Every October, as temperatures drop across the Indo-Gangetic Plain, a familiar ritual begins. Schools shut for a few days, flights are delayed, hospitals see a spike in respiratory admissions, and social media fills up with photographs of monuments disappearing into a grey shroud. By December, the outrage fades along with the headlines – until the cycle repeats the following winter. This annual amnesia is, in many ways, the biggest obstacle to solving India’s air pollution crisis. Pollution is not a winter problem or a Delhi problem. It is a year-round, subcontinent-wide public health emergency that we have chosen, for too long, to treat as a seasonal inconvenience.
The numbers should unsettle us into a different kind of urgency. The World Air Quality Report 2025 placed India as the sixth most polluted country globally, with a population-weighted average PM2.5 concentration of 48.9 µg/m³ – nearly ten times the World Health Organization’s annual guideline of 5 µg/m³. There is a silver lining buried in that statistic: it represents a 3% improvement over 2024 and a 10% decline from 2023. But modest gains at the national average conceal a harsher reality on the ground. Seventeen of the world’s twenty most polluted cities in 2025 were located in the Central and South Asian region, the overwhelming majority of them in India’s own Indo-Gangetic Plain.
Effectively, the entire population of this country – all 1.4 billion of us – lives with exposure to PM2.5 levels the WHO considers unsafe. The World Bank has estimated the resulting economic toll at around $36.8 billion a year, from lost productivity, premature deaths, and pollution-linked illness, a figure that likely understates the true cost once long-term health burdens on children and the elderly are factored in.
I write this not as a distant observer but as someone who works, day to day, on the ground with one of India’s most ambitious state-level attempts yet to fix this problem at scale. And it is precisely this vantage point that makes me want to be honest about both the promise and the limits of our current approach.
National Clean Air Programme’s uncomfortable report card
Launched in 2019, the National Clean Air Programme (NCAP) set out to reduce PM10 concentrations by 20–30% across 131 non-attainment cities by 2024, later revised to a 40% target by 2025–26. Seven years on, independent assessments make for sobering reading. Of the 100-odd NCAP cities with adequate monitoring data, only around 23 have achieved the revised 40% reduction target, and 23 cities have actually recorded an increase in particulate levels since the programme began. Even among the cities that have shown improvement, most still exceed national air quality standards. Researchers tracking the programme have been blunt: the original targets are, in their words, no longer achievable within the programme’s lifetime.
Part of the problem is structural. NCAP funding has disproportionately favoured interventions that are easy to plan and report – principally road-dust suppression, which absorbed roughly two-thirds of allocated funds – while the sectors doing the most damage to human health receive a fraction of that attention. Biomass burning, vehicular emissions, and industrial pollution together account for a large share of the disease burden from PM2.5, yet industrial emission control alone has attracted barely 1% of funding. We have, in effect, been sweeping the visible dust off the streets while the invisible, more lethal particles from tailpipes, chimneys, and cooking fires continue largely unchecked. Compounding this, only around 90 of 131 non-attainment cities have completed source-apportionment studies after seven years of the programme – meaning roughly forty cities are still designing pollution control strategies without reliable data on what, precisely, is polluting their air.
Why airsheds, not city limits, must define the next phase
If there is one lesson that ought to reshape the next phase of India’s clean air effort, it is this: air pollution does not respect municipal or even state boundaries. A power plant in one district, stubble burning in another, and vehicular congestion in a city hundreds of kilometres away can all converge to choke the same airshed. This is the premise on which Uttar Pradesh – home to seventeen non-attainment cities and a population larger than most countries – has built its own state-level clean air programme, in partnership with the state government and the World Bank.
This programme represents a genuine departure from the city-by-city playbook that has defined NCAP’s first phase. It is structured around three pillars: strengthening state-level institutional capacity for air quality management, driving sector-specific emission reductions in transport, agriculture, livestock, household cooking, and industry, and – critically – enabling cooperation across the wider Indo-Gangetic Plain airshed. The nearly $300 million World Bank-financed effort aims to bring clean cooking access to 3.9 million households, put 15,000 electric three-wheelers and 500 electric buses on the roads of Lucknow, Kanpur, Varanasi, and Gorakhpur, and replace over 13,500 heavily polluting vehicles with cleaner alternatives. It is also piloting tools that previous efforts largely ignored – AI-enabled decision support systems, tunnel-kiln technology for brick manufacturing, and a network of scientific “supersites” to finally answer the basic question NCAP has struggled with for years: where, precisely, is the pollution coming from, and which interventions actually move the needle?
I do not raise this experience to suggest Uttar Pradesh has solved anything yet – implementation is only beginning, and history offers plenty of reasons for caution. But the design principle behind it deserves to become the template for the rest of the country, not a state-specific experiment. Haryana has since received similar World Bank financing for an airshed-based programme of its own. Punjab, Delhi, and the remaining IGP states must follow, with genuine cross-border coordination rather than the largely symbolic regional planning that has characterised discussions so far.
Three shifts that would matter more than any single new scheme
First, India’s clean air policy must finally prioritise PM2.5 over PM10. Fine particulate matter penetrates deep into the lungs and bloodstream and is responsible for the overwhelming share of pollution-linked disease, yet our monitoring, targets, and funding formulas remain oriented around the coarser, less lethal PM10. Continuing to chase PM10 targets while PM2.5 does the real damage is a policy choice we can no longer defend on health grounds.
Second, enforcement against industrial and vehicular emissions needs teeth, not further extensions. Deadlines for thermal power plants to meet emission norms have been pushed back multiple times over the past decade. Every extension is a signal to polluters that compliance is negotiable. Meanwhile, small and medium industrial units – numerous, dispersed, and often under-monitored – remain the hardest part of the emissions inventory to regulate, precisely where ongoing efforts to help roughly 2,000 MSMEs in Uttar Pradesh adopt cleaner technology could prove instructive if scaled nationally.
Third, we need honesty about what performance-linked funding actually measures. Cities have received grants tied to air quality improvement, yet fund utilisation has often stayed below 60%, and the metrics used to judge “improvement” have sometimes rewarded activity over outcomes. A rupee spent damping down road dust is not equivalent, in health terms, to a rupee spent helping a household switch from biomass to clean cooking fuel or helping a farmer manage crop residue without burning it. Our funding architecture should reflect that difference explicitly.
The politics of patience
None of this is quick, and that is precisely the difficulty. Clean air policy in India operates on a timescale that outlasts election cycles, budget cycles, and the news cycle’s attention span. The temptation, every winter, is to reach for visible but largely cosmetic responses – smog guns, water sprinklers, temporary construction bans – because they photograph well and satisfy public anger in the short term. The harder, less visible work of building airshed governance, financing clean cooking transitions for tens of millions of households, and holding industry to enforceable emission timelines does not generate the same immediate optics. But it is the only work that will actually move the needle a decade from now.
Uttar Pradesh’s experiment, and Haryana’s alongside it, will be watched closely – by other Indian states, and by neighbouring countries in South Asia facing an identical airshed problem. If the airshed approach can demonstrate measurable PM2.5 reductions over the next five years, it will offer NCAP’s next phase something it has lacked from the start: proof that a different model works. If it cannot, we will have learned that too, and at considerable cost. Either way, the country cannot afford another seven years of well-funded activity that does not add up to genuinely cleaner air. The 1.4 billion people breathing this air every day deserve a policy architecture built for that outcome, not merely for the appearance of one.
About the Author
