Beyond the exit poll: Can SCORE restore faith in India's election surveys?

At a time when election forecasting is battling a crisis of credibility, a new public-opinion platform promises to measure the political mood constituency by constituency.

Indian politics has never suffered from a shortage of predictions. What it increasingly lacks is confidence in those making them.

The broad problem is that election surveys have increasingly become television events rather than exercises in understanding society. Credible institutions and serious researchers must compete for attention with lesser-known operators offering extravagant forecasts, often without adequate public disclosure of their methods. The distinction between scientific research, political consultancy, and televised speculation has become disturbingly blurred.

It is against this backdrop that a new initiative called SCORE has entered the field.

Launched in Bengaluru on October 16, SCORE proposes a departure from the prevailing election-season approach. Its stated ambition is to measure public opinion continuously, constituency by constituency, rather than attempt to decipher the electorate only when polling day approaches.

According to its founder, Naresh Arora: "SCORE is India's first and only opinion polling organisation that will measure public opinion not for media or politicians but directly for the public so that the ordinary voters can be better informed about the issues that matter."

FROM ELECTION-DAY PREDICTIONS TO A PERMANENT POLITICAL BAROMETER

India's opinion-polling tradition has a distinguished intellectual history, even if its contemporary television incarnation sometimes obscures it.

The pioneering work of Eric D Costa and later Prannoy Roy helped introduce scientific election analysis to a wider Indian audience. The Centre for the Study of Developing Societies (CSDS), particularly through its Lokniti programme, developed a substantial tradition of studying electoral behaviour, social identities, political participation, and changing public attitudes.

Researchers such as Yogendra Yadav and Sanjay Kumar helped establish the importance of understanding why Indians vote as they do, rather than merely predicting which party will win.

Commercial organisations, including CVoter, Axis My India, and others, subsequently expanded the reach and visibility of polling. Some developed tracking exercises and large field networks. The record, however, has been uneven, and the pressure to translate complicated findings into definitive seat projections has frequently overwhelmed the more nuanced conclusions that surveys can offer.

SCORE is, therefore, not inventing continuous opinion research. Nor is constituency-level analysis unprecedented. Its potential contribution lies in bringing together regular tracking, granular constituency data, and direct public access within a single platform.

Its first exercise covers all 117 Assembly constituencies of Punjab, where elections are due in 2027. According to its launch announcement, researchers interviewed more than 28,000 voters between September 13 and October 13, 2026.

The organisation plans to release a fresh Punjab tracker on the 15th of every month, with an Uttar Pradesh tracker scheduled to begin on November 20. Results are intended to be accessible through its website and WhatsApp, allowing ordinary citizens to examine political trends beyond the conventional television debate. With a sleek, modern, and accessible interface, SCORE makes it easy for anyone to understand and analyse its various parameters.

In the larger context, an opinion poll that tells a voter only who is likely to become Chief Minister has limited civic value. One that reveals how constituents assess their MLA, which public services they consider inadequate, whether government schemes are reaching beneficiaries, and how these perceptions change over time can serve a larger democratic purpose.

There is, of course, an important difference between measuring public opinion regularly and measuring it reliably. A monthly survey does not automatically become more scientific simply because it is repeated. But a carefully designed series can reveal shifts that isolated surveys miss, distinguish enduring political preferences from temporary reactions and provide an opportunity to test earlier findings.

That, rather than the ability to predict a winner months in advance, ought to be the more consequential promise of SCORE.

PUNJAB'S VERDICT BEFORE THE VERDICT

SCORE's inaugural Punjab survey offers a revealing picture of a state where the ruling Aam Aadmi Party (AAP) appears electorally stronger than the public's assessment of some aspects of its governance might suggest.

If elections had been held on October 15, the survey estimates that AAP would have secured between 69 and 82 seats, comfortably crossing the majority mark of 59, although falling short of the 92 seats it won in 2022.

Its estimated vote share of approximately 38% represents a decline from the 42% it secured in the previous Assembly election. The Congress trails at 24%, with a projected 17-34 seats. The BJP, at 14%, is estimated to win 4-12 seats, while the Shiromani Akali Dal, with 12.5%, is projected to secure 2-6 seats.

An intriguing feature is the emergence of the Amritpal Singh-led Akali Dal (Waris Punjab De), which the survey places at 8.5%, with a projected 2-6 seats. Its reported support among younger men, particularly those below 35, points to the fragmentation of the traditional Akali constituency and the possibility of new political alignments.

The numbers, if borne out, suggest that AAP's principal advantage may lie as much in the disunity of its opponents as in the enduring appeal of its own government.

Punjab's regional variations reinforce this interpretation. AAP's estimated support is strongest in Doaba, at 43.9%, followed by Majha at 40.8%. In Malwa, which makes up 69 of the state's 117 Assembly constituencies, its vote share falls to 35.2%.

Malwa consequently emerges as the most consequential battleground. The survey identifies 22 constituencies where narrow margins could influence the final tally, including six effectively tied contests.

But the more interesting findings concern the relationship between political leadership, welfare delivery, and dissatisfaction with governance.

Chief Minister Bhagwant Mann enjoys a personal approval rating of 63.9%, compared with 57.7% for his government. Among respondents who expressed a Chief Ministerial preference, Mann commands 46.3%, ahead of Congress leader Charanjit Singh Channi at 18.3%, Sukhbir Singh Badal at 12.3%, and Amritpal Singh at 8%.

Only 48.1% of respondents express satisfaction with their sitting MLA, while 42.4% favour re-electing their representative.

The gender divide is even more striking.

According to SCORE, AAP enjoys 51.2% support among women, compared with just 27.4% among men. The Congress, meanwhile, is almost level with AAP among male voters, at 26.6%. Among women aged above 60, AAP's support rises to 57.8%.

Such a difference deserves closer examination. It raises questions about the role of welfare, household economics, political identification, and the growing independence of women's electoral choices.

The survey reports that nearly 80% of households have benefited from at least one state government scheme. It records substantial reach for free electricity, ration distribution, and the Mawan Dheeyan Satkar cash-transfer programme.

These are reported survey responses, rather than independently verified administrative figures. Nevertheless, they suggest why welfare delivery may be an important electoral asset for the incumbent government.

The contradiction becomes sharper when respondents are asked about the state's problems.

Drugs emerge as the single largest local concern, identified by 39.3% of those surveyed. In Majha, the figure rises to 49%. Roads, drainage, sanitation, water, unemployment, and inflation also feature prominently.

For a government seeking another term, the findings present a complicated picture. Welfare appears to have created goodwill, but it has not extinguished anxieties about social conditions, public infrastructure, or local administration.

For the Opposition, the challenge is equally formidable. Public dissatisfaction does not automatically translate into electoral consolidation. Unless the Congress, BJP, Akali Dal, and newer political formations can convert grievances into a coherent alternative, the incumbent may retain its advantage despite measurable discontent.

Perhaps the most important qualification is that SCORE identifies close to 23% of voters as undecided. With several months remaining before the election, this is hardly an electorate whose final intentions can be treated as settled.

A survey taken in October 2026 cannot tell us with certainty what Punjab will decide in 2027. What it can do is establish a benchmark against which subsequent political developments may be measured.

That is where SCORE's proposed monthly exercise could become genuinely useful.

CAN PUBLIC OPINION BECOME A PUBLIC INSTITUTION?

There is a larger democratic argument for taking initiatives such as SCORE seriously.

India's electoral discourse is dominated by political parties, television channels, social-media campaigns, and an expanding ecosystem of political consultants. The ordinary voter, although repeatedly invoked as the ultimate sovereign, is frequently reduced to a number in somebody else's narrative.

The most valuable contribution of sustained public-opinion research would be to reverse that relationship.

Imagine a system in which citizens could regularly examine how their constituency rates its elected representative, whether local dissatisfaction is increasing, which government programmes are reaching households, and which problems remain unresolved.

Such information could potentially improve the quality of political debate. Opposition parties would have to respond to identifiable public concerns rather than merely attack incumbents. Governments would have an additional means of detecting discontent between elections. Journalists could test political claims against evidence instead of relying exclusively on partisan assertions.

There are limits, of course. Public-opinion surveys cannot replace elections, legislative scrutiny, independent journalism, or the everyday accountability of elected representatives. Nor should a survey be mistaken for the considered judgment of an entire constituency.

For SCORE, the real test will not be whether its first Punjab projection eventually proves correct. One accurate forecast, just like one inaccurate forecast, is insufficient to establish the credibility of an institution.

The more meaningful test will emerge over successive rounds. Will its November findings capture identifiable shifts from October? Will changes in party support correspond with developments on the ground? Will its constituency-level estimates prove more informative as the election approaches? And will its researchers explain what changed, why it changed, and where uncertainty remains?

The same questions will apply when the proposed Uttar Pradesh tracker begins. The state's enormous electorate, varied social coalitions, and highly localised political contests will present an even more demanding test of the platform's capacity to measure public sentiment.

For now, SCORE's arrival is encouraging precisely because Indian psephology needs fresh thinking, greater methodological openness, and a renewed sense of public purpose. The experiment will be worth watching, particularly at a moment when the credibility of election forecasting has suffered repeated setbacks.

There was a time when election surveys helped Indians understand the country's extraordinary political diversity. Somewhere along the way, the search for understanding became entangled with the race to announce winners.

SCORE's most worthwhile ambition would be to restore that original purpose.