119 Years of Trust


Saturday, August 28, 1999

This above all

regional vignettes

Tripti Nath and R. Suryamurthy discuss the accuracy (or the lack of it) of the election forecasts made by psephologists
Poll Pundits or Punters?

IN the run-up to the elections, the voter gets flattering attention from political parties and candidates alike. His views are indispensable and his voting intentions become a subject of scientific research by pollsters. His opinion is gauged by opinion poll, exit poll and post-poll analysis.

Polling opinionsIn their book, India Decides, well-known psephologists, David Butler, Ashok Lahiri and Prannoy Roy have described opinion poll as a tool of research that has only recently been exploited. The authors feel that although opinion poll results are fallible, especially in Indian conditions, they are the best means of deriving a more accurate picture of the voting intentions and political attitudes of the Indian electorate.

While opinion polls, psephologists admit, are fallible in Indian conditions, exit polls and post polls are relatively accurate. The difference between the three, according to noted psephologist Yogendra Yadav, is that of timing. He points out that other crucial factors include sample size and sample design and the representativeness of the sample ensures accuracy.

Naveen Surapaneni, a psephologist at the Centre for Media Studies, says that while the opinion poll gives the likely trend of the voting pattern, the exit poll gives a clearer picture as it is taken outside the polling booths. "From past experiences it is evident that voters tell the true and factual opinion about political parties and their intentions," he said.

Yogendra YadavYogendra Yadav, a Fellow of the Centre for Study of Developing Societies, says post poll is an academic exercise in which a random scientific sample is drawn from each electoral roll. The researchers visit the house of the voters after the poll and talk to them in a relaxed atmosphere. The questionnaires used in the post poll surveys are extensive.

He mentions that post poll is not an instrument of forecast but a means of post-mortem for a long-term analysis. The exit polls are relatively less expensive as the respondents "come to you and you do not have to go to them".

He says that pre-election poll suffers from the uncertainty whether the person will vote or not. "Moreover, you cannot have a truly representative scientific sample and cannot truly randomise which is considered the best scientific method in statistics".

The commercial opinion poll in India is conducted by Quota Sampling. Yadav says that the CSDS State Co-ordinators are academics who engage their research associates and students for the opinion polls. "The exercise lasts about a week and involves up to 400 persons."

Titoo Ahluwalia, Chairman of ORG-MARG, one of the country’s largest market research companies, in an interview to a national daily has said that opinion polls in India have proved to be a pretty reliable indication of the way the political wind is blowing.

Talking of exit polls in the same interview, Ahluwalia informs that exit polls are a relatively recent phenomenon in India and "we are still at the steep end of the learning curve."

Mahesh RangarajanThe first national poll was carried out by the Indian Institute of Public Opinion before the 1957 general election. Under the guidance of Eric da Costa, the father of opinion polling in India, the institute has covered almost all subsequent elections. News magazine India Today has also promoted professional opinion polling in India. Since 1980, it has commissioned market research groups and psephologists to conduct country-wide opinion polls.

Opinion polls collect information by interviewing a sample of the population whose opinion we are interested in. In a typical election survey, it means choosing sample constituencies and polling booths and the selection of a few voters who resemble the entire population. Their responses are then recorded. Some researchers use dummy secret ballots for sensitive questions on voting to encourage an honest response. All the responses are then collated and computed.

Election forecasting is essentially a game of numbers involving two steps. First, the pollsters conduct an opinion poll to find out the likely choice of voters. Under our electoral system, the percentage of votes has no direct relationship to the percentage of seats won by a party. The second step involves the translation of votes into seats through elaborate statistical techniques.

But do opinion polls influence the voters in their choice of candiadate or party? A survery conducted by ICSSR-CSDS-India Today in 1996 indicated that an overwhelming majority of ordinary voters had not read or heard about opinion poll-based election forecasts.

Yadav says his assessment after a post-poll survey done by the CSDS three years ago was that only 0.7 per cent of the voters were influenced in their voting decisions by opinion polls.

Psephologists state that some studies have suggested that there is a bandwagon effect --- people rushing to support the winner. Others claim that there is an underdog effect-- - voters backing the apparent loser. They argue that there is no consistent evidence to support the contention that opinion polls have a net influence on voting behaviour and added that most of the opinion polls are published in English-language press whose reach is quite limited.

Yadav says, "These terms have come from the West. In assessing the effect of the opinion polls, we don’t have sufficient evidence in this country to prove such hypotheses.’’

Election analysts, however, admit that the opinion polls help in strategic voting. If the poll projects the candidate of one’s preferred party as an apparent loser, the voters may cast their vote in favour of the second most preferred candidate.

GVL Narasimha RaoG.V.L. Narasimha Rao of the Development and Research Services (DRS) is of the view that opinion poll results cannot alter the party leanings which the electorate acquires over a long period of time. "The opinion polls, however, do influence the ‘undecided’ voters and it does have an effect on the morale of the party cadre. The ‘undecided voter’ influenced by the opinion poll may resort to tactical voting, thereby increasing or decreasing the margin of defeat," he said.

Naveen Surapaneni, psephologist at the Centre for Media Studies, agrees with Rao that only the undecided voters are influenced by the opinion poll results to some extent. He points out that sometimes voters change their views about political parties due to certain events that occur during campaigning.

Narasimha Rao says varying results thrown up by different opinion polls are the result of the sample size and their quantum. He asserts that opinion polls and exit polls are the most scientific methods of assessing the mood of the people and gauging their opinion on the strengths and weaknesses of the political parties.

One may ask whether these predictions can be taken as the gospel truth? Psephologists argue that opinion polls are better indicators of popular mood than cocktail circles, press club gossip or hit-and-miss headlines as pollsters try to speak to ordinary voters to find out how they might vote. They may or may not succeed in giving an exact forecast of the number of seats a particular party will gain, but they give a better idea than anyone else of the direction in which the wind is blowing.

Political commentator and historian Mahesh Rangarajan says that people ought to realise that political commentary is not "P. C. Sorcar’s magic."

Yogendra Yadav, while agreeing with Mahesh Rangarajan, says that election forecasts are unfortunately equated with black magic. "Either there is a sense of complete awe and unadulterated admiration, or poll forecasts are looked upon with hostility and suspicion. I would like people to look at them as attempts to develop scientific estimates," he said.

The most significant pitfalls that opinion surveys and exit polls have to steer clear of are those connected with sample size and selection. The question of size is addressed through a statistical formula applied to the electorate as a whole. Yet, even with the most careful system of selection, as by using forms of stratification by region, locality and class, poll samples often fail to reflect the bewildering complexities of electoral behaviour in India.

At another level, opinion surveys and exit polls are known to be fallible on account of the inherent questionnaire biases. Even the manner in which a question is posed can often influence its answer.

Viewed with "hostility, suspicion, unadulterated admiration and complete awe,’’ poll forecasts, however, attract the attention of the voters and politicians alike.

Poll jargon

Exit poll: An opinion poll which is conducted on election day at the exit gate of the polling booth with the purpose of asking voters whom they voted for. A technique used to predict the election outcome before the actual counting.

Random sample: The ideal textbook technique of sampling that ensures that the sample is free of bias. Random here is not haphazard.

Quota sample: A technique of obtaining a representative sample by fixing in advance a quota for all significant sections of the population. It is cheaper and easier to execute than the random sample, but compromises on quality.

Weightages: While analysing the findings, pollsters estimate the degree by which different sections of society may have been under- represented or over-represented in the sample. They can’t change the sample, but can correct possible distortions mathematically. That is, by reducing or enchancing the ‘weight’ of their answers in the final calculations.

Swing: It means a change in the percentage of votes for a party over two points in time. If the BJP’s vote percentage increases from 20 to 24 per cent, it means a 4 per cent swing for it. Evolved by psephologist David Butler for the British elections, swing has acquired a very simplistic meaning in the lexicon of the Indian pollsters

EC directive

PUBLICATION and dissemination of results of opinion polls and exit polls continue to be prohibited by the Election Commission.

Reiterating the guidelines issued before the 1998 Lok Sabha poll, the EC has restrained the publication or dissemination of results of opinion and exit polls either in print or electronic media from 5 pm on September 3 till closing of the poll in all states and union territories on October 3.

The commission has also directed that organisations and agencies conducting opinion or exit polls must indicate the sample size of the electorate covered by such polls and the geographic spread of the survey.

The polls should give details about the methodology followed by the percentage of errors, the professional background and the experience of the organisation and key professionals involved in the conduct and analysis of the polls, according to the EC order.

Similar directions issued by the EC last year were challenged in a petition filed by Frontline magazine. The case is still pending in the Supreme Court. While declining to pass an interim order, a three-Division Bench comprising Justice M. M. Punchhi, Justice B. N. Kirpal and Justice M. Srinivasan observed, "If we grant a stay, we will be allowing the petitions and similarly if we deny the stay, we will be dismissing them. We are doing neither, we are simply adjourning the matter. The Press might take risks."

Yogendra Yadav says he is in favour of the ban on publication of results of exit poll. "Publication of exit poll results should be banned till the last person has cast his vote. There should be some sanctity of the vote."

Yadav, however, feels that the Election Commission’s order banning publication and dissemination of results of opinion polls from September 3 to October 3 will deprive the voter access to balanced and correct information. "This means, nothing will come out between September 3 and October 3. It is a rather extreme ban. They could have imposed the ban in phases," he said.

G. V. L. Narasimha Rao has no objection to the EC directive banning the publication of opinion poll results. He is, however, opposed to a blanket ban as opinion poll is another means of disseminating information about what what people think about political parties and the issues they think are important. "Opinion polls are similar to what is written in newspapers, magazines and discussed in television. If opinion can be expressed about the outcome of the elections in these media then opinion and exit polls should also enjoy the same rights," he argues.

Apart from the Election Commission, the Press Council of India has laid down guidelines on "pre-poll and exit poll" surveys. Justice P.B.Sawant, Chairman of the Press Council of India and President of the World Association of Press Councils, laying down guidelines for the media, stated that newspapers should neither allow their forum to be used for distortions and manipulations of the elections nor allow themselves to be exploited by interested parties.

The Press Council, therefore, suggests that whenever newspapers publish pre-poll surveys, they should take care to preface them conspicuously by indicating the institutions which have carried such surveys, the individuals and organisations which have commissioned the survey,the method of selection of the sample and the possible margin of error in the findings.

The PCI has said that in order to ensure that the voters are not influenced by any external factors, the publication of exit poll surveys should be deferred till after the last poll is held.

Accuracy of 1998 Lok Sabha elections forecast
Seats forecast BJP allies Congress allies NF+LF Others
DRS 249 155 102 37
Outlook/AC Nielsen 238 149 123 33
India Today/CSDS 214 164 127 38
Frontline/CMS 225-235 145-155 120-130 32-52
Acutal result 252 166 96 23
Forecast of 1999 Lok Sabha elections
Seat forecast BJP allies Congress allies NF+LF Others
Timespoll/DRS 332 138 73  
Outlook/CMS 279-289 152-162 63-68 34-39
India Today/Insight 332-336 132-146    
(The agencies have conducted the survey almost a month before the actual date of voting. They would further update the seat forecast in subsequent surveys a few days before the actual polling. The findings of those surveys might give more accurate trend of the voters perception than those surveyed earlier. These polls only suggest the mood of the people and the popular trend at the time of the survey.)


Home Image Map
| Good Motoring and You | Dream Analysis | Regional Vignettes |
Fact File | Roots | Crossword | Stamp Quiz | Stamped Impressions | Mail box |