FC: Cato's Jerry Taylor reply on fuel efficiency and dead Americans

[Declan McCullagh <declan () well com>]

[I expect this will be the last round unless I get some truly, remarkably 
well-argued responses. Be warned before you submit: The bar will be high! 
Previous message: http://www.politechbot.com/p-03258.html --Declan]

---

From: "Jerry Taylor" <jtaylor () cato org>
To: "Declan McCullagh" <declan () well com>
Subject: RE: Replies to "Raise fuel efficiency standards, kill Americans?"
Date: Wed, 13 Mar 2002 16:15:06 -0500

Thanks for forwarding these comments on to me, Declan.  I am happy to
comment on some of the criticisms levied at my comments on the federal CAFE
standard.

Fred Heutte from the Oregon chapter of the Sierra Club argues that auto
fatalities have been declining even as fuel efficiency has been improving.
True enough.  But that doesn't necessarily tells us how CAFE standards might
have affected the overall trend.  It is, Mr. Heutte, a multi-variant world.
It may well be (and in fact, it is almost certainly the case) that highway
fatalities would have declined even faster had not federal CAFE standards
been put in place.  That David Greene at the Oak Ridge National Laboratory
similarly embraces this "trend" argument (a logical fallacy, by the way,
that would probably flunk a college freshman) speaks volumes about Mr.
Greene's ability to think straight about statistics.  Nor is the fact that
the Union of Concerned Scientists would disagree particularly telling given
that they are primarily an activist group of environmentalists - not a trade
association of scientists.  But neither Mr. Heutte nor the others have
anything to refute the regression analyses performed by dozens of academics
over the years finding a relationship between vehicle weight and highway
fatalities.  Mr. Greene's assertion that no such analyses have ever been
done likewise tells us about his lack of knowledge in the field.  Robert
Crandall at Brookings has done such work.  So have Douglas Coate and James
VanderHoff at Rutgers (see
http://www.cato.org/pubs/regulation/regv24n1/coate.pdf for a recent
analysis).  So have a legion of others (see the recommended readings at the
end of the Coate & VanderHoff article for a brief review).

The contention offered by Mr. Greene through Mr. Heutte - that heavy cars
may save the passengers of those cars but surely increases the risks to
others in less heavy cars to such an extent that the fatality result is a
wash - does not hold up to the aforementioned statistical examinations.
Most immediately, however, the argument fails to recognize that a large
number of highway fatalities are one-car crashes.  Heavy vehicles do not
simply displace fatalities from one class of car buyers to another although
that argument gets a lot play among some circles.

The argument that Ms. Mesnikoff of the Sierra Club makes - that SUVs are
more dangerous on balance than standard automobiles because of the roll-over
problem - is a similar example of sleight-of-hand.  SUVs are indeed more
likely to roll over than standard cars.  But roll-overs represent an
extremely small constellation of highway fatalities.  What SUVs "surrender"
on the safety front in higher roll-over rates they "get back" in greater
protection in conventional crashes.  ON BALANCE SUVs provide more safety to
occupants than standard vehicles even if for one small subset of the
incidences they may prove somewhat more hazardous.  This proposition, by the
way, was tested empirically by the aforementioned Coate & VanderHoff study.
That Ms. Mesnikoff generalizes from one small aspect of the safety basket to
the entire basket of safety features at issue is, well, par for the course.

Tim (no last name of I.D. provided) asks "Let me see if I have this
straight.  Reducing emissions is a bad idea
because cars will become smaller, and therefore offer their occupants less
protection when they're hit by some idiot in a massive SUV."  No Tim, you do
not have it straight.  Reducing tailpipe emissions and improving fuel
efficiency are two separate issues and I, accordingly, did not address the
issue of tailpipe emissions at all.  CAFE standards will not reduce the
amount of pollutants that come out an SUV tailpipe.  In fact, CAFE standards
may actually increase net emissions for some period of time because tighter
CAFE requirements will increase the price of certain cars and trucks that
don't meet the median standard (the only way, after all, for auto companies
to produce the sales necessary to meet the overall mandated average fuel
efficiency standard).  This will slow down auto fleet turnover in that
consumers will hold onto their cars a bit longer before buying a new ones,
keeping older cars - and typically, the most polluting cars - on the road
longer than necessary.  Professor Andrew Kleit at Pennsylvania State
University calculates that Sen. John Kerry's proposed new CAFE standards
would for this reason increase VOC emissions by 1.87 percent, NOx emissions
by 3.41 percent, and CO emissions by 4.57 percent (I have an electronic
version of the study for those who'd like to see it).

Jamie McCarthy accuses me of falsely reporting the findings of the National
Academy of Sciences Report last year.  First of all, a good rule of thumb -
never pay any attention to press releases.  Pay attention to reports.  Press
releases are often exercises in spin but are, in any cases, summaries that
are rarely written by the authors of the report themselves.  So if you want
to know what's in the NAS report, read the NAS report, not the summary or
the press release about it.

Yes, Jamie McCarthy is right to highlight the observation taken from the NAS
report that "isolating the effects of CAFE from other factors affecting U.S.
light-duty vehicles over the past 25 years is a difficult analytical task."
That's because correlation does not necessarily equal causation.  The best
tools we have for this job are regression analyses which merely attempt to
isolate correlations to the greatest extent possible.  Yes, it's difficult.
But that's why economists get paid "the big bucks" (at least some of them,
anyway).  The results of such regressions must prove consistent with logic
and common sense or else they don't hold.  Accordingly, I'm awful surprised
that it isn't obvious to everyone that a lighter and smaller car - all
things being equal - is less safe for the occupants than a heaver and larger
car.

It is correct to note that the NAS report attributes the 1,300 - 2,600
additional deaths on the roadway every year to the reduced weight of the
average car in the U.S. auto fleet.  But most analysts - particularly most
environmentalists - are convinced that were it not for federal CAFE
standards, the auto fleet would have gone back to the pre-CAFE weight
standard after oil prices collapsed in the mid-1980s.  So this is a
distinction without a difference.  Moreover, it's worth noting that the NAS
did not independently run it's own study on the subject; the report simply
summarized the literature on the matter and offered its assessment of the
likely range of fatalities that result from lightening the weight of cars
and trucks (the primary method by which auto companies comply with the CAFE
standards).  If you go back and look at the studies the NAS cites to justify
its 1,300 - 2,600 estimate, you'll find that the relationship between CAFE
and highway fatalities is made quite explicit.

I am, by the way, amused by the manner in which some environmentalists can
turn on a dime in the various science debates.  When it comes to global
climate change, arsenic in drinking water, or a host of other issues, we are
constantly beaten about the head and shoulders about "the consensus of
scientific opinion" as if good science were simply a show of hands.
Minority reports or opinions about those and many other issues are
contemptuously dismissed as "fringe" and, well, embarrassing affronts to
mainstream opinion and confessions of ignorance.  When it comes to the CAFE
debate, however, minority opinions and reports are treated with great
reverence; the existence of dissent supposedly neutralizes the issue or even
discredits it altogether.  Now, my own opinion is that minority views are
often correct and that good science is more than a vote of scientists with
an opinion.  So it may well be that minority views on CAFE's relationship to
automobile fatalities are correct.  But the dissenting opinion in the NAS
report is, in my opinion, unpersuasive and contrary to common sense.

Jerry Taylor
director, natural resource studies
Cato Institute

---

Date: Wed, 13 Mar 2002 11:26:30 -0700
From: "Ralph S. Hoefelmeyer" <ralph.hoefelmeyer () wcom com>
Subject: RE: Replies to "Raise fuel efficiency standards, kill Americans?"
In-reply-to: <5.1.0.14.0.20020312234432.02a66830 () mail well com>
To: declan () well com, jtaylor () cato org

Declan, Jerry,
We drive a 4 door, 4 wheel drive, long bed, dual tired Chevrolet K3500
pickup about half the time.  It gets 10-12 miles a gallon.  It would be hard
to flip, with the dualies, and is safer for us in a wreck, unless we get hit
by a semi.  The other parties in a wreck would be toast.
The critics of our choice of vehicle don't get it; we do not care a whit
about the other parties in a wreck, if we are not at fault; we care much
more if we are at fault, but not enough to take greater risks to ourselves.
We are not communitarians.  As for the resource costs, those are driven by
the market; we can afford to drive our truck.
We bought the biggest passenger truck available.  The fact it uses an
internal combustion engine was decided by the market.  If we could have
gotten it powered by a fuel cell, propane or nuclear plant, we might have
done so if it was cost effective and offered in the market.
Cordially,
Ralph

<opinions are mine>

---




-------------------------------------------------------------------------
POLITECH -- Declan McCullagh's politics and technology mailing list
You may redistribute this message freely if you include this notice.
Declan McCullagh's photographs are at http://www.mccullagh.org/
To subscribe to Politech: http://www.politechbot.com/info/subscribe.html
This message is archived at http://www.politechbot.com/
-------------------------------------------------------------------------