In anticipation of the Senate debate on the Corporate Average Fuel Economy (CAFE) program, The Heritage Foundation (THF) and the Competitive Enterprise Institute (CEI) recently hosted a panel of four recognized experts to discuss the public policy implications of the CAFE program and to examine its effects on consumers and the economy. Panelists included Dr. Robert W. Crandall, Senior Fellow in the economic study program at the Brookings Institution; Barry Felrice, Senior Manager of Regulatory Affairs for the Daimler Chrysler Corporation; Sam Kazman, General Counsel of CEI; and Dr. W. David Montgomery, Vice President of Charles River Associates. Edited excerpts from this discussion are presented below.
Congress established the Corporate Average Fuel Economy program
better known as the CAFE program in 1975. This program requires
auto manufacturers selling in the United States to meet certain
fuel efficiency standards and levels for their fleet of new cars
and light trucks comprised of pickups, minivans, and sport utility
vehicles (SUVs). The standard for passenger cars is currently 27.5
miles per gallon and 20.7 miles per gallon for light trucks. Some
in Congress, such as Senator John Kerry, want to drastically
increase standards to force manufacturers to build cars and light
trucks that would get 35 miles to the gallon by 2013.
The goals of CAFE standards were to reduce US dependence on
foreign oil and decrease the consumption of gasoline. Yet, today
the United States imports more oil than when the CAFE program was
enacted and people are driving more miles. The program has also had
Excerpts from Dr. Crandall
We needed CAFE in 1975 (or the Congress thought it did) not
because of the Arab oil embargo but because the way in which
Congress responded to it. Having put regulatory mechanisms in place
to keep the price of oil and gasoline artificially low, they needed
to offset those incorrect market signals so as not to induce
General Motors, Ford and Chrysler to continue to produce
gas-guzzling cars. Had they let the price of oil and gasoline rise
market mechanisms would have taken care of this perfectly
CAFE is a solution in search of a problem. Is the problem global
warming? Is it energy security? Is it traffic congestion and air
pollution? Or is it simply the envy of some people who can't afford
to buy large SUVs? I think that I can say confidently that except
for, perhaps, a subset of that last problem, CAFE is not the right
solution to whatever the problem advocated is.
That's the problem with any CAFE standard: it only targets a
very, very small subset of those people using energy, namely, those
people who buy new motor vehicles principally for pleasure usage.
It doesn't target those people who drive older motor vehicles. It
doesn't target those people who use oil to burn in industrial
boilers. It doesn't target those people who use oil to burn for
home heating. Therefore, it's an extremely inefficient policy
because it doesn't equate the marginal cost of the policy across
A CAFE policy does not target carbon-rich fuels but only oil
that is used solely in new vehicles. That can't be a very efficient
way to deal with the global warming problem.
The current proposal would cost something like $17 or $18
billion a year in lost consumer surplus. That's what an economy is
designed to do- to produce goods that people really want which
generates surplus for them.
To the extent that we make cars less safe through CAFE people
will trade off against it; that is, they will find ways to protect
themselves against these awful light unsafe cars by buying more
safety equipment, strapping their belts on, buying air bags, and so
Plus, the latest research suggests that air bags don't improve safety. All of the surplus in increased safety is taken up by driving less safely.
Excerpts from Barry Felrice
I will talk about the promise of technology, but also the
limitations of technology. When we talk about CAFE particularly as
part of the current debate in the Senate, it's allegedly a
technological debate. But I will argue that some of the numbers
floating around are way beyond the capabilities of technology as we
in the auto industry or, I think, anyone on this planet can
conceive in the time frames that we're talking about.
During the time of the steep rise in gasoline prices the actual fleet CAFE increased significantly. From model year '79 till about '82 it went up from about 18 to over 25 miles a gallon. Was this due to CAFE standards? I would argue that CAFE standards were somewhat if not totally irrelevant during that period of rising gasoline prices.
And that's one of the fatal flaws of the program: that is, it
assumes that we could set these CAFE standards and that they could
be met through technology. Everyone thought in the mid-seventies
that the real price of gasoline was going to rise. One constant in
the 25 years of CAFE has been the inability of everyone,
government, economists, auto manufacturers, to predict future
prices of gasoline.
So when gasoline prices rise CAFE becomes somewhat irrelevant.
When gasoline prices drop or steady then they become a constraint
on manufacturers and on consumers and they limit consumer
EPA has stated that fuel efficiency -- how much gasoline it
takes to move a ton a mile down the road -- has increased by 1-1/2
percent a year for each of the last 20 years. This means there is
more fuel-enhancing technology going into light-duty motor
Why are vehicles increasing in weight again as they have since
the initial downsizing of the early '80s? A lot of that is safety
and not just regulatory safety improvements but voluntary safety
improvements, everything from antilock brakes to side air bags to
electronic stability controls and traction controls. All these
things add weight.
And yet at the same time fuel economy has not suffered. Our engineers have been working like crazy to keep the fuel economy the same even in the face of all these increases in weight. Some of that technology, some of the benefits, are getting eaten up by moving more mass or providing more power because that's what customers want.
Industry needs sufficient lead time if CAFE standards are to
change to get a return on our investment in technology. We spent $3
billion on the new 2002 Dodge Rams which included a new assembly
plant, a new engine, and a new drive train. We probably will need
six to eight years of production to make that money back and get
some decent return on it.
If CAFE standards change too quickly and are too high, we will
have to scrap that investment where we'll lose more money, which is
hard to imagine for some of us these days. We can't get to these
new levels through technology. When that happens we start cutting
back on the least fuel-efficient vehicles which lessens our revenue
flow so then we have less revenue to invest in new technology. It's
a vicious cycle.
Some say there's a lot of technology out there that you're not
putting on your vehicles that are cost-effective for consumers. I'm
not sure many consumers want to wait fourteen years to get $3,000
[The Kerry bill] is technologically unfeasible. We know of no
basis in science, fact, data, or any report (especially the
National Academy of Sciences report) that would lead one to the
conclusion that the 35mile per gallon average level is achievable
through cost-effective technology in that timeframe (2013). It just
For example, let's look at Daimler Chrysler. Assume we could
achieve at 28-mile-a-gallon truck fleet. For us to average 35 miles
per gallon, our car fleet has to get 78 miles a gallon. Now, if
those trucks only get 27 miles a gallon, then our car fleet has to
get something like 111 miles a gallon.
So any law that talks about a combined standard has disparate
competitive effects which I would argue, the Congress is not fully
So Senator Kerry's [CAFE] bill is really not even a 35-mile-a-gallon bill by 2013. It's more like a 40-mile-a-gallon bill.
Excerpts from Sam Kazman
The debate we're seeing right now over CAFE is a fundamentally
dishonest debate. Before the National Academy of Sciences issued
its CAFE report this past summer I had never met a single defender
of CAFE who admitted it kills anyone. Now, after that report, they
still don't admit it kills anyone at all.
None of ITS defenders had ever said CAFE kills. They didn't
admit CAFE kills anyone after Bob Crandall and John Graham
published their 1989 study, which found between 2,200 and 3,900
deaths annually from CAFE. They didn't admit it kills anyone when,
in 1999, USA Today published its extensive analysis finding
over 40,000 deaths from CAFE over the lifetime of its program in
And then the National Academy of Sciences issued its report last
August, which found that CAFE kills between 1,300 and 2,600 people
per year due to its constraining effect on producing larger cars.
But CAFE's proponents still don't admit it kills anyone at all. For
that reason, the debate over CAFE is a fundamentally dishonest
Why does CAFE kill? It does so because it constrains the
production of larger cars and, in most modes of collision, larger,
heavier cars are more protective of their occupants than are small
About 50 percent of all occupant deaths occur in single-vehicle
crashes. Extra mass in a car involved in a collision with a tree or
a bridge abutment or a brick wall is incredibly protective. You
find differences in survival rates between sub-compacts and large
cars on the order of four times as great or eight times as great, a
four to eight times DELETE 'THE' higher death rate in very small
cars as in the larger cars. There is simply no question whatsoever
that in single-vehicle crashes larger, heavier cars are
The other half of all occupant deaths, however, occurs in
multi-car collisions, largely in two-car collisions, and there the
issue gets a little more complicated.
In multi-car collisions adding mass to your car protects you
more but it does put the occupants of the other car at somewhat
greater risk. And so the question is, what is the net effect?
When the two cars involved in that multi-car collision are
pretty much identical, larger mass helps the occupants of both
cars. When they're not identical but are still pretty similar to
each other, adding mass to your car protects you. It tends to hurt
the occupants of the other car, but its net effect overall is more
protection, and so society benefits from added mass.
In other scenarios involving very different cars, adding mass to
one car may reduce overall social safety. But whatever the
multi-car effect of added mass may be, it is totally outweighed by
the protection of added mass in single-car collisions. For that
reason, to use the words of Dr. Leonard Evans (head of the
International Traffic Medicine Association and a researcher in this
issue for three decades) , CAFE kills and more stringent CAFE
standards will kill even more.
What arguments do proponents of CAFE offer, especially those who
want to make CAFE more stringent?
The first argument is that new technology can get us out of this bind. It can give us much higher fuel economy and improved safety. But picture such a high-tech car in your mind and then add a few additional cubic feet and a few additional pounds to that car. Make it a little bit bigger and a little heavier.
Two things happen. Once you've made it bigger, this high-tech
car is a bit safer and it's also a bit less fuel efficient. In
short, you've still got a tradeoff even at this incredible
high-technology frontier between safety and fuel economy. And so
high technology does not get you out of this bind.
Defenders of CAFE also argue that CAFE can't be deadly because,
after all, it's endorsed by Ralph Nader, Joan Claybrook, and
Clarence Ditlow. But long before large cars became so politically
incorrect these very same folks stated very forthrightly that
larger cars are safer cars. In a 1989 magazine interview in which
he was asked for his advice for buying a safer car, Ralph Nader
said, first buy one with an air bag and second, buy a larger
In 1972, Nader and Ditlow published a book called Small on
Safety: A Critique of the Volkswagen Beetle. Page after page
has such statements as "Small size and light weight impose inherent
limitations on the degree of safety that can be built into a
What's happened? Back then large cars were not politically
incorrect. Now they are.
Why have these CAFE converts like Nader taken this view on this
position? Because for them the line has been this all along: you
want more safety, you need more government. You need another
government regulation if you want a safer product. With CAFE all of
a sudden it's exactly the opposite.
Another example of the political incorrectness of large cars and
how it has skewed consumer information can be found in Consumer
Reports. Every year Consumer Reports has an annual
buying guide issue which includes a very extensive discussion of
how to buy a safe car. You have to dig through that safety article
mid-way or more to find any mention of the fact that larger cars
are more safe. In some years it's not mentioned at all.
On the other hand, the auto industry probably has the most direct stake in accurately accessing vehicle crashworthiness. Go to the web site of the Insurance Institute for Highway Safety, and look under their "Tips for Buying a Safer Car". One of the very first factors they mention is that large cars are safer than small cars. You don't find it in Consumer Reports because I believe they have somewhat of a political agenda.
Proponents of more stringent CAFE standards always present the
image of a Ford Explorer devastating a GEO Metro, a large SUV just
killing everyone inside a subcompact when there's a collision
between the two. In the words of Adrian Lund at the Insurance
Institute, that is a highly unrepresentative example; the real
issue isn't so much the larger size of that Explorer but the small
size of that subcompact. If you want to improve auto safety you
have to do something about those subcompacts.
It is instructive to look at the Ford/Firestone tire fiasco.
According to a detailed report in Public Citizen last April, Ford
had requested Firestone to redesign the Explorer's tires so it was
more fuel efficient. Firestone attempted to do that, and the result
was a defective product. Here is an example of one company in its
quest for higher fuel economy producing a defective product. Yet,
that was never the focus of any attention whatsoever.
No one pushing for higher CAFE admits that it kills anyone. One
thing they do argue is that higher CAFE would be keep us out of all
of these blood-for- oil wars. But in a sense CAFE itself is a
blood-for-oil battle and it's waged not with knowing soldiers, not
with a military that knows that it's being put at risk, but with
civilians who have no idea that they're being put at any risk
What does happen when the public learns about the safety costs
of CAFE? Well, lately we've seen a slew of polls about how 300
percent of the American public wants higher CAFE standards. But the
questions in those polls never mention the safety issue.
Once you tell the public about the safety issue, their views about CAFE change dramatically, and that is something that folks in Congress should keep in mind. The safety issue is an issue that can change this debate as well as public opinion.
Excerpts from Dr. Montgomery
The economy is an engine for producing goods and services that
consumers want. And what you've heard about so far today is exactly
how CAFE standards fundamentally interfere with efficiently
allocating the resources we have in order to produce the goods and
services that people want. And that means they're going to be a bad
idea in terms of how well the economy operates. We need to work
through where those inefficiencies in the resource allocation
process come from and how large they are. We should also consider
other ways we could go about remedying whatever defect in the
economy CAFE standards are supposed to remedy but without so much
interference the economy.
So let me start by talking a bit about what the costs of CAFE
There are three kinds of costs that need to be considered. One
is the cost of technical fixes - what you would do to a car in
order to make it produce more fuel economy without changing
anything else about it.
Second, there's the cost of what is called mix-shifting in the
debate. Mix-shifting means meeting near-term standards. In
particular meeting tight CAFE standards requires figuring out some
way of getting motorists to buy smaller, less powerful vehicles
that are inherently more fuel efficient with constant technology
than larger vehicles.
And the third cost is the cost of lost performance and other
attributes of vehicles which consumers show that they value by
actually going out and spending more for vehicles that have these
attributes than vehicles that don't have them.
All of these [costs] tend to be either underestimated or left out completely in the studies cited by proponents of CAFE standards.
I've never seen a CAFE standard proposal that took into account
the fact that there's a cycle that manufacturers go through of
design and tooling and production which, if interrupted, can
increase costs many fold in actually getting these hypothetical
technologies into vehicles. Second, there's a cost associated with
mix-shifting. What are you going to do in order to meet the fuel
economy standard where you simply can't get cars up to 100 miles
Well, this does force you to price your product so that people
will stop buying your high-fuel consumption vehicles like trucks
and start buying your smaller, higher fuel economy vehicles
instead. And, obviously this is what occurs in the industry
whenever CAFE standards are binding from the first year they take
What CAFE standards do is force the subsidization of unsafe
purchases by causing the subsidization of purchases of these small
vehicles that then get crushed when they run into anything or when
something heavier runs into them.
This kind of cross subsidization, which is driven by a
regulatory program, induces people to choose vehicles which would
not have been their first choice if they'd been allowed to see
correct market signals as to what it costs to produce those
vehicles. That's why the cross subsidization produces an
inefficiency in our allocation of resources within the
This is basically a frustration of consumer preferences. We're
not equating the marginal cost of everything that we produce. We're
not equating marginal costs to the marginal value that the consumer
puts on something.
Manufacturers spend a great deal of time and a great deal of
money figuring out how much their customers value certain
attributes in order to figure out what package to put together to
make as much money as possible when they sell a vehicle. This is
not something they do for PR purposes. They do it for purposes of
making money. They don't it because they have a cause.
Defenders of CAFE don't take into account factors such as
acceleration, what's called noise vibration and harshness, towing
capacity, about driveability, and other attributes that consumers
value. And all of these things tend to be affected adversely by the
technologies that are generally proposed for making cars that are
cheaper and have greater fuel economy.
The one that I like best is front-wheel drive. You know why
there are not very many front-wheel drive pickup trucks? Because
they can't tow anything and towing is an important attribute for
those who purchase trucks. So if we want to pretend that trucks are
really cars, then we can impose fuel economy standards that turn
them into cars when you are losing something that has a clear,
identifiable value in the market place that consumers reveal when
they pay something for a vehicle with the characteristics it has
today. So these, I think are the costs of CAFE standards.
We (Charles River Associates) put together some estimates of the
cost of technology fixes. What we came up with was that a proposal
for something like the 28-mile-per- gallon SUV standard that if it
was implemented over a 12-year period with a careful phase-in to
match the cycle of rebuild, would increase the cost of a new SUV by
about 10 percent. Anything above 35 miles per gallon for new cars
is hard to even conceive of over the period that we're looking at
but it would add more than 20 percent to the cost of a new car.
That includes both the technology fixes and the downgrading of the
quality of the vehicle in the consumer's preference. It's not the
hardware. It's the combination of everything that either increases
the cost or increases the value of the vehicle.
So what we're talking about there is probably something like a
reduction in new car sales of about 15 percent when a standard like
this comes in and starts to bite.
The larger economic modeling we've done suggests that that would
produce about a 10-percent drop in employment in the auto industry
during the shock period.
There is clearly a shock effect in the short run.
But, remarkably, the costs we came up with were almost identical
to those that Bob Crandall talked about, which was a loss of about
$17 billion per year in consumer surplus from these changes.
CAFE standards are about the worst strategy you could think of
if you wanted to solve any of the problems that people have talked
about, as Bob mentioned, except envy. Because it is true that CAFE
standards would do something about the fact that some people have a
lifestyle campaign to prohibit SUVs in the guise of fuel economy
The problem with applying this only to new cars is that it
certainly has the effect that I just talked about on new car sales
but that's not just bad for oil companies and bad for autoworkers.
It's actually bad for consumers and it's particularly bad for the
First of all, it means that a large part of the fuel savings you
would expect from CAFE standards go away because, what happens if
people don't buy new cars? The first thing that happens is that
they keep their old cars longer. Even today old cars get worse fuel
economy than new cars, so delaying the replacement of old cars by
new cars takes back some of the fuel economy improvements.
With CAFE standards we're also imposing on people higher fuel
economy in their new cars than they would have chosen. What's that
mean? Since they're not paying a higher price for gasoline, better
fuel economy means it costs less to drive so people will drive
more. There's another take-back effect, the additional driving
that's stimulated by the CAFE standard.
So here we have two effects which I would call new-source bias.
And basically the liberal environmental bias in environmental
policy is let's require new sources of pollution, that is, new
investment that makes the economy grow, pay considerably higher
costs than what's out there already. That discourages new
investment. In this case it discourages replacing older, dirtier
cars with new cars and it encourages actually driving cars more.
That only defeats the purpose of CAFE standards.
Furthermore, the take-back effect of lower new car purchases, slower turnover of the fleet, on fuel consumption is not very large, but it's huge on air pollution because the difference between a 10-year-old car and a new car on air pollution is immense compared to the difference in fuel economy between a 10-year-old car and a new car. So, in addition to killing people in accidents , what the CAFE standards do is that they kill people through higher levels of urban air pollution because that's what you get out of this new source bias and reduced turnover of the fleet.
Thank you to panelists and audience.