This page explains how the Indian Taxpayer Voice survey works, how responses become published statistics, and what the results can and cannot tell you.
Purpose
To understand the concerns reported by people who choose to take part, about job-loss protection, health-related costs and public spending, broken down by state or union territory and broad employment sector.
The questionnaire
Version 1, currently in a pilot that is not yet open to the public; dates will be recorded here when it opens. After a consent question there are 11 questions. Three are required: state or union territory, broad employment sector, and main concerns (more than one can be chosen). The rest are optional, and each personal question offers “Prefer not to say”: employment type, income band, gender, age band, financial buffer, health cover, priorities, and a free-text comment.
Income is asked as a broad band of annual individual income before tax: up to ₹4 lakh, ₹4–12 lakh, ₹12–24 lakh, ₹24–50 lakh, ₹50 lakh–₹1 crore, and above ₹1 crore. The boundaries follow the income-tax slabs (₹4, ₹12 and ₹24 lakh) and surcharge thresholds (₹50 lakh and ₹1 crore), and are re-checked against the rates in force before each survey version.
Who can take part
Anyone aged 18 or older who chooses to respond. Recruitment channels will be listed here as they are used, so readers can judge who is likely to have seen the survey.
Anonymity
The survey stores no name, contact details, IP address, invitation code, or date and time with a response. State and union territory are as reported by respondents; we never infer location from IP addresses. Free-text comments are never published word for word. See the privacy notice for details.
Protecting the survey from automated responses
Before the survey opens, each visitor’s browser completes a short proof-of-work check, and submissions are rate-limited. This makes large-scale automated submission costly; it cannot prove that each respondent is a unique person, and determined manual repetition remains possible. Patterns that suggest abuse are investigated, and affected responses are removed and the change recorded here.
Which responses count
- Only completed responses with consent are counted; partly completed responses are not.
- Responses with answers outside the questionnaire’s options are excluded.
- Repeated submissions of the same response record (for example, after a network retry) are counted once.
How figures are calculated
- Percentages use valid responses as the denominator unless stated otherwise.
- Published counts are rounded to the nearest 5, and percentages to whole numbers calculated from the rounded figures, so totals may not add up exactly. This, together with hourly updates, makes it hard to work out a single person’s answers by comparing one update with the next.
- Respondents may select more than one concern, so concern percentages can add up to more than 100%.
- In the sector-by-concern table, percentages are the share of that sector’s respondents.
- Groups with fewer than 20 responses are not shown. Some larger groups are also hidden so that a small group cannot be worked out by subtraction.
- States and union territories with no responses are shown as “No responses”, which is different from “Not shown”.
- Because the survey records no dates, trends over time are the weekly change in the published total between scheduled updates, never individual response times. Weeks with fewer than 20 new responses are not shown, and the first week includes any earlier responses.
Limitations
This is a voluntary online survey. People who choose to respond may differ from the wider population, for example in internet access, language, interest in the topic, or how they heard about the survey. Results describe respondents only and are not statistically representative of Indian taxpayers, however many people take part.
Refresh schedule
Published statistics are recalculated every hour and show the time of the last update. If an update looks wrong, for example a sudden large drop in responses, the previous figures stay published until it has been checked.
Corrections
If we find or are told about an error in a figure, a method or this page, we correct it and add a dated note here describing what changed and why. Report an error to hello@indiantaxpayervoice.org.