Today the Environmental Protection Agency announced new sulfur emission limits, which they virtually guarantee will add five years to Ms. Agnes Riolato’s life. Don’t believe me? Well okay, maybe that’s stretching things a bit. But they did say this new regulatory requirement would prevent as many as 2,000 respiratory deaths per year without costing refiners and automakers any significant increase in costs.
Of course, as you might expect, the American Petroleum Institute disagreed. Their analysis – no less biased than the EPA’s I am sure – claims the new regulation would raise gas prices while providing little environmental benefit – and could in fact cause interrupted supply lines. House Energy and Commerce Committee Chairman Fred Upton (R-Mich.) expressed concern as well, noting that, "the American people have endured an unprecedented three straight years of gasoline above $3 per gallon. This winter’s cold snap underscores just how vulnerable American families and businesses are to any increases in energy costs, and yet the administration is moving forward to raise prices at the pump … this rule is all pain and no gain."
While that may be true, what about poor Agnes? Is her life worth nothing? Now, I should note that Agnes is a fictitious character – at least as far as my Google search indicated. I am pretty confident out of 7.2 billion people in the world there is at least one person with that actual name, and in the 1 in only slightly less than that amount chance she were to read this blog post, I apologize in advance for any perceived derision.
But what I am driving at here is that I’m not at all sure the EPA’s claim of reducing respiratory deaths by 2,000 a year isn’t just as fictitious as Agnes. I don’t know that it’s fictitious. I have no reason to doubt – but that’s the point. Now stop and think for a minute – no, I mean really stop and think (if you’ve read this far, you’ve already committed more of your life than you had planned, so what’s a few more seconds).
How would you go about projecting a reduction in respiratory deaths (however hundreds of ways that might be defined) attributable to regulatory standards at such a macro level that it would be nearly impossible to predict the many variations in implementation? Just imagine trying to construct a model that would account for those possibilities. Head hurt?
I’m sure there are folks much smarter than me out there who can do that sort of thing. How they convince someone else that doing so produces value sufficient to earn a living is another matter. But even if you can model it, you are still left with making all kinds of assumptions on how things will actually play out. To me, the complexity involved here makes me really question the validity of the assertion: 2,000 lives a year saved. But there you have it.
Let’s face it, we’re living in a sound bite media reality where we toss around portended facts & stat’s as if they were as ingrained in our communicative lexicon as contractions. It’s hard to tell what’s worse: that sound bite reporting does little to build knowledge, awareness and understanding – or that it has been so overused that its effect is now largely ignored (i.e., there is no informing and educating whatsoever going on ).
So I think what we need is a new governmental agency. This is where my conservative colleagues frown and get angry that I was leading them on. But really, why not – who is going to notice one more agency at this point? I want a nonpartisan group of very smart thinking types with strong analytical and mathematical skills (something similar to the CBO) whose responsibility it is to score every official promulgation that comes from another federal agency.
I want them to build the type of model that I described above and understand it in a way that I most certainly could not. And I want them to assess the variability around all of the key assumptions. So, for example, if one of the assumptions used to project the 2,000 lives saved was to assume Agnes would spend 600 hours a month living at her home in the suburbs and 130.5 hours a month in the city (this of course is assuming that’s the only two places she will travel in a month – guess we’d have to assess that assumption as well) I want to know what the probability is of the assertion being made coming true given the variability around those assumptions.
And then I want them to use whatever mathematical techniques are necessary to produce a scoring of the reported assertion. They can play with the reporting language, but I am envisioning something like this using the real life example shared above:
Last year at this time it was reported by the Environmental Protection Agency (I’m just trying to be pragmatic here) that annual respiratory deaths would be decreased by 2,000 through implementation of new emission standards that are just about ready to be released as a first draft. The Agency of Factual Standards and Selective Omissions has determined that there is approximately a 10% chance that between 1,800 and 2,200 lives will actually be saved; that there is a 25 % chance that between 500 and 4,000 lives will actually be saved; that there is a 50% chance that between –2,000 and 10,000 lives will actually be saved . . . etc. You get the point.
If those projections don’t bear out, we would have another agency whose purpose it is to discipline the modelers who made the bad assumptions about the other agency’s assumptions. And then you can see how a scenario similar to the opening credits of Monty Python’s Holy Grail would play out with various, repeated sackings . . .
There is, of course, a more serious side to all this. Today with electronic and social media being what they are, public policy is often discussed and debated with varying amounts of ultimate ability to have any impact in venues like this Policy Pub. And in the passion of those debates it is often overly tempting to pick and choose facts & stat’s that help support our beliefs. That might help the ego through the night – but it really doesn’t do much in the interest of promoting educational discourse.
Okay, lesson over. I’ll have a Macallen 18 y/o neat please, Sam!
Image: Painting by Flemish artist Quentin Matsys around 1513