Show Me the Data

Sources: Centers for Disease Control, EPA Toxic Release Inventory. Red dot = Virginia.

I’ve been pondering Vivian E. Thomson’s book, “Climate of Capitulation: An Insider’s Account of State Power in a Coal Nation.” Thomson, a University of Virginia professor who served on the State Air Pollution Control Board during the Warner and Kaine administrations, stresses the high cost of air pollution in human health and mortality. Her treatise quotes various studies predicting that tighter air quality standards would yield major reductions in premature deaths.

I wondered, has anyone gone back to see if the expected reductions actually materialized? If society is going to spend billions of dollars to make gains in air quality, it would be nice to know that there actually is a payoff in the form of better health. My sense is that no one ever checks. If anyone does, the public never hears about it.

Out of curiosity, I ran a correlation analysis between two types of data: the pounds of toxic emissions reported to the Environmental Protection Agency and the 2016 incidence of cancer reported by the Centers for Disease Control, broken down by state. All other things being equal, one would predict that larger toxic emissions would be associated with a higher incidence of cancer.

As can be seen in the chart above, there is almost zero correlation — the R² is .0028 — between toxic emissions and the incidence of cancer in a state. (The chart omits Alaska and Nevada, huge outliers in terms of volume, which would have made the trend line to an even more negligible .0008.)

I readily concede that this is a superficial analysis. Among other factors one might consider would be the size of the state in square miles, on the theory that the same volume of toxic chemicals spread over more acreage would dilute human exposure and result in lower cancer rates. Also, any sophisticated comparison would account for differing toxic release profiles of the 50 states. The Toxic Release Inventory tracks some 143 chemicals, from acetaldehyde to zinc, some of which are more toxic and/or carcinogenic than others.

I publish the graph above not to dispute the idea that there is a connection between toxic chemicals and human health — of course there is — but to push back against the idea that spending billions of dollars tightening regulations on toxic chemicals ineluctably leads to better health outcomes. Perhaps the health benefits are everything Vivian Thomson purports them to be. But perhaps pollution abatement is subject to the laws of diminishing returns which means smaller benefits for larger expenditures. Conceivably, Virginians would see greater benefits to their health if they spent the money in other ways.

When Thomson criticizes Virginia’s “climate of capitulation” — the idea that industry exercises a controlling influence over the political and regulatory system — one might expect her to demonstrate that Virginia’s health is worse off as a result. We would expect to see (a) that states with “traditionalistic” political cultures like Virginia impose laxer pollution restrictions than “individualistic” and “moralistic” political cultures, and (b) that those laxer restrictions are reflected by worse health statistics. But in her book, she makes no effort to demonstrate such connections.

Indeed, Thomson misses what would seem to be an easy opportunity to do so. She quotes a Harvard study as stating that installing “Best Available Control Technology” at the Mirant coal-burning plant in Alexandria would avoid 40 deaths, 43 hospital admissions, 560 emergency room visits and 3,000 asthma attacks per year. Well, from her point of view, things worked out even better than adopting best-in-class technology — the Mirant plant shut down altogether, emitting zero pollution. Did asthma rates in Alexandria decline as advertised? According to a 2015 health profile, 21.4% of Alexandria 10th graders in 2014 had been diagnosed with asthma. I can’t find older data on the Web, but surely it exists. Was the asthma incidence lower than 10 years previously? If not, what does that tell us?

Thomson has total faith in the validity of the studies she cites, which, of course, align with her ideological proclivities. But if she wants people to accept her argument that Virginia’s regulatory policy is bad for Virginians, some of us would like to see the evidence.

There are currently no comments highlighted.

9 responses to “Show Me the Data

  1. It appears to the agnostic observer that a core element of the liberal catechism is that if corporations and business have influence, the results will be bad for society but if government and activists have the upper hand, we will all be better off. That conviction appears to be the first filter for any data.

    So, if it doesn’t come from an activist group or if there are any corporate sources of funding the data is pre-determined to be biased and worthless.

    How Liberals square this conviction with what they can see around them every day of the results of government control and even involvement is a pean to the power of faith and ideology.

  2. Several observations about your analysis:

    I think it was a good start. I’ll assume your cancer incidence is new cases per 100,000 people per year. Let’s assume that emitted toxins have a low correlation with cancer. What might have a higher correlation? Your graph seems to show a range of 125 (low) to 200 (high). That seems like a wide variance for states within the same country.

    1. Health impacts of air pollution go beyond cancer. You are only measuring one class of malady.
    2. Air pollution may not impact all types of cancer equally. If you eliminate the kinds of cancer that are not more likely in places with high air pollution you might get a stronger correlation.
    3. As you mention, you might be better advised to use “per sq mi” than per capita when analyzing toxic waste emissions. You might even have to analyze by metropolitan area. Toxins tend to come out of the air as time and distance from the emission source increase. The Mirant plant was smack dab in the middle of one of Virginia’s most densely populated places. There was virtually no buffer between the plant and the nearby residents. What the plat spewed the residents breathed.
    4. Not all toxins are created equal. If you are analyzing the health effects of air pollution caused by electrical generating facilities son’t you have to focus on the toxins produced by such plants rather than all toxins?

    As for the Mirant plant – let me provide some first hand experience. The soot from that damn plant would settle on cars leaving a visible film over the car. When I used to fish from a boat near the hot water exhaust from the plant (into the Potomac River) I’d feel like I had a sore throat for some time after I left the river. Alexandria has a population density of over 10,000 people per sq mi. Having an antiquated power plant spewing whatever in an area that densely populated was inexcusable.

  3. I would think from a scientific point of view it would be hard to test your thesis. We are surrounded by so many harmful substances that it would be difficult to prove that the reduction or absence from one source could result in specific health outcomes. Statistically it is easier to prove harm when something exists than to show an improvement when it does not exist.

    Much of what we are exposed to comes from what we eat. Most processed foods contain corn or soy (80% of what’s in a grocery store). Setting aside the issue of the health effects of GMOs (no large scale studies have been done, the studies that have been done show they are harmful) the chemicals applied during the growth of the plants are endocrine disruptors and cause a whole host of illnesses including cancer.

    It would also be hard to show societal improvements from the reduction in just one point source. Airborne pollutants are carried long distances as was shown in the days when we were dealing with Nox and Sox. Given that the general trend is that we are getting sicker, and the younger generation is not likely to live as long as their parents, it would be hard to show a small improvement against a background of general decline.

    Your comment about diminishing returns is a good one, although some compounds are toxic in extremely low concentrations. Fortunately, in the case of energy production, in many applications the cleanest solution is also the cheapest.

    This is where we need to collaborate. Rather than get into arguments about small changes at the margins, we can design new approaches that provide clean energy, at the lowest cost, that put the most people to work of any alternative. Folks all along the political spectrum should be able to get behind those types of solutions.

  4. How dare Ms. Thompson publish a book critical of your generous patron! Good thing they have a knight in shining armor to defend them. But if your rather pedestrian exercise in science is the answer they intend, well, they need to do better than that. Even as a non-scientist I could spot that effort to as a significant stretch, a high school statistics exercise.

    She was known as a hard line, environmental purist and I suspect that was known when she was appointed. Elections have consequences and governors get to pick those boards. She sparked a lot of hand wringing in the business community. Kept some lobbyists busy. But in the long run the process worked and she rotated off. Mirant was no longer in a viable location, and if Dominion was going to build what better be the last coal plant in the United States, they should have expected it would be under super-stringent emissions controls. I have zero knowledge of the other issue you cited. I do not lie awake worrying that environmental toxins or GMO corn will kill me, but in general I’m a clean air stickler myself.

  5. there are many pitfalls for amateur data “analysts”.. this is sort why there is an actual 4-yr degree in the field.

    we not only have the experience of the US citizens but citizens in other countries.. and Europe – and Japan are far tougher than we are… in general. “regulations” don’t “cost” money – what they do if force those who would “externalize” costs to other people – not just as cancer but in shortened lives and lives themselves for children, elderly and compromised immune systems.. these folks become the fodder of less regulation..

    The irony here is that if you talk to the anti-regulation people about what countries are better.. for instance 3rd world countries with far fewer regulations – the complaint is that those countries don’t have rule of law or protect property rights.. as if the “property” rights of humans is a different thing than the “property rights” of someone trying to made a buck by dumping waste in a river or airshed..

    Not about “liberals” .. it’s about unethical and corrupt philosophies when it comes to human beings.. and business.

    I’m opposed to regulation for regulation sake.. and we do have some of it but a lot of this is about someone who wants to produce a product that has hazardous waste as a by-product and they basically want to get rid of it on the cheap.

    We have superfund sites across the country including about 2 dozen in Va that are there because getting rid of “excess” regulation.

  6. I distrust this report although I haven’t read Prof. Thomas’s book. Throwing out CDC data in a homemade analysis is foolish. One big question is how long is the time that cancer actually develops in a person’s body and the period for which he or she was exposed to carcinogens?

    • As I said in the post, I’m well aware that my quickie analysis was very superficial. I was trying to make a larger point — it would be desirable if someone actually took a retrospective look to see if the predicted benefits to human health actually materialized. How could any reasonable person oppose that idea?

      • how would you oppose it ? Because the starting premise is wrong… but more than that do you not think the industry does not do it’s own studies to bring to the table?

        The problem is folks who don’t understand data .. misinterpreting it then asking why no one is looking into it.. and yes.. I’d oppose that as not understanding data – trying to disaggregate it .. and ignoring the fact that industry and industry groups are well into this process and have been guilty in the past of cooking the data to benefit them.

        This started with cigarettes when the CDC and other govt scientists first did studies on cigarettes and the industry impugned the science and the scientists.. to claim that cigarettes were not bad.. and anyone who has bothered to follow that history – KNOWS that this went on for years.. decades.. before the Cigarette industry got caught.. and the proof is in the tobacco settlements.

        And this kind of thing continues… que VW and other producers of diesel engines who got caught – red-handed.. cooking the data..

        who should I trust? in determining the harm that chemicals cost?

        Well if I had half a brain, I’d look at the long and factual history of industry playing with data… before I believe we need to keep the govt more honest..

Leave a Reply