, political leaders and parties have armed opinion polls to shape and influence, rather than simply reflect, public preferences.
Out of 100 Indians who voted in the 2019 national election, only 35 were engaged voters, who knew which party to vote for even before the campaign began. The remaining 65 voters decided who to vote for in the last days or weeks before election day (Lokniti National Election Study). Thus, 65 percent of Indian voters make their voting decisions quite late in an election.
Further away, a significant 43 percent of those uninvolved voters make their choice based on hawa – who they think he is likely to win over. This is, indeed, nearly 30 percent (43 percent of the 65 percent of unengaged voters) of all Indian voters – a significantly high number, especially when the margins of victory in multi-party elections in India are weak.
A rigged pre-poll disguised as an independent scientific study by a professional polling agency and widely disseminated by compliant and compromised media can play a powerful role in creating hawa or a “wind of victory” for a particular party and influence the party. 30 percent of floating voters to vote for this party. Since these polls wear a veneer of objectivity, voters are more easily misled. Thus, false opinion polls can be extremely powerful weapons in Indian elections.
Certainly, misleading opinion polls in favor of one party do not always help, as evidenced by the recent Bengal state elections, when nearly all of the media predicted the BJP as a clear winner in their pre- polls when in reality the BJP has lost heavily.
But that doesn’t change the fact that manipulated opinion polls propagated by compromised media are a powerful weapon in Indian elections, actively deployed by some political parties. A widely circulated rigged opinion poll with the potential to influence 30% of the vote is as dangerous to Indian democracy as fake news on social media, if not more.
Some academics and political leaders have called for a ban on pre-election polls in India, which can be an extreme measure. However, the Election Commission can help voters detect bogus opinion polls by setting standards and guidelines.
Any opinion poll that does not reveal its survey methodology, sample selection technique, sample size and the exact questionnaire should be considered suspect. A robust survey will also make all of its raw data public, so people can use and replicate the analysis. Most of the polls published by Indian media, except a few, do not even disclose the basic details of their investigative methods, let alone the release of their data.
Most people judge a survey only on their sample size by assuming that a large sample size means a good survey. Pre-ballot polls for the 2020 US presidential elections with 160 million voters had an average sample size of 3,000 people. The Reserve Bank of India determines interest rate decisions for a $ 3 trillion economy based on a survey of 6,000 of India’s 250 million families. The herd immunity levels of Covid were determined on the basis of serological surveys of 30,000 people in Delhi for a population of 20 million.
The notion of sample size is very poorly understood. The selection of the sample is much more important than the size. To investigate a state election in India, a robust sampling methodology should choose people from every constituency, identity, age group, and sex. A survey of 2,000 well-represented people is far superior to a survey of 20,000 people selected from a few assembly constituencies or a specific religion or age group. As a rule of thumb, a credible survey for a state election in a highly diverse, multi-party, first-past-the-post democracy like India will need to survey 500-1,000 randomly selected people from each constituency in the assembly, in order to predict the seats with reasonable precision.
The Election Commission can require the disclosure of detailed survey methods, raw data and prescribe minimum standards for stratified sampling for pre-election surveys. By these criteria, almost all investigations published by Indian television media are suspicious and questionable. A poll that passes certain elections is not proof of its credibility or its robustness, just as a broken clock proves it right twice a day. The vast field of empirical science, of which investigations represent only a small part, is an established discipline for which Nobel Prizes have been awarded, and not something that night flight operators can easily replicate.
Opinion polls not only reflect people’s opinions, but influence them to create an aura of gain. If 30 percent of Indian voters are vulnerable to such influence, then there is a dire need to regulate this danger and protect the sanctity of Indian democracy.
This column first appeared in the paper edition on November 8, 2021 under the title “The Perils of Opinion Polls”. Chakravarty is President of Congress Party Data Analytics and Political Economist