EPA’s New Ozone Rule: Part 22

The goal of our discussion is a cost-benefit analysis. What benefits would lower ozone bring us, how much would it cost, and do the benefits justify the costs? These questions are addressed in two EPA documents:

• Final Ozone NAAQS Regulatory Impact Analysis (March 2008). To view, click here.
• Regulatory Impact Analysis Final National Ambient Air Quality Standard for Ozone (July 2011), which is a supplement to the March 2008 document. To view, click here.

As these documents are at the heart of our discussion, I really should take the time to read and understand them thoroughly. But my time being short and the documents together totalling 645 pages, unfortunately I can’t do them justice. But you can read them, and I can point to certain highlights that can give us food for thought.

These papers can be challenged. But critics who would argue with their conclusions can’t just glibly dismiss their claims out of hand. They need to demonstrate that either their assumptions or their methods are wrong. They need to argue the issue with the same level of detail that these documents do.

What attracted my attention most were a few charts in the beginning of the July 2011 document. The first chart, Table S1.1 on page 6 of the document, lists the costs and benefits of ozone and PM_2.5 (particles suspended in the air 2.5 microns in diameter and larger) reduction. Please open up the chart by clicking here.

Let’s describe the elements of the chart. There are three main rows, each row showing the costs and benefits of each of three possible limits on ground-level ozone: 0.075 ppm, 0.070 ppm, and 0.065 ppm. Each row is divided in half: the upper half for multi-city analyses, the lower half for meta-analyses, where the authors did not collect raw data but rather gathered data from other studies. Each half-row sites statistics from three studies: six studies in all. The studies, listed in order of appearance in the chart by author’s name are:

• Bell, M.L. et al, 2004, Ozone and short term mortality in 95 US urban communities, Journal of the American Medical Association 292(19) 2372-2378. For the article, click here.
• Schwartz, J., 2005, How sensitive is the association between ozone and daily deaths to control for temperature?American Journal of Respiratory and Critical Care Medicine, Vol. 171(6):627-631. For the article, click here.
• Huang, Y., F. Dominici, M.L. Bell, 2005, Bayesian Hierarchical Distributed Lag Models for Summer Ozone Exposure and Cardio-Respiratory Mortality, Environmetrics, 16, 547-562. For the article, click here.
• Bell, M.L., F. Dominici, J.M. Samet, 2005, A meta-analysis of time series studies of ozone and mortality with comparison to the national morbidity, mortality, and air pollution studies, Epidemiology, 16(4):436-445. For the abstract, click here.
• Ito, K., S.F. DeLeon, M. Lippmann, 2005, Associations between ozone and daily mortality: analysis and meta-analysis, Epidemiology 16(4):446-457. For the article, click here.
• Levy, J.L., S.M. Chemerynski, J.A. Sarnat, 2005, Ozone exposure and mortality: analysis and meta-analysis, Epidemiology 16(4):458-468. For the abstract, click here.

There are three major columns in the chart: total benefits, total costs, and net benefits (total benefits minus total costs). Total benefits and net benefits are divided into two half-columns: 3% discount rate and 7% discount rate. I don’t really understand what these are, but I can guess from what I’ve read. As I understand it, social discount rates are the rates of return one could expect if money spent on a social good was invested in financial markets instead. Let’s say you invested a large amount of money in 200 mutual funds chosen at random. Some funds would get a high rate of return, some a low rate of return, but over 10 years time, the rate of return would likely average out to some figure no matter what funds you chose. This rate of return is what we call the social discount rate.

Now the author prepared the chart showing amounts in 2006 dollars that would accrue in 2020. That suggests to me that the author is asking: if we go to a lower ozone standard in 2006, what are the costs and benefits we can expect in 2020? We can expect adopting a stricter ozone standard to cost us so much in 2007. If instead of adopting the stricter standard, we immediately invested that money instead at a 3% or a 7% rate of return, how much money would we get in 2020? We do the same for costs in 2008 and 2009 and so on. We would also see benefit in 2007. We can estimate the financial value of that benefit (harder to do than determining costs) and ask the same question: if we immediately invested that money at a 3% or 7% rate of return, how much money would we get in 2020? We do the same for benefits in 2008 and 2009 and so on. We sum up the financial returns from costs and benefits, and compare the results.

Now if you look at the numbers, you’ll see that for each combination of ozone limit, type of study (multi-city vs. meta-analysis) and cost/benefit column (for example, costs estimated for an ozone limit of 0.075 ppm, multi-city analyses) that the numbers in the combination are quite close to each other; the differences between the studies are not great. I took the average of each combination and put them into a condensed chart. I also calculated the size and midpoint of each net benefit range. Figures are in billions of 2006 dollars. A negative net benefit is a net cost.

What I found interesting about these numbers is that total costs are the same for each limit of ozone both for the multi-city studies and the meta-analyses. However, for total benefits and net benefits, the meta-analyses are consistently higher than the multi-city studies.

Also interesting is that the range of estimation of net benefits widens as the ozone limit gets lower. The range is $8.6 billion for 0.075 ppm, about $20 billion for 0.070 ppm, and about $33 billion for 0.060 ppm. That tells me that as the ozone limit gets lower, there is more uncertainty in estimating costs and benefits.

Now if you look at the midpoints of the ranges, the midpoints for the meta-analyses are fairly consistent: about $4 – $5 billion. But the midpoints of the ranges for the multi-city analyses go down as the ozone limit gets lower: from a net benefit of $2.4 billion for 0.075 ppm to a net cost of $1.8 billion for 0.070 ppm and then finally to a net cost of $4.6 billion for 0.065 ppm. But even the meta-analyses predict high net costs at the lower end of their ranges: up to $6 billion for 0.070 ppm and up to $11.7 for 0.065 ppm.

This tells me that as we choose lower limits for ozone, the uncertainty of estimating what the net benefit will be increases as well as the risk that the net benefit will be negative (i.e. really be a net cost). Of course, this evaluation depends on how much financial value we attach to a human life.

But it is also important to consider the benefits alone. If the benefits were purely financial, then it would make sense to be very utilitarian and forget about those benefits if they were outweighed by costs. But if those benefits are in a substantial number of lives saved and illnesses alleviated, then they become much more desirable, even urgent. Even if the economics dictate that it is wiser not to pursue those benefits now, they can remain in our sights as a goal we want to achieve eventually.

Following the table we just discussed is Table S1.2: Summary of Total Number of Ozone and PM2.5‐Related Premature Mortalities and Premature Morbidity Avoided: 2020 National Benefits, page 8 of the document. Please open the chart now by clicking here.

According to this chart, the number of lives that can be saved by both reducing ozone and particulate matter 2.5 microns and larger is substantial. To put it in perspecitve, on 9/11 2,753 New Yorkers were killed. Surely, if we were aware of a plot by Al Qaeda to kill 4000 Americans, we would expect our government to react. If we can save that many lives by protecting them from air pollution, shouldn’t we try?

There is one more topic we need to discuss on this subject, and that is compliance.