As Delhi’s winter pollution season sets in, the author — the founder project director of SAFAR — argues that emergency restrictions under GRAP can contain peaks but cannot cure a problem whose sources lie across an entire airshed.

The author notes that this year’s peak pollution season arrived roughly two weeks late. He links this to a strong El Niño phase and to rainfall from a late cyclone, which delayed both the winter and the paddy harvest in Punjab and Haryana.

• Meteorology works both ways: calm, slow winds let pollutants accumulate, while a milder, delayed winter keeps the inversion layer less severe. The author cites NIAS research tying ENSO to North Indian air quality.
• The satellite blind spot: NASA’s MODIS sensors on Terra and Aqua pass over Punjab at fixed times in the late morning and early afternoon. The author cites reports that over 90% of stubble fires in 2024 and 2025 occurred after 3 PM, compared with under 3% in 2021, so official fire counts increasingly undercount.
• A geostationary corrective: South Korea’s GK-2A satellite, with its Advanced Meteorological Imager (AMI), watches the region continuously and detects fire counts that are, per the author, orders of magnitude higher than official estimates.
• Modelling implication: once these fires are included, NIAS modelling attributes roughly 20% of Delhi’s PM2.5 on 1 November to stubble burning. The author’s point is that forecasts are only as good as their emission inputs.

An airshed is a region within which air and its pollutants mix freely because of shared topography and meteorology, regardless of state lines. Delhi lies inside the vast Indo-Gangetic Plain (IGP) airshed, hemmed in by the Himalayas to the north.

• The author argues that measures confined to Delhi tackle only about 50–70% of the sources affecting the city, even before counting stubble smoke.
• A NIAS study under the Office of the Principal Scientific Adviser proposes a framework called “AIRSHEDEM”. The author says it could bring mitigation planning close to full source coverage.
• Its pragmatic middle ground is a planning unit wider than Delhi NCT but narrower than the whole NCR, which the author presents as both scientifically sound and administratively workable.
• The ICU analogy: GRAP is like intensive care for a chronic patient — it stabilises, but treatment must target the underlying disease. It also imposes recurring costs on citizens.

• Strength: it rightly reframes air pollution as a regional, multi-jurisdictional problem, which is the logic that created CAQM in the first place.
• Strength: it flags data integrity — if fire counts fall partly because farmers burn after satellite overpasses, “success” statistics may be misleading.
• Gap: airshed planning needs fiscal and enforcement arrangements across states with different priorities — for instance, farm incomes in Punjab versus air quality in Delhi. The author leaves the institutional design largely open.
• Year-round sources (vehicles, road dust, construction, open waste burning, industry) require sustained rather than seasonal action — a point the author also concedes.
• Federal angle: Entry 6 (Public health) and Entry 14 (Agriculture) sit in the State List, so effective airshed governance depends on cooperative federalism.

On the eve of its tenth anniversary, new research uses pharmacy-level sales data to show that demonetisation led households to cut back on medicines for sudden, short-term illnesses. The authors conclude that liquidity is itself an economic determinant of health.

• Data: the All India Organisation of Chemists and Druggists (AIOCD) database of monthly sales of every stock keeping unit (SKU), across 805 firms, 31 regions and 18 illness categories. The authors describe it as the first product-region-time dataset used to study demonetisation.
• Hypothesis: chronic drugs (e.g. for diabetes, hypertension) are bought monthly by price-aware buyers. Acute drugs (for infections, pain, gastrointestinal ailments) are bought suddenly, briefly and with little shopping around. Under a cash squeeze, the unplanned, unsearched purchase is the one most likely to be deferred. The idea builds on Alan Sorensen’s (2000) work on prescription-drug price dispersion.
• Method: a difference-in-differences design controlling for region-, company- and illness-specific trends.

• Not supply or prices: the number of unique SKUs and the inter-quartile range of prices did not change.
• Not fewer prescriptions: IQVIA prescription data showed no break in doctors’ prescribing.
• Not fewer illnesses: outbreak data from the Integrated Disease Surveillance Programme (IDSP) showed no change in acute disease incidence.
• Tests on product characteristics, inventory channels, physician behaviour and price rigidity also pointed to the search-cost explanation. The authors note that observational data cannot exclude every mechanism.

• Assess welfare costs in advance: whatever the objectives of a currency operation, cash dependence creates similar risks. The concern is distributional impact, not intent. The authors cite evidence of only modest tax gains from the cashless push.
• Public health as a buffer: using HMIS data, they find that regions with fewer PHCs saw larger declines. They suggest building a health-sector protocol — credit lines or digital-payment support at pharmacies — into any future large currency operation.
• Better data: sales data cannot show household health outcomes. The authors call for linked consumption–health datasets and pre-registered evaluations of extreme policy instruments.
• Balance sheet: they acknowledge that demonetisation accelerated digital adoption and reduced counterfeit stock, but argue that its costs were real for households on the margin.

• Hidden costs of macro decisions: the study shows that monetary and currency policy can have health effects, and that aggregate indicators can hide them.
• Inclusion lens: digital rails help, but households without smartphones, connectivity or bank access remain cash-dependent. A shock’s burden falls on the poorest and least-served.
• Evidence-based policymaking: the call for ex-ante impact assessment and pre-registered evaluation fits the wider case for regulatory impact analysis in major executive decisions.
• Caveat: the effect is a modest, relative decline measured from sales. Without outcome data, it shows reduced access, not measured health harm.