Post Content
Santosh Harish leads the Air Quality Fund at Coefficient Giving, a global philanthropic advisor and funder that has directed more than five billion dollars in grants towards health, scientific research, pandemic preparedness and potential risks from advanced AI.
Formerly Open Philanthropy, Coefficient Giving’s foundational donors are Dustin Moskovitz, co-founder of Facebook and Asana, and philanthropist Cari Tuna.
At the fund, Santosh focuses on particulate air pollution and lead exposure. He was earlier a researcher in air quality governance at the Centre for Policy Research, New Delhi and the India Center of the Energy Policy Institute at the University of Chicago.
He holds a B.Tech. from IIT Madras and a PhD in Engineering and Public Policy from Carnegie Mellon University.
Santosh spoke to indianexpress.com on the opportunities for tech interventions in the air quality space, the learnings from their grants, and the challenges the air quality innovation ecosystem faces in the country. Edited excerpts:
Must Read | Don’t drop everything just to do AI: Tech Matters’ CEO Jim Fruchterman warns nonprofits
Venkatesh Kannaiah: Tell us broadly about the journey of Coefficient Giving and the themes you engage in.
Santosh Harish: Coefficient Giving began in 2011 as a collaboration with GiveWell, the charity evaluator. Our founding and most significant partnership has been with Good Ventures, a foundation led by Dustin Moskovitz and Cari Tuna.
Story continues below this ad
We have directed more than $5 billion in grants across 13 focus areas. We focus on causes that are important, neglected, and tractable — areas where additional resources can make a difference.
Air quality is a relatively newer programme for us. So far, we have committed more than $35 million to air quality. This is still an area where we are learning, especially around the absorptive capacity and where funding can have the most impact.
Over the last few years, we have increasingly worked with other donors as well. We recently launched two major collaborative funds: the Abundance in Growth Fund and the Lead Exposure Action Fund.
Venkatesh Kannaiah: Tell us about your India engagement strategy.
Santosh Harish: The South Asian air quality programme was launched in January 2022. The health burden of air pollution in India is enormous. It is responsible for around two million premature deaths annually, according to the Global Burden of Disease estimates.
Story continues below this ad
Beyond mortality, there are also major impacts on cognitive development and productivity, and a disproportionate share of these costs is borne by children.
Also Read | ‘Tech for good ideas’ are fine, but it is impact at scale that matters: Anjali Bansal, founder, Avaana Capital
We also felt that, relative to the scale of the problem, philanthropic and aid funding for air pollution was limited. We looked at whether philanthropic funding could make a difference, and now we think it can deliver.
We work on improving measurement and strengthening cost-effective air pollution monitoring systems. We have invested in modelling, which helps identify pollution sources and determine policy priorities by comparing different intervention scenarios. Our grantmaking has therefore spanned a wide range of interventions.
For example, on stubble burning, one team has been embedded within the Punjab government to support programme execution and provide project management assistance. Other groups have worked directly with farmer collectives to understand pain points and improve access to crop residue management equipment.
Story continues below this ad
Venkatesh Kannaiah: Tell us about your India grants on air quality and their impact.
Santosh Harish: We funded a programme at IIT Kanpur to deploy around 1,400 low-cost sensors across Uttar Pradesh and Bihar, at least 10 times more than what existed earlier.
Focused on rural blocks, we wanted to demonstrate what sensor deployment at scale looks like and to understand the challenges involved.
There has been some scepticism from government agencies about whether large-scale sensor networks really work. We took the initial risk, tested the model, understood the challenges, and demonstrated how it could be done in practice.
Story continues below this ad
The second goal was to use sensors for remote monitoring of air pollution in rural areas. Low-cost sensors, along with satellite-derived estimates, are among the major innovations of the last decade that have matured enough to help us understand pollution levels. This project was testing whether that could work reliably at scale.
The third goal was to make rural air pollution visible. We have known for some time from models that air pollution, especially across the Indo-Gangetic Plain, is not dramatically different between cities, towns, and villages. It is almost a contiguous airshed, or a series of connected airsheds, where pollution from one area affects others. The pollution levels are often similar, though the sources may differ.
Also Read | Responsible AI cannot be an afterthought, needs to be part of DNA: Infosys CTO Rafee Tarafdar
I think this project has done quite well. Bihar has now had these sensors running for nearly two years, and Uttar Pradesh for a little over a year. Mumbai and a few other major cities have recently approached IIT Kanpur to help set up similar dense sensor networks in urban areas to complement regulatory monitors.
The second example is a grant provided to the Center for Study of Science, Technology and Policy (CSTEP) involving a different technology — reduced complexity models.
Story continues below this ad
To give some context, fine particulate matter (PM2.5) comes from multiple sources. Some particles are emitted directly, while others are formed in the atmosphere through chemical reactions involving gases such as sulphur oxides and nitrogen oxides from burning coal, petrol, and diesel.
Some pollution is produced locally, while some travels hundreds of kilometres. This makes atmospheric chemistry models extremely important because they help us understand how pollutants form and disperse over distance.
The challenge is that traditional atmospheric chemistry models are computationally very intensive. They take significant time to run scenarios and require specialised expertise, which is still limited relative to the scale of India’s problem.
Reduced complexity models — developed over the last decade in scientific literature — help fill that gap. They are faster, cheaper, and easier to deploy.
Story continues below this ad
CSTEP and its partners have worked on this significantly. IIT Bombay has taken the lead in developing one such model for India, while CSTEP has worked with Carnegie Mellon University on another.
The advantage of having multiple models is that they can be used together in an ensemble approach. Each model has different strengths and weaknesses, and comparing them helps make stronger policy recommendations.
Also Read | Climate tech is moving away from a green premium: Ankur Shrivastava, Founder, Momentum Capital
This helps ensure that policy and funding are directed toward the most important sources rather than solving the wrong problem.
Both models are currently in beta testing, and it is encouraging that the West Bengal Pollution Control Board has already engaged CSTEP to run these models and support action plan development and policy analysis.
Story continues below this ad
The third example is the work being done by the International Council on Clean Transportation (ICCT) on remote sensing of vehicular emissions.
The current system in India is the PUC regime, where emissions are checked at fuel stations by inserting a probe into the tailpipe. This mainly measures idling emissions and uses fairly imperfect proxies for pollution. It does not reflect how vehicles behave in real-world driving conditions, and in many cases, it measures the wrong indicators.
ICCT has been working on remote sensing methods, for example, installing equipment at toll gates that can measure emissions as vehicles pass through. It helps us understand which categories of vehicles are disproportionately responsible for emissions.
ICCT’s early findings were quite striking. In some cases, vehicles were found to emit nearly 10 times the prescribed or assumed limits under real-world conditions.
Based on those early results, the Supreme Court directed an official pilot to explore what a policy based on remote sensing could look like, and that work is currently underway.
Venkatesh Kannaiah: Tell us about your global grants on air quality and their impact.
Santosh Harish: The first is that we helped seed the EPIC Air Quality Fund. EPIC refers to the Energy Policy Institute at the University of Chicago.
We helped seed this fund to set up air quality monitors or sensor networks in countries with little or no public monitoring data. They made about 30 grants across nearly 20 countries.
In some of those places, the first monitors are already leading to conversations with local agencies about setting air quality standards for the first time, developing action plans, and identifying pollution sources that need attention.
Also Read | Love Machines are here to monetise the loneliness economy: James Muldoon, author and sociologist
The second example is a collaboration led by Stanford University and the International Centre for Diarrhoeal Disease Research, Bangladesh, focused on brick kilns.
The partners worked with brick kiln owners and workers to improve operations in what are called zigzag kilns. Across South Asia, there has already been a policy push for kilns to move from older technologies to this more efficient zigzag design.
They found substantial reductions in coal use and carbon monoxide emissions, with strong indications that particulate matter emissions also declined. This is likely to be a highly cost-effective intervention because nobody needs to redesign the kiln itself, it is mostly about improving day-to-day processes.
The hope now is to scale this across Bangladesh and eventually replicate it across Indian states as well.
Our role in these cases is primarily as a funder. The technical expertise — understanding kiln operations, improving methods, working with governments, and implementation know-how — comes from the grantees.
When we talk about air pollution, it is not always about breakthrough inventions or dramatic new technologies. In many cases, it is about implementing better methods at scale and applying well-established science rigorously. Sometimes the biggest opportunity is not a new invention, but using what already exists far more effectively.
Venkatesh Kannaiah: Tell us about the technologies you are betting on to solve air quality issues.
Santosh Harish: I think we can do a lot more in terms of cost-effective monitoring. Low-cost sensors and satellite-derived estimates play an important role. Some parts of the country are now fairly well monitored, and the credit for that goes to the National Clean Air Programme, which has been around since 2019.
For the longest time, even major cities like Bengaluru and Mumbai had only two or three monitors, which made it very difficult to understand the scale of the problem or identify where pollution was coming from.
Venkatesh Kannaiah: Your views on air capture technologies.
Santosh Harish: My view on air capture technologies like large-scale outdoor purification systems, smoke towers, and similar interventions is that it does not help much. The actual reduction in pollution has to happen at the source.
There have been attempts — large smoke towers, outdoor air purifiers on streets, and similar ideas. But from a cost-effectiveness perspective, these are generally poor solutions. At best, they may reduce pollution within a few metres or perhaps a few tens of metres.
Even if you argue that creating those small “clean bubbles” has some value, the costs are extremely hard to justify. It is often more cosmetic than meaningful.
Now, if we are talking about indoor environments — like hospitals, schools, or other places with vulnerable populations — then air purifiers are a very different story. Indoor air purifiers are a proven and well-tested technology.
Venkatesh Kannaiah: What has been the experience in developed countries on air pollution issues?
Santosh Harish: Many Western cities had extremely severe air pollution in the first half of the 20th century.
London had the famous Great Smog of 1952, which killed thousands of people in a few days. Many British and US cities, like Pittsburgh, had pollution levels that were more toxic than what we see today because they were heavily coal-based.
What is encouraging is that those countries made substantial progress in pollution control with far less scientific understanding and fewer tech tools than we have today.
China has made major progress over the last decade, with estimates suggesting roughly a 30% reduction in pollution levels between 2011 and 2020. Importantly, those gains came not from public air capture devices, but from reducing pollution at the source and stronger policy enforcement.
Venkatesh Kannaiah: How is the Indian innovation ecosystem around air quality?
Santosh Harish: There is obviously scope for innovation on the emissions control side. Better tailpipe technologies, cheaper ways of cleaning industrial stacks or chimneys, and various things that can make the electrification of vehicles cheaper and more widely used.
Take this idea of emission inventories. It’s basically a gridded database of how much pollution is coming from different sources. For instance, we don’t have an official public national emission inventory. There are multiple institutions capable of putting this together, and there have been efforts at developing such a national inventory. But we don’t have one that could be used to track progress over time.
Source apportionment is similar. You would want to be able to do this every few years because we have many rapidly developing cities.
While scientific talent exists, we need to ensure there are bridges between policymakers and public problems that need to be solved and what science can do for them. The other is about ensuring that incentives are in place for scientists to be part of this enterprise.
Venkatesh Kannaiah: As an air quality researcher, what are your asks from the Indian government?
Santosh Harish: The first would be to strengthen the science-policy bridge. That responsibility largely sits with the Central Pollution Control Board, which already has the mandate for it.
There are, of course, many government institutions working on air quality, and in some places these bridges work better than in others. There are also states like Gujarat and West Bengal where this kind of coordination has been relatively strong. But it is not consistent across the country, and more importantly, it is not routine or seamless.
The second would be to refocus the National Clean Air Programme squarely on PM2.5.
In the previous NCAP period, which ended in March, the emphasis shifted more toward PM10, largely for practical reasons. PM10 had a longer monitoring history and more available data across the 130-odd cities, so it was easier to track progress over time.
But the consequence was that a disproportionate share of NCAP resources went into dust management. From a health perspective, PM10 is important, but it is the finer particles — PM2.5 — that are far more harmful and responsible for much of the health burden.
So while the shift to PM10 was understandable from a monitoring standpoint, it had unintended consequences. Going forward, we need to pivot back and keep PM2.5 at the centre of policy design. PM10 is, in some sense, yesterday’s pollutant. The world has moved on, and we should too.
The third would be to move from city-level planning to at least state-level planning, or more accurately, to operationalise the concept of airsheds.
For years, we have treated air pollution primarily as an urban problem. But even if your goal is to improve air quality in cities, you cannot solve it by looking only within city limits.
