sam i am

Save the planet, reduce inequality.


I submitted a verbal comment in person on Tuesday at the public hearing on the proposed SAFE Vehicles Rule in Dearborn (MI). This rule would basically freeze fuel economy standards at current levels for a to-be-determined number of years. The standard for new passenger vehicles is currently about 37 miles per gallon ; the current policy, developed during the Obama Administration, would push that number to 54.5 by 2025.

It was good to see how the public comment process works and what people had to say. It felt like some 80-85% of comments were against the proposed rule. There were definitely some industry representatives pledging support for the rule, but there were also many who criticized it. I think that in general, those involved in the making of cars did not support the rule, and the most common stated rationale was regulatory certainty and consistency. The fixed and transition costs of adjusting car production to satisfy a new set of standards are enormous. A long, steady fuel economy path is laid out by existing rules; suddenly (and in all likelihood temporarily) blocking that path would impose huge costs on automaking. Just as importantly, California's intention to keep its fuel economy standards high threatens to fracture the U.S. market. Having to make cars to one specification in California, another one in the rest of the U.S., and likely others elsewhere in the world -- that does not sound good for business.

There were seemingly more than a hundred speakers in all, and they touched on health impacts, affordability, vehicle and travel safety, and jobs, among other things. Anyways, here is my comment below -- I had three minutes to talk.



Thank you for the opportunity to comment here today. My name is Samuel Stolper; I am a professor of environmental and energy economics and policy at the University of Michigan, Ann Arbor and I’m here as a private citizen. Policies like the SAFE Vehicles Rule are what I study for a living.

I would be remiss if I didn’t begin my comment by highlighting the moral responsibility of the United States to reduce greenhouse gas emissions. In 2011, per capita emissions in the U.S. were 20 tons; in the rest of the world they were 7. The U.S. is responsible for 27% of greenhouse gas emissions in the last 160 years, more than any other country, and more than 28 European Union countries combined. The moral imperative of climate change mitigation continues to be ignored by our representatives in government.

That said, climate change mitigation policy will always have effects beyond the main target of emissions levels. For that reason, regulatory impact analysis is a vital – though not on its own sufficient – step to inform and justify policy decisions. The environmental benefits of higher fuel economy should absolutely be considered together with impacts on driving, congestion, and safety. The problem is that the SAFE Vehicles Rule is currently being justified by an incomplete, unscientific, and non-transparent consideration of costs and benefits. Others here today have already exposed this fact and highlighted a litany of important flaws. Among these, I’d like to quickly reiterate two: first, the estimate of reduced fatalities from accidents under the SAFE Vehicles Rule is inflated and indefensible; and second, regulatory certainty is integral to cost-effective policy and yet widely ignored in the proposal and analyses of fuel economy freezes.

I’d like to elaborate on a third flaw: the woefully incomplete and short-sighted consideration of environmental benefits, from reductions of both of local and greenhouse gas pollution. I’ll focus on the latter. The social cost of carbon – which the government uses to place a monetary value on greenhouse gas emissions – has been dropped from a 2020 central estimate of 45 $/ton, to between 1 and 6 $/ton. The assumptions that underlay this drop – higher discount rates, and omission of damages outside the U.S. – are inconsistent with scientific modeling practices, the recommendations of the Council of Economic Advisers, and the logic of the climate problem. Moreover, even our most complete estimates of the social cost of carbon fail to account for a large portion of benefits. Labor productivity, crime, conflict, the impacts of sea level rise, natural disasters, ocean acidification – none of these are considered. Just because you can’t easily monetize it, doesn’t mean you don’t count it.

If the government is going to justify policy decisions on the grounds of net benefits, it must give a complete and scientific consideration of all costs and benefits. Such consideration would reveal the “no-action” alternative to be superior to any freezes on fuel economy increases. I strongly advise DOT and EPA to maintain existing fuel economy standards. Thank you.


Ayoooooo my time-to-peer-reviewed-publication just dropped from infinity to 7 years. The Journal of Development Economics is publishing "Can environmental policy reduce infant mortality? Evidence from the Ganga Pollution Cases." Shout out to my coauthors Shareen Joshi and Quy-Toan Do.

We're talkin' river pollution in India – which has been rampant for decades, all over the country. There’s a sense that regulation has not worked, and it's backed up by research. Shareen, Toan, and I basically did a quantitative case study of one specific environmental regulation in India. We are asking, did this regulation work? What, if any, were its impacts? Just as importantly, what were the mechanisms of its impacts? Finally, what can we learn more generally about successful policy design and the relationship between water pollution and health?

The area in focus is Kanpur – a city of approximately 3 million inhabitants, sitting right on the Ganga River (The Ganga is India’s longest river, revered as a goddess by Hindus; industrial operations along the river number in the hundreds of thousands, and more than 330 million people reside in the Ganga River Basin). Kanpur’s got a large leather tanning industry that has historically dumped its voluminous waste directly into the Ganga. In 1987, the lawyer M.C. Mehta sued Kanpur industry and govt for violation of the Constitutional “right to life”. In his own testimony, Mehta describes the straw that broke his camel's back: in 1984, a matchstick tossed into the Ganga in the pilgrimage town of Haridwar resulted in the river catching fire for 30 hours, due to a toxic layer of chemicals produced by a pharmaceutical firm.

This was India’s first environmental public interest litigation. The Supreme Court decided to hear the case, splitting it into two parts to separate the suit against industry and the suit against government. Together, these are now known as the "Ganga Pollution Cases". In an unprecedented October 1987 verdict, the Court ruled in favor of the plaintiff both times. In particular, tanneries would have to clean their wastewater or shut down.

To investigate the rulings’ impacts, we merged river pollution monitoring data and birth histories from a national survey. Did the rulings do anything? We compare Kanpur and its downstream neighbors to other districts in the Ganga Basin, before and after the rulings. Here's a map of the Ganga Basin with "treated" districts highlighted. You can also see the location of government river pollution monitors.

Pollution drops significantly in the aftermath of the verdicts. The graph shows biochemical oxygen demand (BOD>3 means river water fails to meet India's standard for bathing). Calcium, sulfate, and chloride ions – all associated with tanning – drop too. Chromium is anecdotally the worst tanning pollutant, but we don’t have data on it.

Infant mortality drops significantly, too. This is driven entirely by changes in neonatal (one-month) mortality risk, shown below. You can see that mortality risk is decreasing through the 70s and 80s, but the trend is roughly the same in both the Kanpur area and the rest of the Ganga Basin. The gap between these two groups, however, shrinks after the verdicts.

So, the Supreme Court rulings appear to have caused a reduction in both pollution and infant mortality. Was the latter caused by the former? I.e., did the policy affect infant survival entirely through reducing pollution, or are other channels (e.g., avoidance, or household water treatment) important? This matters for designing well-targeted environmental policy.

To answer this question, we first measure how much BOD harms infant survival in general. The challenge, of course, is that pollution is correlated with other stuff (urbanization; economic activity) that affects health. In our case, in-district BOD doesn’t predict mortality. But upstream BOD does! That’s our instrumental variable. Take a look below: the predicted effect of high BOD (column 1) on in-district mortality is negative and statistically insignificant. But the predicted effect of high upstream BOD (columns 2-5) is positive and quite significant.

According to our estimates, surpassing India’s bathing-class standard of BOD<3 raises the risk of neonatal mortality by 10-15 percentage points (Panel B in the table below):

Are the magnitudes of the policy impacts on pollution and health consistent with the story that health improved entirely because of the pollution reduction? Yes – statistically, we can’t rule that story out.

To summarize:

  1. Here we have a rare instance of water quality regulation that actually improved outcomes in India. It's interesting to think about *why*, though we can’t credibly answer that question here. Also, note that water pollution is still an enormous problem in Kanpur (and elsewhere in India).
  2. With the Ganga Pollution Cases, health improved primarily through pollution reduction. But in general, policy can affect health through awareness, behavior, and incomes as well. We need to better understand *mechanisms* of policy impacts.
  3. The analysis strongly suggests that industrial pollution kills infants. While this is surprising to some, it could be explained by in utero exposure or transmission of river pollution through much more than just drinking water (cooking, bathing, irrigation, or just person-to-person contact).
  4. We've found hard evidence that the health burden of pollution is not confined to the origin of that pollution. River pollution kills infants where it is produced and also in downstream communities. That makes it all the more important that policy span jurisdictions; it won't do for individual localities to only think of their own well-being here.


Check it -- I wrote a letter to the Detroit News. It was a response to an opinion piece published two weeks earlier. That first piece had the title "Coal is still a necessary resource." Don't really agree with that statement. The newspaper told me they'd run my letter and then edited it without asking for my permission, so here is my full submission below. Shout out to Max Auffhammer and Meredith Fowlie, whose post on the Energy Institute Blog provided a lot of inspiration.


Gary Wolfram wrote an opinion piece two weeks ago in this newspaper, arguing that “coal is still a necessary resource.” He based his argument on two primary assertions: first, that coal is the best way to ensure reliable, consistent electric supply; and second, that regulation of coal power generation is unjustified. Neither of these assertions are correct. Both are based on partial logic; both are unsupported by evidence. An unbiased, complete accounting of costs and benefits reveals the truth about coal: it should be rapidly phased out.

Americans have come to expect extremely reliable electricity supply, and for good reason: it is unpleasant to suddenly have the lights turned out, and unpredictable power outages severely restrict economic productivity. But regional system operators already compensate power plants for reliability, explicitly paying them to be available to generate electricity in the event of emergency. And coal is not the secret to further reliability improvements: only 0.0007% of major U.S. power disruptions in the past five years were due to fuel supply shortages. Most disruptions – including widespread outages in Puerto Rico, which Wolfram tries to use as evidence of the need for coal – were due to fallen power lines caused by extreme weather.

Coal is no more helpful in the effort to supply electricity to Americans at lowest cost. Those who argue otherwise, as Wolfram does, are defining the word “cost” incorrectly. The economic cost of power generation is not simply the price of fuel inputs, equipment, and labor; it’s also the damage done to human and environmental health. This is not a point of contention among economists. The delivered price of coal in November 2017 in the United States was $2.03 per MMBtu. Meanwhile, the non-partisan National Academy of Sciences has estimated that the health and environmental damage of coal consumption is $6.60 per MMBtu – 325% larger than the production cost! During his State of the Union speech last week, President Trump proclaimed his love for “beautiful, clean coal”, but that description could not be more false.

More than being “unnecessary”, coal is actually making society worse off on the whole. But let’s not forget that this resource provides livelihoods to thousands in the United States, and fears that its phase-out could devastate some communities are justified. A transition to cleaner energy that does not make these communities whole would be a failure.


Two weeks ago, I attended the Transportation, Energy, Economics, and Environment (TE3) Conference at University of Michigan. Great words to put together, subject-wise. Also a valiant effort to bring together academia with government and private enterprise. There were some very smart people in attendance, and I learned some interesting facts. For instance, I learned that: (a) the conventional wisdom about rising fuel economy in the U.S. auto fleet is that it leads to lighter cars and more deaths from car accidents); (b) if you look at changes in the whole distribution of vehicle sizes and weights, the results suggest that the opposite is true -- that is, accidents will largely involve lighter cars on both sides, and deaths will go down; (c) the 'rebound effect', in which more fuel-efficient cars get driven more miles, flips the prediction back to (a) -- in a back-of-the-envelope calculation, the rise in accidents from increased driving swamps the drop in accidents from lighter cars. This was an emotional roller coaster for someone who thinks reducing greenhouse gas emissions is important.

Anyhow, that sequence of findings, among others, did not answer what, for me, is the most pressing question in climate economics/policy. It's the same question that I suggested recently when someone asked me what to ask Bill McKibben during his visit to U of M last month. Here's the question:

  • Is electricity storage the holy grail solution to climate change?

What I mean is, if we need to totally decarbonize the economy in order to stabilize the Earth's climate; and if the sun and wind provide enough zero-carbon electricity to power the economy (including transportation); and if the only obstacle is being able to economically store that electricity; then does anything else besides solving electricity storage even matter?

I haven't really put anybody on the spot with this question, but also nobody has really given me an explicit answer to this question. Economics research is overwhelmingly focused on retrospective analysis, because that's where the empirical data are (my own research perfectly exemplifies this skew). We don't have electricity storage in any economically meaningful amount yet, so I personally haven't seen economics research on the subject (there is certainly research on the subject in other fields though; just type 'electricity storage' into Google Scholar). At this point, I just wonder whether studying Corporate Average Fuel Economy (CAFE) standards is still fruitful, when electric vehicles are seen as the future and electricity storage may be the only way to actually keep the planet from overheating.