The tragic explosion last week at the retail fertilizer plant in West, Texas, illustrates some of the pitfalls of hazard reduction policies. Public and media discussion of the case also shows how the standard economic model of human decision-making — used over the past 200 years — is deficient.
The upshot is that, as we often do, we are likely to waste resources on ill-thought-out regulation of such plants while ignoring greater dangers that could be reduced at much lower cost.
Start with economic theory, which long was grounded on the assumption that humans were highly rational in making decisions. This extended to assessing risks of adverse outcomes.
In the late 20th century, that assumption of rationality was taken to extremes, both in theorizing about the effects of macro-economic policies and, at a micro level, about the behavior of financial markets.
However, more recent work in experimental and behavioral economics shows how we often are not rational and how this up-ends some predictions of conventional theory.
Amos Tversky and Daniel Kahneman, two Israeli psychologists, did much path-breaking work in this area, with Kahneman winning the 2002 economics Nobel for their work. (Tversky had already died.)
They coined the term “availability heuristic,” translated as: “if you think a lot about it, it must be important.” Right now, we are thinking about fertilizer plant explosions. Hence we think it must be an important issue.
Moreover, we tend to underestimate risks that are frequent and familiar — driving while talking on cell phones — and overestimate those that are rare and unfamiliar but in the news, such as a terrorist attack or a fertilizer explosion.
Fatal explosions at a retail fertilizer blending plant are very rare. Indeed, the one in Texas is the first one in our country in the 90 years that ammonium nitrate has been used as fertilizer. Yes, there have been explosions at manufacturing plants. And yes, millions of tons of the chemical in the Texas explosion, when mixed with a precise proportion of fuel oil, are used annually as a commercial explosive in mining and construction. And yes, terrorists have used it.
I am familiar with both ammonium nitrate and anhydrous ammonia, the other contentious material in the Texas tragedy, from years of farming. So I admittedly may fall into the camp that underestimates risks because of excessive familiarity. But media stories and comments posted in response to online ones show that reporters and the public are at the other pole.
Much of this stems from the difficulty that people have with putting apparently large numbers in context. This applies equally well to news of large layoffs, oil spills, deforestation and budget deficits.
The plant reportedly had 54,000 pounds of anhydrous on hand. This seems like a lot, and the media describe it using terms like “enormous tanks.” But it is about the smallest tank a plant can have without needing to be completely empty before accepting a single-semi truck resupply. Already 40 years ago, the farm coop where I used to buy anhydrous would empty and refill such a tank five times in one busy day. So what seems enormous to national media is minuscule to farmers and farm suppliers.
Moreover, the safety measures in West were not all that different from those at thousands of similar facilities across the nation. I have never seen a retail plant with blast walls nor been in one with a sprinkler system. This was not a case of particularly lax regulation by an ideologically anti-regulatory state government.
Yet at least 14 people are dead. Isn’t that sufficient proof that big changes are needed? Don’t we need more inspections, mandatory fire suppression systems, blast walls and separations from inhabited areas as specified by the American Standard Table of Distances for the Storage of Explosives?
The answer is no, probably not. Consider that about 90,000 Americans die in nontraffic accidents each year. We really have never had federal or state efforts to look at the different hazards involved and the cost of reducing them.
But if we did, it is clear that devoting resources to reduce hazards around retail fertilizer plants would be low in terms of lives saved or injuries averted per million dollars spent. Devoting the money to hazard reduction with higher payoffs would make us better off. But there is little public support for prioritizing.
We faced a similar situation in 2008, when a collision near Chatsworth, Calif., between a commuter train (the driver was texting) and a freight train killed 28 and injured more than 100.
Congress hastily passed a pending railroad regulation bill, and George W. Bush signed it 24 days after the accident. Among other things it required a GPS-based system of “Positive Train Control” to avoid collisions.
Mandated to be in place by 2015, this already has cost railroads an estimated $8 billion to $12 billion. Indications are that it will do little to reduce the death toll from train collisions, currently about 50 per decade.
Clearly, if one wants to reduce transportation-related deaths or even railroad related deaths, Positive Train Control is far from the best way to do it. But large majorities in the House and Senate voted for it because of the hubbub after the Chatsworth accident.
This railroad legislation made us worse off rather than better off. We should not repeat the mistake in the wake of the West explosion.