Delhi-NCR’s air pollution response has hardened into a seasonal ritual, with emergency measures repeatedly invoked but little evidence of lasting impact.
The Graded Response Action Plan (GRAP) has drifted from a corrective framework into a performative exercise, marked by delayed escalation and weak public explanation.
As pollution exposure becomes continuous rather than episodic, public protest reflects democratic frustration rather than panic.
Data abundance has not translated into accountability, with dashboards proliferating even as access to auditable datasets declines.
Until clean air governance delivers measurable outcomes, transparent data and proportional responsibility, dissent will remain a recurring feature of Delhi’s winters.
I moved to Delhi from San Diego in 2021, at the start of winter, and by chance found myself working on air pollution and sustainable mobility. For the first time, I was not just reading about the national capital’s air crisis, I was breathing it.
What struck me almost immediately was how thin the line was between perceived action and lived reality.
One early winter morning, the city air slipped from “poor” to “very poor” category on the Air Quality Index (AQI) almost overnight. Alerts poured in, news reported that high level meetings were scheduled, and advisories circulated. On paper, the system appeared responsive. But when I tried to understand what would actually change as pollution crossed each threshold, the answers were vague. Measures would be announced, lifted and reintroduced, often without clarity about what they were meant to achieve or how success would be judged.
I assumed this ambiguity would be temporary. Four winters later, it has hardened into routine.
Every year, Delhi-National Capital Region responds to polluted air with a familiar script: Construction bans, traffic restrictions, work-from-home advisories and the familiar ladder of the Graded Response Action Plan, or GRAP. The rituals are precise, the outcomes, on the other hand, are not. When winter passes, restrictions recede. And the bad air returns the following year, just as unforgiving.
At some point, repetition stops being coincidence and becomes governance.
GRAP was designed as a graded, evidence-linked framework, escalating action as pollution worsens. In practice, it has increasingly become a seasonal ritual rather than a corrective mechanism. This winter was no exception. Pollution levels have repeatedly entered the “severe” category this year, yet escalation remained cautious and delayed. Lower-order measures were preferred even as health risks mounted.
What was striking was not only which GRAP stages were invoked, but how little public explanation accompanied these choices. As emergency frameworks are repeated without clear criteria, timelines or outcomes, they lose their signalling power. What were meant to be warnings become background conditions.
This year, that background condition intensified.
Pollution episodes were longer, peaks more persistent and recovery periods shorter. Poor air quality was no longer confined to early mornings or outdoor exposure. It extended through the day and into enclosed spaces. Indoor concentrations rose despite mitigation measures, and commonly relied-upon buffers such as limited outdoor activity, masks and air purifiers proved increasingly inadequate.
This shift matters because it altered who was exposed and for how long.
Public protest did not emerge simply because pollution occurred. It emerged when pollution breached the thresholds that had historically insulated the most protected residents. When exposure became continuous rather than episodic, and personal mitigation ceased to offer meaningful relief, air pollution moved from an inconvenience to an inescapable risk.
Protest, in this context, is not panic or disruption. It is a response to systemic stagnation. In a democracy, dissent arises when institutions fail to demonstrate learning. Years of data, committees and plans have not translated into visible, sustained improvement, nor into clarity about long-term strategies, accountability or the distribution of sacrifice.
Citizens are repeatedly asked to comply with restrictions without being shown whether these measures work, how success is measured, or whose behaviour is being regulated most aggressively. When this cycle repeats year after year, public dissent becomes the only remaining audit mechanism available outside the state.
For many in Delhi, however, polluted air has never been episodic. The impacts of air pollution exposure have always been a year-long affair, particularly for marginalised communities. Outdoor workers, informal commuters and residents of high-exposure neighbourhoods have always experienced pollution as a permanent condition. They cannot retreat indoors on “bad air days”, and in many cases, indoor exposure differs little from outdoor exposure. Over time, air pollution control has become gentrified, with mitigation strategies protecting those with resources while normalising chronic exposure for those without. Their pollution exposure is the direct result of what policy delays have allowed.
Environmental crises in unequal cities tend to become urgent only when protective buffers fail for the privileged. That delayed recognition does not invalidate today’s protests, but it does explain why trust in institutions is fragile and uneven.
A similar pattern of discretion is visible in how air pollution data itself is governed.
India does not lack air pollution data. What it increasingly lacks is accountable data governance.
Dashboards now dominate how air quality is communicated. Real-time AQI numbers flash across screens, but consistent historical datasets are difficult to access. Bulk downloads are fragmented. Station-level metadata, calibration histories and monitoring changes are rarely disclosed. Application programming interfaces meant to ease access to raw data exist unevenly — and where they do, they are unstable or poorly documented.
This matters because data opacity is not neutral. It shapes what questions can be asked, who can ask them, and where responsibility ultimately lands.
The problem became particularly evident with the National Clean Air Programme’s PRANA portal. When it was launched in 2021, it felt like a rare opening. City action plans and funding details were visible. Early analysis revealed what many had suspected; a large share of funds was flowing into road widening, mechanical sweepers and water sprinkling. These were visible actions, but ones with limited and often unverifiable long-term impact. Measures aimed at source control and enforcement capacity received far less attention.
The expectation was that regular updates would enable course correction. Instead, reporting became irregular. Over time, several dashboards and reports that had once been accessible were no longer consistently available during portal transitions. Whether by design or neglect, transparency receded just as scrutiny became sharper. For a programme built on the promise of data-driven action, this silence was telling.
Meanwhile, even as raw data became less accessible, reports proliferated. Clean air action plans, source apportionment studies and progress reports are released with regularity. Many rely on outdated emissions inventories and broad recommendations that sound reassuring but are difficult to test or challenge. The result is a strange paradox. We are surrounded by information, yet starved of accountability.
Recent parliamentary calls to update India’s air quality standards and strengthen monitoring networks signal an important moment. India’s ambient air standards have not been revised since 2009; revisiting them is overdue.
But the AQI is not just a technical tool. It defines what level of risk the state is willing to accept for its citizens.
Despite these limitations, some datasets continue to offer uncomfortable clarity. Continuous monitoring stations still tell us what people are breathing. Satellite observations show how pollution moves across the Indo-Gangetic Plain, ignoring city and state boundaries. Fire detection data leaves little ambiguity about seasonal biomass burning. Vehicle registrations, fuel sales and power generation figures quietly reveal whether policy narratives match reality.
Together, these datasets allow us to ask difficult questions. Are pollution peaks actually declining? Are we investing in the sectors that contribute the most? Are public funds being spent where they can deliver lasting change?
That citizens and researchers must stitch these answers together themselves is not a technical flaw. It is a governance choice to allow a modified reality and translucent accountability.
When accountability weakens, something else rushes in to fill the gap. In the absence of stable, auditable data, air pollution discourse drifts easily into blame.
This has been visible in media debates for years. My colleagues — and more recently I — have participated in air pollution coverage since before the pandemic, and one pattern has remained stubbornly consistent. When scientific evidence is filtered through political narratives, discussions shift from what must change to who must be blamed. And as long as blame circulates, structural reform remains optional.
The contrast between 2019 and 2025 discourses makes this clear. In 2019, air pollution debates were dominated by a predictable and polarised blame cycle. Seasonal stubble burning in neighbouring states became the central villain. Public health emergencies were declared whenever AQI repeatedly breached the “severe” category. Experts pointed to multiple contributors, but political discourse focused on deflection and mutual accusation. The outcome was a stalemate.
By 2025, the evidence has shifted. Stubble burning declined sharply, and satellite data, source apportionment studies and emissions inventories forced a pivot. Vehicular emissions and year-round urban sources could no longer be ignored, and airshed management entered the vocabulary.
And yet, even as the science became clearer, debates continued to circle back to the same question: Where does the buck stop?
This persistence of the blame game is not accidental. Fragmented data allows fragmented responsibility. When datasets are incomplete, outdated or inaccessible, accountability becomes negotiable. Each actor can claim partial truth, selective uncertainty or procedural constraint.
As long as blame can be externalised, governance can remain reactive.
This is how Delhi’s air crisis gets trapped in firefighting mode. Emergency measures substitute for structural reform. GRAP stages are invoked and withdrawn. Construction bans come and go. Advisories are issued. Dashboards update. But the underlying emissions trajectory remains largely untouched.
Delhi does not need more emergency responses, more action plans or more dashboards. It needs fewer patchwork actions. Clean air governance must move beyond activity lists to measurable, sector-wise outcomes. Funding must follow emission sources, not administrative convenience. Responsibility must be proportionate to consumption and exposure, not political leverage.
Most importantly, data must return to its democratic purpose: enabling citizens to see, question and hold authority to account.
Delhi’s air crisis is not a mystery. We know what pollutes the city. We know who bears the burden. We even know what would work. When people take to the streets, they are not rejecting the credibility of grounded, science-backed actions. They are reacting to a system that knows better, measures endlessly, but acts selectively.
Without visible and continuous accountability, dissent will remain a recurring feature of every winter, signalling unresolved institutional failure.