[How will the changing demographics of SUV buyers influence rollover rates?]
A big factor holding down the number of rollover deaths in SUVs until now is
that these have mostly been $35,000 luxury vehicles being bought by prosperous
middle-aged families with children who don't go out much after dark, aren't
drinking a lot. The problem is coming in that teenagers like SUVs more than any
other age group, according to the automakers' research. ... They lack the
driving skills to keep a vehicle like a sport utility on the road with all four
wheels down. Then you are probably going to see more rollovers. ...
Does it make any more sense to focus on the Explorer itself than it does to
focus on the tire itself?
There is a much bigger problem than just Explorers. Explorers have one of the
lower rates of rollover of any SUV on the market. If those same tires had been
installed on another model, it is possible that they would have had a higher
rate of crash deaths per million vehicles. But the Explorer particularly caught
people's attention because it was the best-selling family vehicle of the 1990s,
and because it became the new status symbol for baby boomers in this country.
[In 1986, Congressman Tim Wirth petitioned NHTSA to issue a rollover
standard. Can you tell us about the Wirth petition and how it affected SUV
It is rare for the regulators to have an opportunity to address a huge safety
problem at the beginning. Regulators had that opportunity with the sport
utility vehicles. The SUV market began to take off with the introduction of the
Jeep Cherokee in 1984, the first four-door sport utility really designed for
families. The regulators were taking a look at sport utility vehicles. They
were taking a look at the rollover question.
[NHTSA considered the Wirth petition and] decided to do nothing about it,
because government regulations are not allowed to ban a class of vehicles. This
is why, for example, we also didn't have rollover rules in the 1970s for
convertibles. If you were to have a rule that every vehicle had to protect its
occupants when it flipped over it would have been very hard to meet that with
convertibles. They considered that in the 1970s, and the legislation adopted by
Congress is very clear, which is [that] you can't ban a class of vehicles; you
can only address problems which are specific to one model or to a few models.
That really limited what they can do.
Again, it was the search for defect? Defect dominates the auto safety
The search for defects dominates the auto regulatory process, the search for
problems that can be addressed by changing one part as opposed to the whole
system, as opposed to forcing the change or even disappearance of an entire
class of vehicles. The object has always been to preserve customer choice, even
while tolerating sometimes fairly large risks to the driving public.
But that model relies on the market knowing, the market understanding, the
market factoring it in. And that isn't always the case, is it?
The emphasis on defects relies, to a considerable extent, on consumers being
able to tell that some classes of vehicles have safety drawbacks. And it's not
clear -- in fact, there is a lot of evidence that consumers didn't understand
that SUVs were not safer than cars, that they were more prone to rolling over.
So what happened? The regulators began drafting rules in the late 1990s to
begin providing consumers information on which vehicles were more likely to
rollover, and were quickly stopped from doing so by Congress. Only with the
Firestone mess has this finally broken loose. Finally we have numbers that are
available to the public. And lo and behold, on a scale of one to five stars,
for which five stars is the best, the sedans tend to get five stars. The
minivans tend to get four stars. And the SUVs get two or three stars -- and
occasionally, only one star -- for rollover stability.
And many deaths have occurred in between?
We have had thousands of rollover deaths in the last decade, many of which
might have been prevented if more had been done to address the rollover problem
[Is it true that] in general the plaintiff lawyers -- in suing over the
rollover issue purely, without the Firestone tire -- have not really been
successful in changing the behavior of auto companies very much? They have
settled cases, and that has [been factored in as] a cost of doing business?
[Yes,] the cost of settling rollover lawsuits has been a cost of doing business
for sport utilities. But because the vehicles have become such a fad, that cost
just vanishes into the overall profit margins that the automakers earn on SUVs.
These profit margins can reach $15,000 for a full-size SUV, which is nothing
more than an old-fashioned large pickup truck with a longer cab that covers a
couple extra rows of seats.
So litigation by itself was not a sufficient measure to really address this
problem, especially because they were settling these things confidentially?
Litigation prompted the automakers to take a few steps to address the overall
problem of rollovers, but didn't produce big changes. Publicity and regulation
tend to force much bigger changes. ...
And since publicity and the market's knowledge is so key, confidential
settlements and gag orders are especially a problem?
Confidential settlements mean that less information gets out about a vehicle's
propensity to roll over or other safety problems with vehicles. But these
settlements go on, because they are usually in the interest of the individual
client and the individual lawyer. They can get a larger settlement -- and
therefore a larger contingency fee -- if they agree to let these documents be
sealed. Now, that said, a lot of these documents get informally shared anyway
by the lawyers; but they can't introduce them in court, because the documents
are officially sealed.
What was Detroit like in the 1970s, early 1980s, when the SUV was just a
twinkle in somebody's eye?
By the early 1980s, Detroit was in terrible shape. The automakers were laying
off, not just tens of thousands, but hundreds of thousands of workers. Their
market share was plunging because of Japanese competition. High gasoline prices
had wrecked demand for the biggest and most profitable models, and very
stringent government regulations were keeping them from making the real gas
guzzlers, by requiring them to meet very high fuel economy standards for cars.
The bright idea that people at American Motors and Chrysler had was to start
building more vehicles that could be classified as light trucks. That meant
that they had to meet much more lenient fuel economy standards, so they could
have the gas guzzling engines that many Americans liked and began buying again,
once gas prices went down by 1982.
Light trucks also could meet much less stringent pollution standards, so you
didn't have a lot of very expensive catalytic converters and other equipment
like that. And they met less stringent safety standards so that you do not have
to have, for example, those head restraints behind your head to keep you from
getting whiplashed during the crash. So there are a lot of regulatory
incentives to build light trucks.
During the mid-1980s, the automakers began making lots of SUVs, beginning with
the Jeep Cherokee. What they noticed was the booming popularity of Western
attire popularized by Ronald Reagan; the rebirth of American patriotism in the
mid- to late-1980s, which prompted people to choose vehicles that had a more
military, more outdoor flavor.
The automakers actually sent engineers to watch movies like "Top Gun" and the
"Rambo" series, and tried to adopt visual cues that suggested a much more
masculine, macho image in the way they designed these vehicles, instead of the
more feminine curves that had been traditional on cars. They began making
vehicles with grills that were supposed to look to people, subconsciously, like
the teeth of a jungle cat. They began making flared fenders that were supposed
to look like the bulging muscles in a savage jaw. So these were details that
were designed to cater to a more assertive, more aggressive American
[How did the automakers save money in designing the SUV?]
When the automakers decided in the mid-1980s that they were going to build more
light trucks to take advantage of all the regulatory loopholes, they were able
to do so very cheaply, even though they didn't have a lot of truck engineers.
Ford had 12,000 car engineers and 400 light truck engineers in 1983, for
example. What did they do? They took the steel underbodies of the pickup
trucks, and they simply lowered onto them different passenger compartments and
bolted them on. ... It was very cheap. You could build pickups and SUVs on the
same assembly line.
Initially, the market segment they were after was a fairly small one. There
was a surprise here. What happened to these SUVs when they started shipping
Nobody anticipated that SUVs would be as popular as they became with urban and
suburban households. SUVs had actually been around as vehicles for hunters and
fishermen for decades. But, for example, American Motors began selling the
Cherokee at $40,000 a year and was selling 200,000 a year within five or six
years. Ford began selling the Explorer in 1990 and could not meet demand for a
It got to the point where Ford leased three Boeing 747s to airlift 200-pound
transmissions from a factory in Bordeaux to its factories making Explorers in
St. Louis, in Louisville, because they couldn't get the transmissions fast
enough by boat. Rather than wait for the transmissions, they were going to sell
every one they could.
And [what were the] profits on these vehicles?
The profits were enormous, because you were basically selling a pickup truck
for a luxury car price. It cost them $20,000, and they were selling them for
$30,000 and $35,000 apiece.
[Tell us about the Explorer's design. Did it contribute to rollover
Ford basically designed itself into a box when it came up with the Explorer in
the late 1980s. And then it struggled for a decade to get out of that box by
tweaking the design, but it couldn't really fix the basic stability problems.
Explorer was designed on a shoestring, because there was a struggle within Ford
with the more powerful car executives wanting to build another luxury sports
All of these car executives love to drive fast. And the small crew of light
truck engineers really wanted to do a sport utility. But they knew they had to
do it so cheaply that it looked like a good bargain to the Ford board of
directors. So they came up with the idea of simply lowering a longer passenger
compartment onto an existing Ford Ranger pickup trick underbody. They could
build it in the same factory, using the same tools that were already being used
to build the Ranger.
The problem was that you were putting all of this extra weight up high on that
pickup truck underbody, and that creates a rollover issue. So as this vehicle
very quickly turned into a family vehicle being driven by people at high speeds
on family vacations, and people who weren't really familiar with the extra
caution and slower speeds that were needed in a sport utility, they began
trying to tweak the design.
The problem is that whenever you build a vehicle on an existing underbody, it
is extremely expensive to change the underbody. You can make all kinds of
changes to the passenger department, but changing that underbody costs a
fortune. They made some changes to the front suspension which helped a little
after the 1994 model year, but a lot of the changes
didn't get made until they completely redesigned the vehicle for the 2002 model
So between 1990 and 2000, they are doing these little things -- but the
essential vehicle is staying the same?
The essential vehicle stayed the same from 1990 to 2000, because once you got
the basic underbody designed and engineered for a pickup truck or a sport
utility vehicle, it is very difficult and expensive to change it. You
practically end up starting with a new vehicle, particularly if you want to put
the wheels wider apart to improve stability. That is the most important thing
that you can do to improve stability. You can't widen the wheel track without
doing the whole vehicle.
Which is an unbelievably expensive and time-consuming process.
It is unbelievably expensive and time consuming to widen the wheel track of a
pickup truck or sports utility vehicle. It is much easier to make changes to
the underbody of a car, which is put together as more of a lattice of steel
parts. But with a sport utility vehicle or pickup in which you are sitting on a
steel frame, changing the design of that steel frame or widening where the
wheels are basically requires redesigning the entire vehicle at a cost of
billions of dollars. Ford just wasn't prepared to make that investment; and no
other automaker really was, either.
Is part of the story here a vehicle that was a runaway success and was in
some respects a savior for the auto industry, a profit engine for the auto
industry, and that they just couldn't get in the way of that freight
Sport utility vehicles became the motor of Detroit's financial success in the
1990s. They couldn't afford not to sell them. Real estate prices rose faster in
Detroit in the 1990s than anywhere else in the country except Salt Lake City.
The automakers' stock prices were zooming because of the profits they made.
They never made more money on vehicles than they made on the sport utility
vehicles, because they could make as much as $15,000 apiece in profits.
One factory that makes the Ford Expedition and Lincoln Navigator was making
more profits than any other factory than any other industry in the world, and
was making more profits than all but several dozen entire corporations in the
world. That is how profitable this business was.
The memos that have surfaced as a result of the [Ford-Firestone] story, the
internal Ford documents, do tell a story. What have you made of them? What is
the story that they tell to you?
Ford was practically done designing the Explorer in 1989. We know that
Consumer Reports was about to publish a report criticizing the stability
of the Explorer's predecessor, the Bronco II. So Ford engineers began working
very hard to try to tweak the design of the Explorer, to make sure it would be
more stable than the Bronco II, and more stable than other midsize sport
utility vehicles on the market. The problem was that, while you could make some
tweaks, like changing the pressure on the tires, like changing the suspension
springs, the biggest change that you can make to make a vehicle more stable is
to widen the wheel track.
You cannot widen the wheel track on a sport utility vehicle without basically
scrapping it and starting over, because that is the foundation of the whole
vehicle. And particularly if you are putting it on an existing pickup truck
underbody, it needs to be able to fit down the same width on the assembly line.
It needs to be able to have the suspensions mounted in the same place by the
same tooling. They just felt they couldn't widen the vehicle; and they didn't.
And that is in the memo?
And they put it as, "We've got to meet Job 1" [the first date of
Is this a financial decision?
There was a financial and safety decision that they had done enough to improve
safety, to make it safer than other midsize sport utility vehicles in terms of
rollovers. Therefore, they weren't going to go to the enormous additional
expense of widening the wheel track. They didn't end up doing that until they
came out with the 2002 Explorer, and they started over, basically. ...
Let's go inside the culture of auto companies. [How do automakers deal with
safety concerns?] What is the norm?
Every automaker faces many tradeoffs when they design a vehicle -- not just
Ford. Sometimes you have a choice between providing comfort or safety --
convenience, ease of access or ease of egress from vehicle or safety. There are
many, many tradeoffs you have to make when you design a vehicle. Another of the
tradeoffs is, of course, cost. If you make a vehicle that costs $100,000 to
produce, and you try to sell it as a mass vehicle, you are not going to
succeed. So they are constantly making these tradeoffs.
Where does safety fit in? The automakers try to build vehicles that are equal
and better in safety to other vehicles of the same size then on the market.
Midsize SUVs on the market in the 1980s were considerably less safe than cars,
because they were built to much more lenient regulatory standards because they
were prone to rolling over. And so being best in class among midsize sport
utility vehicles -- which was Ford's goal, which they apparently obtained --
was a fairly low target.
What is the impetus (if there is one) for an automaker to be thinking about
safety, given that they have now factored lawsuits into the cost of doing
business so they are not really threatened by lawsuits?
Automakers have four reasons to address safety. One, they don't want to get
sued, because settlements are expensive. Two, they don't want to have a recall
-- they don't want regulators to crack down on them, because recalls are
expensive. Three, they don't want bad publicity, because that can really hurt
sales. And, finally, automakers, auto executives, auto engineers do have
consciences. They tend to drive the vehicles they build, and so do their
families because they tend to get deep, deep discounts on driving them. They
don't want to get killed any more than the next guy.
They do have corporate cultures which vary. Ford actually has one of the
stronger safety corporate cultures among the automakers. Others are much less
conscious of safety than Ford. But all of this feeds into the market, which is,
if nobody is going to buy it, it doesn't matter how safe it is. So they have a
lot of tradeoffs.
And the fact is that Americans seem to like to ride high. They like to be able
to look over the cars. They like to be able to look down on other motorists.
And so they are going to sell these tall vehicles, even though tall vehicles
are more likely to roll over.
So in the last analysis, the real guarantor of safety in the design of the
vehicles is the marketplace, is us demanding it? I don't think a lot of that
was happening in the late 1980s.
The best guarantor of safety is an informed consumer who demands a safe
vehicle. And that wasn't really happening in the late 1980s and early 1990s,
when, despite frequent reports in Consumer Reports and on television
about rollovers, consumers just love the image of SUVs. These are image
vehicles. These are not vehicles that are built for safety. And partly because
the government also wasn't putting out a lot of information about the safety of
these vehicles, the result was a fad for a class of vehicles that were
inherently less stable than cars.
Is this a classic case for regulation? Does it fall on the federal
government to protect consumers in a case like this, where the market isn't
demanding it and there is no real pressure on the automakers to do anything?
Many have argued that it is a classic case for regulation when consumers cannot
readily perceive the safety drawbacks to a faddish new class of vehicles. And
yet the regulators weren't really doing that much. Part of it was that there
was political pressure -- not just from the automakers -- but also from groups
financed by the automakers, and even the public. The fact was that the public
really wanted these vehicles, and the politicians didn't want to get in their
In fact, the history of auto safety shows that auto companies not only do as
little as they can in regard to safety, but sometimes actively oppose that
progress in safety. They are, very generally, culturally opposed to it -- isn't
Automakers have a long, long history of opposing new safety innovations. They
were slow to adopt seat belts. They were slow to adopt head restraints to
reduce whiplash. They were slow to adopt air bags. They have been slow to adopt
practically every safety technology, and only in the last ten years has the
public been rather concerned about safety. [And only] as a new and younger
generation of safety-conscious auto engineers have begun to rise through the
ranks have you begun to see more of an interest in safety in Detroit. But it is
still very much a financial tradeoff. ...
Did Ford have an opportunity to make their SUVs safer over the course of a
decade and basically pass on those opportunities until the redesign [of the
Explorer]? They knew they had a problem with the Bronco II.
Ford and other automakers had lots of opportunities in the 1990s to make this
vehicle safer, and they didn't take them. All of their market research was
telling them that there was a large group of baby boomers out there who just
wanted to be as high off the road and have as macho an image as possible. They
were not that concerned with rollovers. So they didn't take steps that they
They didn't redesign the vehicles to widen the wheel track, which would have
been very expensive. But they also didn't take smaller measures as quickly as
they could have. For example, new electronic anti-skid technology helps keep
vehicles going straight, so they don't start going sideways and rollover. And
yet automakers introduced that first on their luxury cars and are only now
beginning to put it on some of their SUVs. If they had introduced that
technology first on their SUVs, cutting into these $8,000 and $10,000
[profits], it might have reduced somewhat the frequency of the rollovers.
And yet they were easily meeting federal standards. It is not as if they
were violating federal standards.
They were meeting federal standards because the rollover standards and the roof
crush standards, for that matter, are pretty lenient. So they were more than
meeting those standards, as they love to tell their critics. And yet still
there was a rollover problem.
Characterize the Bronco II, the predecessor vehicle of the Explorer. What
was known about it?
The Bronco II had a fairly high rollover rate, as did the other midsize sport
utility vehicles of the 1980s. And rather than simply start over or rather than
design SUVs on car platforms, which tend to be more stable, Ford and other
automakers chose to put out ever greater number of SUVs that were not that much
of an improvement over the sport utility vehicles of the 1980s.
The automakers were not convinced the SUVs would be the huge market that they
became. They repeatedly underestimated how big the market would be. If you look
back through the late 1980s and throughout the 1990s, practically every year
people were asking when is this boom going to peak, as it went from 2 percent
or 3 percent of the market to 18 percent and 20 percent of the market.
Everybody was saying, "Well, we will keep selling them this year," and "We are
making a lot of money, but this fad can't last, because these vehicles aren't
as practical for the typical family as cars are."
And every year, the sales kept growing. Every year, more and more people who
weren't used to driving these tall top-heavy vehicles were buying them and
putting them to family use and driving them at high speeds on the highway,
instead of driving them at ten miles an hour up dirt tracks the way they were
designed to be driven.
NHTSA actually at one point began a defect investigation into the Bronco II.
What happened with that?
A decade ago, the National Highway Traffic Safety Administration opened an
investigation into the stability of the Bronco II. There were a lot of lawsuits
pending against Ford because of Bronco II rollovers. The regulators closed the
investigation ... declaring that all sport utility vehicles rolled over
and, therefore the Bronco II was not defective because it simply shared a trait
of all sport utility vehicles.
That decision haunted the trial lawyers, because they lost a lot of money that
they spent on preparing Bronco II lawsuits. And it haunted consumers, because
it meant that the rollover issues were not really addressed. Rollovers had been
declared, not a defect, but something that simply happens in sport utility
NHTSA was one thing under the Carter administration, and then it became
something else. What was the before and after?
During the Carter administration under Joan Claybrook, NHTSA was a very
aggressive investigator. They did things that drove the automakers nuts, like,
for example, releasing television footage of the Ford Pinto bursting into
flames. That infuriated the automakers, who were delighted when President
Reagan won election. The automakers contended that the regulatory burden on
them had become excessive. NHTSA was gutted in terms of its budget. Its staff
was cut very, very deeply, and it has never really covered from that. Its
funding remains a third below where it was in 1980 after adjusting for
And you had a different kind of administrator?
In the Reagan and Bush administrations, you had a succession of administrators
who were much more cautious about criticizing the auto industry, and who tended
to go to work for the auto industry after they left government.
What do you make of that piece of the revolving door?
We have had a pattern for the last 20 years of regulators going to work for the
automakers after they leave the National Highway Traffic Safety Administration.
And that has created incentives for the regulators to be less harsh to the
people who are later going to employ them. It has also made the trial lawyers
much more reluctant to tell NHTSA about problems, because when you get a safety
investigation and then the regulators close it, the regulators then also leave
government and then begin testifying for the automakers that there was not a
problem; they investigated, and they didn't find a problem. ...
What role are the plaintiff lawyers playing in the auto safety system today?
They characterize themselves as the last line of defense, or the only
Trial lawyers play a sometimes contradictory role in automotive safety now. On
the one hand, they like to portray themselves as the last line of defense
protecting consumers. On the other hand, their obligation is not to the general
public -- their obligation is to their client. But very often, their client's
interests are different from those of the general public.
In the case of the Firestone tires, each individual client's interest was that
there be no information given to regulators, for fear that there might be an
investigation. Their interest was that the settlement and any documents
obtained be sealed, because that way they could get more money from the
automakers. The lawyers share the client's interest in the sense that they get
a third of whatever the client does [get] in the financial settlement. ...
Trial lawyers love rollover cases in particular, above any other kinds of crash
cases, because rollovers are particularly likely to produce paralysis. And as
the lawyers like to say, there is no substitute for a living but severely
impaired witness. It is hard to get a jury to be sympathetic for a dead person
whom they can't see. But because rollovers particularly cause paralysis, they
produce very, very sympathetic victims for juries. ...
Did [the lawsuits] have an effect on the culture within an auto company, on
the level of openness, on the willingness to admit defect?
The enormous size of the litigation industry has made automakers much less
willing to discuss safety issues, much less willing to acknowledge any defect
that will be used against them in court. If you look back 25 and 30 years ago,
a lot of information was available in technical papers presented at engineering
conferences. Now you seldom see that except in areas where the automakers think
they're not likely to be sued. ...
The enormous volume of litigation is both helping and hurting auto safety. On
the one hand, automakers are concerned that they don't want to pay too much
money in settlements. They don't want the trial lawyers to be able to find
documents that showed that they made the wrong decision on safety, particularly
for financial reasons. On the other hand, the trial lawyers also do not have an
incentive to share information with the government that might produce some kind
of decision that might stop further crashes by forcing a recall, but might also
produce a finding that hurts their lawsuits.
You described it as an industry, yet we think of plaintiff lawyers as being
the last freelancers, the last cowboys out there. Is there more to it than
The trial lawyers who sue automakers are not the Erin Brockoviches of this
world. These are very wealthy people who fly from meeting to meeting in their
private jets. They have much more lavish budgets even often than the
automakers' own lawyers. They can make $10 million if they get a $30 million
settlement. In the Firestone cases, they have made huge fortunes, which will
probably be used to endow all kinds of buildings at law schools. A lot of
buildings are already endowed at law schools by successful trial lawyers.
They share information extensively. With Firestone, there is a compact disc
available for a few hundred dollars giving you all the documents you need to
sue Ford or Firestone on behalf of anyone who might walk in your door
complaining about their tires. So this is a big, big highly organized industry
with warehouses of data, with exchanges of information, with conferences. It is
an industry. ...
Tell us about the incompatibility issue [with SUVs].
The tires are just one of the safety issues posed by sport utilities. A bigger
problem than the tires is that the Explorer and other sport utilities are built
in such a way that makes them extremely dangerous to cars. In fact, a federal
study found that the Explorer is 16 times as likely as the typical family car
to kill the other driver in a crash. If you look at the numbers, almost as many
people are being killed unnecessarily, additionally, in cars each year by
Explorers as died in tire-related crashes of Explorers over the last decade.
So it is a terrible problem that can't readily be addressed by little things
like changing the tires. It really requires lowering the steel rails in the
underbody. That has been done actually for the 2002 Explorer. But there are
still other sport utility vehicles out there that haven't been redesigned. And,
of course, we have got 20 million sport utilities already on the road now, that
are not only a menace to their own occupants but also to other motorists.
Is this a classic example of a safety-related problem that isn't a defect,
isn't subject to lawsuits, isn't subject to lots of media attention and
congressional attention, so it falls through the cracks of the auto safety
system -- because it doesn't fit this mold?
The incompatibility of SUVs with cars, their tendency to drive over the bumpers
and doorsills over collisions is a classic example of a huge auto safety
problem that falls through the cracks of our defect-oriented approach to
automotive safety. The media mostly focuses on cases where they've got the
trial lawyers putting all the documents together for them.
Members of Congress tend to focus mostly on cases also where they've got the
trial lawyers putting documents together for them. The regulators are focused
on areas where there is litigation to produce the documents, or where there is
a problem that can be easily and quickly fixed by replacing some auto part. The
result is that nobody really has been paying much attention to it, even though
it is a far deadlier problem than the tire problem.
Is it disingenuous for Ford to say that their SUV, or SUVs in general, are
safer than cars?
It's a myth that SUVs are safer than cars. People in SUVs die just as often as
people in cars; they just die differently. They are more likely to die in
rollovers, and they are much more likely to kill other people in the process.
Even the largest sport utility vehicles don't really make you appreciably more
safe than you would be in a large car or minivan, and you will get much better
gas mileage [in a car or minivan], you will produce a tiny amount of the air
pollution, and you will not be putting your neighbors at an enormous risk.
home · introduction · unsafe on any tire? · before you buy... · interviews · nixon & detroit
discussion · links & readings · video excerpt
FRONTLINE · wgbh · pbs online
photo copyright © richard hamilton smith/corbis
web site copyright 1995-2013 WGBH educational foundation