Voter intelligence
No single issue decides an election. Voters weigh household finances, public services, identity, trust, leaders and local candidates against one another, and the weights shift between countries, regions, constituencies, elections and households.
This page sets out the framework Polity Lab uses to study voter choice: what voters weigh, how a concern turns into a vote, and how to measure the difference between the two. The framework applies to any democracy; the evidence comes from India, where much of our work is based.
What voters weigh
This is an inventory, not a ranking. Each dimension comes with a question the voter is, in effect, asking. Something can matter a great deal in a voter's life without deciding how they vote.
Income, prices, debt and savings.
Is my household better off or worse off?
Whether work exists, how secure it is and what it pays.
Can I find work, keep it or earn more?
Crop prices, input costs, irrigation, procurement and land.
Is farming becoming more viable or more precarious?
Pensions, food support, cash transfers, housing and health cover.
Did I get what was promised, and will it last?
Water, power, roads, transport, sanitation, schools and clinics.
Has everyday life become easier?
Integrity, competence, reputation and how easy they are to reach.
Can this person solve problems, and can I hold them to account?
Trust, record in government and the credibility of the leader.
Who can govern well?
Caste, ethnicity, language or kinship; representation, status and inclusion.
Does this arrangement recognise and protect people like me?
Belonging, values, freedom of belief and recognition.
Does this party or candidate stand for what matters to me?
Crime, harassment, policing and access to justice.
Do I feel safe, respected and protected?
Bribes, favouritism and how officials treat people.
Is the system fair to people like me?
Schooling, health, jobs and the chance to do better.
Will my children have better chances than I had?
Investment, infrastructure and a fair share for the area.
Is my area getting its fair share?
Security, the economy, stability and standing in the world.
Is the country heading where I want it to go?
Past promises, past performance and broken commitments.
Should I believe these people again?
Evidence from India
Two large Indian surveys agree on the headline: livelihoods come first. Shortly before the 2024 Lok Sabha election, more voters named unemployment as their most important issue than anything else, with rising prices close behind.
Share of respondents, %. Source: Lokniti-CSDS pre-poll survey, 2024 (10,019 respondents in 19 states), as published in The Hindu, 11 April 2024.
Rural India · 2018
Share of rural respondents naming each issue among their priorities, %. Respondents could name several. Source: ADR, All India Survey on Governance Issues and Voting Behaviour 2018 (October to December 2018).
Urban India · 2018
Share of urban respondents naming each issue among their priorities, %. Respondents could name several. Source: ADR, All India Survey on Governance Issues and Voting Behaviour 2018 (2,73,487 respondents in 534 Lok Sabha constituencies).
Read with care
A stated priority is not a cause. Someone can name unemployment as the country's biggest problem and still vote on the local candidate, on a leader or on a long-held loyalty. Surveys like these tell us what is on voters' minds. They do not, on their own, tell us what moved their votes.
Five measurements
Most voter research blurs five separate things. Keeping them apart is the difference between knowing what voters worry about and knowing how they will vote.
The decision process
A working model in six stages. It is a way of testing behaviour, not a claim that every voter moves through six tidy steps.
The voter meets a price rise, a closed clinic, a new road, a job offer or a personal problem.
They decide whether it is good or bad, fair or unfair, temporary or lasting.
They decide who is responsible, and who actually has the power to fix it.
They compare the record, credibility and likely conduct of the alternatives on offer.
They weigh material interests against identity, ideology, leadership, the candidate and trust.
They back a party or candidate, vote tactically for whoever can win, or stay at home.
It rarely runs in a straight line
Many voters start with a party preference and read everything else through it. Others change their minds after a personal setback, a change in household circumstances or the arrival of a credible alternative. Some decide months ahead, some during the campaign, and some only near polling day.
Rural and urban
Rural and urban voters often care about the same broad things. What differs is how an issue shows up in daily life, whom they blame for it and how much it counts. These are tendencies to investigate, not fixed profiles.
One policy, two meanings
In a village, electricity can mean water for the crop and light in the evening. In a city flat or workshop, it can mean steady supply, a business that keeps running and the cost of keeping cool. The policy heading is the same; what is judged, and how, is not. And geography is no substitute for economics: a landless labourer and a prosperous farmer in the same village may see the same election very differently.
Welfare and delivery
Receiving a benefit and experiencing it are different things. Two households can receive the same entitlement and judge the government behind it quite differently, because one was treated reliably and with dignity and the other met delays, exclusion and petty humiliations.
In India, research on the 2024 National Election Study finds that welfare recipients weigh their experience of getting a benefit, not only whether they received it (Kailash, 2024).
Trade-offs
The hardest questions in voter research begin when two things a voter cares about point in opposite directions. Which way the balance tips differs between constituencies, and between elections in the same constituency, so it has to be measured each time rather than assumed.
From views to votes
Beyond the fifteen dimensions, several mechanisms decide how much each one counts on the day.
The present is judged against remembered governments, old promises and how things used to be.
A modest improvement can satisfy a voter who expected little and disappoint one who was promised a lot.
Is this politician competent, honest, responsive and likely to follow through?
Governments win credit and blame for things they only partly control, and voters may separate responsibility from the power to act.
Family, neighbours, colleagues and community bodies inform choices, though their pull varies and should not be presumed.
Media shape what voters know and which issues feel urgent. Exposure alone is not persuasion.
Voters choose from what is offered. Candidate selection, alliances and who looks able to win can change the choice.
Registration, migration, work, health and the distance to the polling station decide whether a preference becomes a vote.
India · several levels at once · 2018
Share of respondents rating each reason important when voting for a candidate, %. The report ranks the candidate's caste or religion fourth of five but gives no percentage in its text. Source: ADR, All India Survey 2018.
India · whose opinion counts · 2018
Share of respondents naming whose opinion mattered most in deciding their vote, %. Self-reported. Source: ADR, All India Survey 2018.
India · turnout by sex · 2009 to 2024
Lok Sabha turnout among registered electors, %. Source: Election Commission of India, official results.
Measuring it properly
A serious study measures each part of the decision separately, then tests how the parts interact. These are the variables Polity Lab builds into voter research.
In the field
Good interviews start from concrete experience, not abstract opinions. These questions separate what has happened to people, what they think caused it, whom they trust and what might change their mind, without putting answers in their mouths.
Common errors
Each of these can make a confident analysis wrong. We check our own work against them.
For strategy
The useful question is never just what voters care about. It is which factors matter, to whom, under what conditions, and whether they change a vote or only the way a voter explains it. That is the question Polity Lab's research is designed to answer.
Measure what local voters weigh, instead of borrowing national averages or last election's story.
Separate the issues voters worry about from the ones that could actually change their choice.
Know which level of government voters hold responsible before deciding whom a message should address.
Treat every assumption about voters as a hypothesis, and check it with evidence before resources are committed.
Sources and further reading
The framework is a set of hypotheses to test, not a universal ranking of voters' motives. These are the sources for the figures above and good starting points for anyone who wants to go further.
