Introduction
Since the OPEC oil embargo of 1973-74, the U.S. Congress has sought to establish an energy policy that would reduce our nation's dependence on foreign oil while preserving nonrenewable energy resources such as petroleum. The Energy Policy and Conservation Act of 1975, among other things, created a federal regulatory scheme intended to save oil by compelling vehicle manufacturers to produce more fuel efficient passenger cars and light trucks.1 This scheme is administered by the National Highway Traffic Safety Administration (NHTSA), and is known as the Corporate Average Fuel Economy (CAFE) program.2
Substantial progress in fuel economy has been achieved. A combination of regulatory and market pressures induced a doubling in new car fuel economy from about 14 miles per gallon (MPG) in 1973 to 28 MPG in 1991. The fuel economy of new light trucks, which account for a growing fraction of new vehicle sales, has increased only slightly from 18 MPG in 1979 to 21 MPG in 1991.3
While the supply of gasoline is now plentiful and prices at the pump are low (about $1.25 per gallon in the U.S.), recent world events have stimulated Congress to consider new legislation that would revamp and tighten CAFE standards. The Persian Gulf war raised questions about whether America had engaged in costly military intervention due to excessive dependence on Mideast oil. Moreover, a global warming trend is occurring that could be catastrophic if the U.S. and other nations do not reduce fossil fuel consumption. Energy conservation is seen by some policy analysts as a promising measure to address both national security and environmental concerns.4
Although fuel economy legislation is well intended, concerns have been raised that overly stringent fuel economy rules could compromise the safety of motor vehicle occupants.5 Safety concerns are rooted in the observation that, in the event of a crash, occupants of small cars are at greater danger of death and serious injury than occupants of large cars.6 Even when occupants are restrained by belt systems, small car occupants incur elevated risks of serious injury and death in crashes.7 The introduction of air bags into new cars should lessen occupant injury risk, but preliminary estimates suggest that air bags are less effective in small cars than in large cars.8. If fuel economy legislation stimulates more sales of small cars (or downsized large cars), the number of crash-related deaths and injuries on the highway may be increased compared to what would occur without tighter fuel economy legislation.
Safety concerns played surprisingly little role in the public policy debate until the Bush Administration reversed the position of the Reagan Administration, which had downplayed the link between fuel economy regulation and safety.9 One of the reasons for this reversal was growing recognition of the findings of the first peer-reviewed study of CAFE and safety, which estimated that 1989 models are 500 pounds lighter than they would have been without CAFE rules, resulting in a 14 to 27% elevation in the occupant fatality rate in these vehicles.10
While researchers at the NHTSA were originally skeptical of these findings, the research they initiated in response ultimately led to similar qualitative findings.11 Independent research performed by the Insurance Institute for Highway Safety has raised additional concerns about the adverse safety consequences of a smaller fleet of vehicles, especially when the fleet of light trucks, which are aggressive in two-vehicle crashes, is growing.12
Not everyone is convinced that ambitious fuel economy legislation will necessarily compromise vehicle safety. Some analysts have suggested that major gains in fuel economy can be achieved through technological improvements that would not compromise occupant safety.13 Others have raised important theoretical and empirical questions about the relationships between vehicle size, vehicle weight, and the safety of occupants.14 Some analysts believe that small car safety can be enhanced through a new round of safety regulation.15 Meanwhile, a series of new empirical studies have been published that shed additional light on the relative safety of small cars.16
In light of the new empirical studies and the issues raised by OTA and GAO, this report updates the Crandall and Graham study of 198917 and analyzes the potential safety risks of new fuel economy legislation. The ultimate purpose of the report is to provide a range of estimates of the safety risks of tighter fuel economy legislation, taking into account uncertainty about how vehicle manufacturers will comply with tighter CAFE standards. The safety estimates in this report are based on the most comprehensive examination of the scientific literature on vehicle dimensions and safety that has yet been published.
Since this report addresses only the safety issue, it does not offer any global recommendation about whether new fuel economy legislation should be adopted. While occupant safety is a critical concern, policy makers should consider all the costs and benefits of fuel economy legislation, including policy alternatives other than regulation of new cars.18
CAFE and New Legislative Proposals
The CAFE program requires each vehicle manufacturer to meet a fleetwide average fuel economy target for its new cars, and a separate target for its light trucks. Since manufacturers must design vehicles roughly five years before being sold, they are not always able to sell as many fuel-efficient vehicles to consumers as they had planned. Consumer interest in fuel economy is sluggish because of the low price of gasoline and the declining relative importance of fuel costs in the total costs of operating a motor vehicle.19
The 1975 legislation called for steady increases in the MPG targets for passenger cars from 18 MPG in 1978 to 27.5 MPG in 1985. The Department of Transportation now has the discretion to adjust MPG targets on a year-to-year basis, and has set a 27.5 MPG target for each of the last three years. Under the CAFE program, manufacturers may use MPG credits earned in good years to pay off MPG deficits experienced in bad years. Deficits may also be carried forward if manufacturers can show credible product plans that promise MPG credits in future years.
The collapse of gasoline prices in the early 1980's and increasing consumer interest in large, high-performance cars have caused some manufacturers to experience difficulties in meeting their CAFE targets. General Motors (GM) and Ford Motor Company (Ford) exhausted most of their fuel economy credits in the late 1980's. Mercedez and other European manufacturers of large, fuel consuming cars have chosen to pay noncompliance penalties for years.
The major domestic vehicle manufacturers are reluctant to miss CAFE targets consistently. Failure of manufacturers to meet CAFE standards is considered "unlawful conduct." The ultimate penalty for noncompliance with CAFE is $50 per MPG per vehicle of shortfall. If GM was assessed a penalty for a 1 MPG shortfall in its passenger car fleet, the total fine could exceed $150 million. Manufacturers also recognize that failure to satisfy CAFE requirements may undermine their credibility in Washington on a wide range of public policy issues. Also, major manufacturers wish to avoid engaging in behavior that Congress has deemed "unlawful". Hence, CAFE compliance is taken very seriously by most vehicle manufacturers.
Due to frustration with administration of the CAFE program, several legislators have proposed bills that would revamp and tighten fuel economy standards. Senator Richard Bryan (D-NV) has introduced the most ambitious bill, calling for a 20% increase (using 1988 as the base year) in fleetwide MPG by 1996 (to about 34 MPG) and a 40% increase by 2001 (to about 40 MPG). The Bryan bill would apply the same percentage increases to both imports and domestics, and to both passengers cars and light trucks.20 In the House of Representatives a bill introduced by Representative Barbara Boxer (D-CA) calls for a 60% increase in fuel economy by 2001.21 In order to assess the safety consequences of such proposals, it is necessary to consider how manufacturers might respond to intensified regulatory pressures.
Compliance Strategies
Faced with more stringent fuel economy targets, manufacturers would need to modify their product designs and marketing plans. The following five, long-term strategies might be employed to comply with new legislation:
* Make technological improvements in the design of engines, transmissions, tires and vehicles that improve fuel economy without reducing interior volume, vehicle weight, or the external dimensions of the vehicle;
Reduce vehicle performance by curtailing engine size, horsepower, and acceleration capability;
* Achieve vehicle weight reduction by substitution of lighter materials and incorporation of other technological improvements that facilitate weight reduction without reducing interior volume or the external dimensions of the vehicle;
* Engage in vehicle downsizing by reducing the exterior dimensions of the vehicle and associated vehicle weight; and
* Engage in mix shifting toward smaller, lighter vehicles by discontinuing certain lines of large cars and/or offering more marketing incentives to encourage sales of small cars.
Since the future decisions of vehicle producers and consumers are difficult to predict, no one knows exactly what combination of strategies would be instituted to meet, say, the ambitious MPG targets in the Bryan bill. As we shall see, the strategy chosen by manufacturers, which is not constrained in recent legislative proposals, would influence strongly the magnitude of the safety consequences that are likely to occur.
Learning from History
Perhaps the most instructive exercise is to examine how manufacturers achieved the doubling of new car fuel economy since 1973. All of the strategies cited above played a significant role.
Technological modifications of the vehicle and its components gave an important boost to new car fuel economy. The key technical improvements included the proliferation of automatic transmissions with lockup, the rapid introduction of computerized fuel injection systems, and the transition from rear-wheel to front-wheel drive (and associated weight reductions). Figure 1 tracks the penetration of improvements into the new car fleet.22
Vehicle size and weight were also reduced substantially to improve fuel economy. The average new car's "shadow," defined as vehicle length times vehicle width, has declined 16% from model year 1974 to model year 1990. The curb weight of an average new car sold in the U.S. has declined by about 20% since 1974.23 These declines in vehicle size and weight are estimated to account for roughly half of the improvement in new car fuel economy achieved since 1974.24
Figure 1
Fuel Economy Boosters 1977-90 [Omitted]
Table 1 reveals how Ford's 1972 and 1992 new-car offerings differ in curb weight and shadow. While both shadow and weight have declined, most of the changes have occurred in the mid-size and large cars. This pattern is mirrored in all new car sales. Vehicle weight reductions have been achieved primarily in the medium and heavy weight classes, which has resulted in a less variable distribution of car weights than existed ten years ago.25
The fraction of the new car market accounted for by large cars declined from about 60% in 1978 to 40% in 1990.26 Some manufacturers, such as Chrysler Corporation, have reduced their offerings of large cars. Other manufacturers, such as Ford, have sought to stimulate small car sales through special pricing, incentive, and advertising strategies.27
Table 1
Downsizing of Ford Cars, 1972-92 [Omitted]
While performance reduction played a positive role in the early years of the CAFE program, consumers have rejected this strategy. Perhaps the most important drag on fuel economy has been renewed consumer interest in engine horsepower. As Figure 2 indicates, average horsepower per pound of vehicle curb weight has been increasing rapidly since 1980, and is now exceeding the levels achieved by the "muscle" cars of the 1960's.28
Figure 2
1955-91 Trends in New Cars: Horsepower per Curb Weight [Omitted]
While the passenger car fleet has become lighter and somewhat more uniform in curb weight, sales of trucks (particularly light trucks) have been increasing rapidly since 1980. Light trucks alone have increased from 17% of the "light-duty" market in 1980 to 32% in 1991.29 While light trucks are subject to fuel economy regulation, they are much less fuel efficient than the average passenger car.
In light of contemporary market pressures, manufacturers would find it extremely difficult to achieve the ambitious MPG targets outlined in the Bryan bill.30 In the absence of a crystal ball, the most sensible prediction would seem to be that manufacturers will respond to intensified fuel economy demands by employing a combination of strategies used between 1974 to 1990. While the precise mix of strategies may change due to subtle changes in technology, economics and consumer demand, there is no reason to expect that any one will play a dominant or exclusive role.
Key Safety Relationships
This analysis of vehicle safety begins from the premise that fuel economy legislation will not change the frequency of vehicle crashes. (This premise is discussed further and ultimately relaxed below). The key concern that has been raised is whether a smaller, lighter fleet of vehicles induced by regulatory pressure will experience a higher rate of occupant injury and death in crashes that do occur (sometimes called diminished "crashworthiness").
As the real-world crash data in Figure 3 suggest, small car occupants face greater risks of injury in crashes than do occupants of large cars. Note that collision frequency has been held constant in this comparison by dividing the number of driver injuries by the number of tow-away crashes. The effect of car dimensions on injury risk persists when potential confounding factors such as driver age, crash severity, and type of vehicle damage are controlled.31 The experimental crash-test data from the NHTSA's New Car Assessment Program (NACP) also indicate that small car occupants are exposed to more injurious crash forces than are occupants of large cars, holding constant crash severity and restraint use.32
Figure 3\33
Probability of Driver Injury (AIS >= 3) in a Tow-away Crash, by Car Size [Omitted]
In Figure 3, the disadvantage of the small car occupant is apparent in both single-vehicle and multi-vehicle crashes. The importance of the single-vehicle crash tends to be neglected in discussions of the safety issue.34 Single-vehicle and multi-vehicle crashes each account for about half of all passenger car occupant fatalities.35 Hence, both crash types merit serious investigation.
Injury Risk in Multi-Vehicle Crashes
Why are small-car occupants at greater danger than large-car occupants in multi-vehicle crashes? Both vehicle weight and vehicle size (e.g., exterior dimensions or shadow) play a role. We begin by analyzing the role of vehicle weight and then address the role of vehicle size.
Physical intuition suggests that the lighter the vehicle in a multi-vehicle crash, the more energy (and hence risk) will be transferred to its occupants. When light cars are struck by heavy cars, the occupants of light cars are in greater danger of injury than are the occupants of heavy cars.36 Yet lighter vehicles transfer less energy (and hence risk) to the occupants of other vehicle(s) in a multi-vehicle crash.37 Hence, light vehicles make a contribution to safety because they are less "aggressive" in multi-vehicle crashes than are heavy vehicles.
If we imagine one car being made lighter in a two-car crash, the occupant risk in the car with diminished weight should increase while the occupant risk in the second car should decline. Physical intuition does not tell us whether the total amount of injury risk in the two cars will increase or decrease.38 The two effects might cancel each other completely, or one effect might outweigh the other. A recent simulation study illustrates that either outcome is possible.39 Only good empirical evidence can resolve this mystery.
A recent study of two-vehicle crashes in Texas examined the impacts of the weights of vehicle 1 and vehicle 2 on the probability of serious driver injury in vehicle 1.40 When crash type, crash speed, driver age and driver sex are held constant, the investigators found, as expected, that the weights of the two vehicles exerted opposing effects. More weight in vehicle 1 decreased the risk of injury to the driver of vehicle 1 while more weight in vehicle 2 increased the risk of injury to the driver in vehicle 1.
Interestingly, each 100 pounds of weight subtracted from vehicle 1 was associated with a net increase in net risk to both drivers. The sum of the injury risk to both drivers increased by 1.4% for each 100 pound reduction in the weight of vehicle 1.41 This result replicates a similar finding reported in the 1970's.42 The finding should be interpreted cautiously because the protective effect of the extra weight of vehicle 1 may partly reflect the protective effect of more vehicle size (e.g., external dimensions) that is usually associated with extra vehicle weight.
It is natural to speculate about whether a fleet of vehicles of uniform weight is more or less safe than a mixture of heavy and light vehicles, holding constant the mean weight of the fleet. Early theoretical and empirical work suggested that a uniform weight distribution produces more fleet safety than a variable distribution of vehicle weights.43 The most recent empirical study came to the opposite conclusion,44 suggesting that a fleet of variable vehicle weights is safer than a fleet of uniform weight (holding constant mean fleet weight). In both studies the magnitude of the difference is slight and can safely be ignored in first-order calculations of the safety impacts of smaller new cars.
While vehicle weight is a mixed blessing in two-vehicle crashes, vehicle size is unequivocally protective. Physical intuition suggests that more vehicle size (e.g., more vehicle shadow and structure) is a protective factor in a multi-vehicle crash.45 Increased vehicle size makes more "crush space" available to manage energy, and more space to "ride down" the crash. A large car occupant is decelerated over a longer distance than a small car occupant and therefore is exposed to weaker crash forces.46 In two-vehicle crashes, increased vehicle size provides superior occupant protection without imposing extra risk on the occupants of the other vehicle (assuming more size is provided without adding weight).
Why are small cars at such a disadvantage in crashes with large cars? Is it their lack of size and inferior crush space, or is it their weight disadvantage? This question is difficult to answer because vehicle size and weight are highly correlated. The only study to address this question used wheelbase as the indicator of vehicle size. When cars of similar wheelbase collide, vehicle weight is a strong predictor of driver fatality risk. When cars of similar weight collide, wheelbase is a weak predictor of driver fatality risk. This evidence suggests that the disadvantage of the small car occupant in multi-vehicle crashes is attributable primarily to its weight disadvantage rather than to inferior size and less crashworthiness.47
There is, however, suggestive evidence that more vehicle size offers significant occupant protection in multi-vehicle crashes. Several independent studies have found that more people are injured or killed when two small cars collide than when two large cars collide.48 Since the opposing vehicle weights (and hence crash forces) are roughly equal in the two collisions (assuming speed at impact is comparable), most analysts attribute the disadvantage of the small car occupants to smaller vehicle size (i.e., smaller exterior dimensions).
Moreover, in two-vehicle crashes involving one vehicle of essentially infinite weight (e.g., a large truck), occupants of small cars fare worse than occupants of large cars. Since the weight difference between the large and small car should make little difference when striking a large truck (unless the exterior of the truck is penetrable or crushable), analysts suspect that it is extra vehicle size (perhaps more structure and crush space) that offers the superior protection to occupants in the large car.49 Some analysts believe that light-truck/small-car collisions will produce more deaths and serious injuries than large-car/small-car collisions due to the stiff design of light trucks (which is harmful to the occupants of both vehicles). This hypothesis deserves additional study in the future.
In summary, more vehicle size (e.g., larger exterior dimensions) enhances fleet safety in multi-vehicle crashes due to improved crashworthiness. Extra vehicle weight helps those occupants who have it and hurts those who are struck by it due to the aggressivity effect.50 The overall safety impact of reducing vehicle weight (without reducing vehicle size) in multi-vehicle crashes is unknown. The outcome may depend significantly on the relative strength and flexibility of substitute materials used to accomplish weight reduction.51 If both vehicle size and weight are reduced, the net impact on occupant safety in multi-vehicle crashes is negative, although there may be a few circumstances when the safety benefits from reduced aggressivity are dominant.52 In the analysis that follows, we assume that weight reduction alone (without size reduction) causes no net increase in the number of deaths and injuries in multi-vehicle crashes.
Injury Risk in Single-Vehicle Crashes
If a vehicle strikes an absolutely immovable, unbreakable, impenetrable object, the vehicle's weight per se has no impact on the occupant's risk of injury. When the object is movable, breakable, or penetrable, more vehicle weight reduces the risk of occupant injury because the object absorbs some of the crash energy. Since many guardrails, sign posts, walls, bushes and trees are somewhat movable, breakable, or penetrable, we should expect occupants of heavy vehicles to suffer fewer deaths and serious injuries in single-vehicle crashes than occupants of light vehicles.53 While the aggressivity of heavier vehicles is harmful to occupants of other vehicles in multi-vehicle crashes, aggressivity is a protective feature for occupants in single-vehicle crashes.
The magnitude of the protective effect of vehicle weight per se is unknown because studies of real-world, single-vehicle crashes have not disentangled the effects of vehicle size and weight. Vehicle size exerts a protective effect in single-vehicle crashes for precisely the same reasons that it protects occupants in multi-vehicle crashes. More size and structure leads to better energy management and less energy applied to the occupant.
The combined impact of vehicle size and weight on injury risk in single-vehicle crashes has been studied by numerous investigators. While some investigators have found equivocal evidence of an association,54 many studies find a systematic association between vehicle size/weight and injury risk in single-vehicle collisions.55 Figure 4 presents the findings of one of the more recent, well-designed studies of this question.56 One recent study estimates that each 100 pound reduction in vehicle weight/size is associated with a 1.0% increase in the risk of injury in single-vehicle crashes.57 Crash tests of vehicles into bridge rails with varying degrees of flexibility suggest a similar protective effect of vehicle size and weight.58
The point has been made that some small cars outperform in crash tests than some large cars.59 While true (and may reflect better small car design), the statements is somewhat misleading. Fixed barriers used in crash tests are designed to be impenetrable and immovable, which removes weight advantages of large cars. In real-world, single-vehicle crashes, barriers are often penetrable, breakable or movable.60
Figure 4\61
The increasing population of small cars has been a special concern to highway
engineers because many guardrails, sign posts and highway fixtures were not
designed for such small vehicles. Small cars may be caught in guardrails or
unable to dislodge breakaway sign posts, at least until highways are redesigned
to accommodate smaller, lighter vehicles.62
Some analysts have speculated that car weight may be a hostile force in
crashes involving pedestrians.63 In fact, the studies on car weight and
pedestrian injury have found no evidence that car weight affects the severity
of pedestrian injury.64 This finding is reasonable on physical grounds since
the lightest car is much heavier than the heaviest pedestrian.
Vehicle Dimensions and Collision Frequency
Intuition suggests that small cars are more maneuverable than large cars.65 Is
there any evidence that small cars are less likely to be involved in crashes
than large cars? The answer to this question is not known with certainty
because driver behavior dominates vehicle attributes as a predictor of
collision frequency. Powerful human factors (such as driver risk taking,
vision, skill, attentiveness and inebriation) are not measured reliably in
crash data systems. Hence, it is difficult to determine whether vehicle
dimensions play a significant role in collision frequency.
Analysts generally find that small cars have higher crude rates of collision
(per registered vehicle or per vehicle mile of travel) than do large cars.66
This association is not very meaningful because the drivers are not the same in
small cars and large cars, and small cars tend to be driven in urban areas
where collision rates are higher. A single confounding variable, driver age,
exerts a dramatic influence on this association.67 Young drivers are more
likely to be involved in collisions than old drivers, excluding the oldest
drivers.68
When analysts control for driver age in studies of vehicle size and collision
frequency, no consistent association is found.69 Some analysts find that small
cars are in fewer collisions,70 while other analysts find that small cars have
higher age-adjusted collision frequencies.71 Since the conflicting studies
often use different definitions of collisions (e.g., police-reported crashes
versus crashes that lead to an insurance claim), it is difficult to resolve the
discrepancies. None of these studies control for other important behavioral
attributes such as driver attitudes toward risk taking and the degree of driver
inebriation.
Some analysts have speculated that large cars, which tend to be equipped with
powerful engines, may be involved in more crashes than small cars due to
speeding or rapid acceleration. The only empirical study of this effect did not
control for differences in driver behavior.72 The opposite phenomenon may also
occur to some extent: Cars with greater acceleration capability may avoid some
crashes (e.g., in passing maneuvers) that cars with less acceleration
capability cannot avoid. Overall, there is no sound scientific evidence that
the acceleration capability of vehicles per se is a major factor in collision
frequency.
Another possibility is that pedestrians and cyclists are more likely to be hit
by big vehicles than be small vehicles. Not only is there a difference in
maneuverability and stopping distance, but there is the simple matter of
vehicle width that, all else equal, should influence the probability that a
pedestrian or cyclist is struck in various collision modes. This hypothesis
deserves careful study, although driver and pedestrian/cyclist behaviors also
need to be carefully considered.
The only causal relationship between car dimensions and collision frequency
that is fairly secure involves the rollover crash, which is a crash mode that
is likely to result in serious injury or death to occupants. A vehicle's
directional stability refers to a vehicle's likelihood of off-road excursion. A
vehicle's rollover stability refers to the tendency of the vehicle to remain
upright during an off-road excursion that is interrupted by a tripping
mechanism. Small cars tend to have less directional and rollover stability than
large cars.73
Early empirical studies found that small cars have larger rates of rollover
fatality than large cars.74 No evidence was found that these differences were
related to differences in vehicle crashworthiness. The differences in rollover
propensity are difficult to attribute to vehicle dimensions per se because
driver behavior may be a confounding factor.
A recent study used a vehicle's frontal crash fatality rate as a control for
driver behavior and still found that small cars have higher rates of fatal
rollover crashes than do large cars. Each 100 pounds of added vehicle weight
was associated with a 3.6% decrease in the rollover fatality rate.75 This study
was not able to determine which vehicle dimensions (weight, length, width, or
height) were responsible for the propensity of small cars to be involved in
rollover crashes, although the vehicle's center of gravity is clearly
important. Since this estimate is based on data from 1970-1982 models, it may
not be indicative of the experiences of cars produced today.
In summary, the safety literature has not established a causal relationship
between vehicle size and collision frequency,76 except in the case of rollover
crashes. The difficulty in establishing such relationships, if they exist, is
the dominant influence of driver behavior on crash frequency. A method and data
base for studying vehicle dimensions and collision frequency is urgently
needed.
Summarizing Recent Empirical Estimates
Our review of the safety literature suggests that decreasing vehicle size and
weight will increase the overall (net) risk of injury to occupants in
multi-vehicle and single-vehicle crashes, despite the beneficial influence of
reduced vehicle aggressivity in multi-vehicle crashes. If vehicle weight is
reduced without reducing vehicle size, then no net increase in occupant injury
risk is predicted in multi-vehicle crashes, although some elevation of risk
will certainly occur in single-vehicle crashes. There is no strong empirical
evidence suggesting that vehicle dimensions affect collision frequencies,
except for the frequency of rollover crashes.
In Table 2, recent empirical estimates of the impact of vehicle size and
weight on injury and fatality risk are summarized. While vehicle weight is
reported as the primary independent variable, it should be understood as
representing the joint effects of vehicle size and weight. The estimates in
Table 2 have been selected on the basis of recency of data and quality of study
design. The data based on passenger cars are also applied to light trucks,
which introduces an unknown degree of uncertainty into the analysis. These
estimates are used in the next section to make a range of estimates of the
impact of new fuel economy legislation on traffic injury counts.
Table 2
Estimating the Safety Consequences of New Legislation
In order to estimate the safety consequences of new legislation, it is
necessary to predict how manufacturers will respond to ambitious fuel economy
targets, such as those prescribed in the Bryan bill. For illustrative purposes,
we consider an optimistic scenario, a pessimistic scenario and a most-likely
scenario.
Under the optimistic scenario, vehicle manufacturers achieve ambitious fuel
economy targets primarily through technological improvements,77 and a modest
reduction in average vehicle weight (300 pounds per vehicle). A small amount of
downsizing and mix shifting causes a slight reduction in vehicle size, and an
associated 100 pound reduction in vehicle weight. Overall, the scenario entails
a 400 pound decline in average vehicle weight. Readers should recognize that
this scenario is extremely optimistic. The recent OTA study, which has produced
optimistic estimates of fuel economy gains from new technology, states that
manufacturers could not meet the early MPG target in the Bryan bill without
some mix shifting or downsizing, especially if consumer interest in engine
horsepower continues to grow.78 The NAS study concluded that such targets are
not achievable in the prescribed time frame.79
Under the pessimistic scenario, manufacturers make little incremental progress
through technological improvements and are forced to rely primarily on size
reduction and mix shifting.80 The result is a 900 pound reduction in average
vehicle weight with corresponding declines in the exterior dimensions of the
vehicle fleet. Scientists at Ford believe that a 600-1000 pound reduction in
size and weight would be necessary to achieve the 40 MPH targets.81 This
scenario is pessimistic because it grants little plausibility to the various
technological improvements that have been proposed to improve fuel economy.
Under the most-likely scenario, manufacturers engage in a mix of the
compliance strategies described earlier. They accomplish part of the fuel
economy gain by technological improvements and substitution of lighter
materials. This leads to a 300 pound weight reduction, as described in the
optimistic scenario. The remainder of the fuel economy gain is achieved by
another 300 pound reduction in vehicle weight, which is caused by corresponding
declines in the size of the vehicle fleet (due to downsizing and mix shifting).
This scenario is considered most likely because the historical evidence
suggests that manufacturers tend to use a mix of strategies to comply with fuel
economy regulation. For purposes of policy analysis, the actual behavior of
manufacturers is more relevant than technological possibilities whose
probability of implementation is remote.
The baseline scenario (without new fuel economy legislation) is a safety toll
of 30,000 fatalities and 150,000 serious injuries in crashes involving
passenger cars and light trucks. A nonfatal "serious injury," defined as at
least level 3 on the Abbreviated Injury Scale, would entail at least eight days
of hospitalization and four weeks of lost work.82 The most serious cases entail
lifelong disablement and disfigurement. The baseline counts corresponds roughly
to the fatality and injury experience of vehicle occupants in the late
1980's.83
The incremental safety impacts of the three compliance scenarios are presented
in Table 3. Steady-state estimates of the safety impact of new legislation were
calculated under the assumption of 100% fleet penetration of redesigned
vehicles. In reality, it would take 15 years for new vehicles to replace most
of the existing fleet. During this transition period, it is assumed that the
net effect of other safety policies (e.g., air bags and relaxed speed limits)
is no net change in the baseline counts of fatalities and serious injuries. An
independent panel of safety experts concluded recently that, due to various
offsetting influences, it is unlikely that the annual traffic injury counts
will improve or worsen dramatically in the foreseeable future.84
The estimates in Table 3 are interesting in several respects. First, the
safety consequences of new fuel economy legislation are adverse in all three
compliance scenarios. In particular, single-vehicle crashes become more
dangerous to vehicle occupants under each compliance scenario. Second, the
magnitude of the adverse safety effects vary dramatically in the three
scenarios. The compliance decisions of manufacturers emerge as the critical
factor in the safety analysis. Third, the uncertainties reported in the
optimistic and most-likely scenarios are significant. This arises from the
inability of safety analysts to distinguish the impact of vehicle size and
vehicle weight on injury risk in single-vehicle crashes. Finally, the results
of the most-likely scenario are broadly consistent with the illustrative
predictions of more complex simulation models.85
Table 3
Despite the uncertainties, the results in Table 3 lend credence to the safety
concerns that have been raised about tighter fuel economy legislation. The
findings also suggest that the adverse safety impacts can be minimized (but not
eliminated) if legislation is modified to discourage downsizing and mix
shifting. If the optimistic claims made for technological improvements prove to
be illusory or exaggerated, the adverse safety consequences of the Bryan bill
could be quite large.
Offsetting Behavioral Responses
Insofar as new fuel economy legislation does induce smaller, lighter vehicles
that pose greater danger to vehicle occupants, some drivers may respond by
taking greater precautions through fastening safety belts and driving more
cautiously. It is difficult to predict how much offsetting behavior might
result from increases (say, 5%) in occupant injury risk that have been
projected to occur in the most-likely scenario.
Some studies suggest that drivers of small cars compensate somewhat for the
extra dangers that will transpire in the event of a crash. For example, one
study estimates that occupant death rates in small cars are only half as large
as they were projected to be on the basis of physical calculations.86 There is
also some observational evidence that drivers of small cars take less risk in
everyday driving than to drivers of large cars87. The behaviors observed
include separation between vehicles in heavy freeway traffic, speed on a
two-lane road and belt use. The inverse association between car mass and belt
use may be confounded by factors such as geography, vehicle age, vehicle
manufacturer and belt design.88
While some offsetting behavior may occur, there is no evidence that such
behavior is profound enough totally to offset the physical disadvantages of a
smaller, lighter fleet of vehicles. Long-term, time-series analysis of highway
fatality rates in the U.S. has demonstrated that vehicle occupant fatality
rates are inversely associated with the mean weight of the vehicle fleet, after
other highway, vehicle, and driver characteristics are controlled.89 The
findings in these studies suggest that little offsetting behavior occurs,
although these macro-level studies are prone to specification error.90
There is also some direct evidence that occupant death rates increased in
those makes of GM cars that were downsized in the late 1970's.91 Although this
evidence provides little indication of offsetting behavior, it cannot be used
to estimate fleetwide safety impacts because it does not incorporate the safety
benefits of downsizing attributable to diminished aggressivity in multi-vehicle
crashes.92
More Safety Regulation?
If new fuel economy legislation does exacerbate the highway safety problem,
one might expect that NHTSA could respond with new safety regulations to
prevent the higher risk to occupants.93 While this response might seem logical,
it would not necessarily occur.
The technical basis for future motor vehicle safety standards has been
undercut by a persistent underfunding of biomechanics and safety engineering
research.94 Moreover, the effectiveness of NHTSA's vehicle safety program has
tapered considerably since its aggressive start in the late 1960's.95 The
1980's revealed that the intensity of NHTSA's regulatory program is highly
vulnerable to political apathy in the Administration and Congress -- and to
economic troubles in the auto industry.96 Since the federal government is
likely to be placing increasing pressure on the auto industry to reduce
emissions from tailpipes, it is not clear that sufficient political commitment
will exist to add another round of safety regulations.
Conclusion
The basic finding of this report is that ambitious fuel economy legislation,
such as proposed in the Bryan bill, will result in more deaths and injuries on
our nation's highways than would otherwise occur. The precise magnitude of this
effect is unknown because it is not clear how manufacturers will comply with
stricter fuel economy targets. If significant vehicle downsizing and mix
shifting occur, the safety toll is likely to be large. If fuel economy targets
are achieved entirely through technological improvements, the safety toll will
be modest. In the most likely scenario, the effect will be a significant
increase in traffic fatalities and serious injuries.
The physical intuition behind the roles of vehicle size and weight in
determining occupant safety can be summarized. If two colliding cars have the
same size and crush space but different weights, the occupants of the lighter
(heavier) car experience a greater (lesser) velocity change in a collision and
therefore have a higher (lower) fatality risk. If the lighter car was smaller
and had less crush space, its occupants would face even higher fatality risks
in the collision due to less crashworthiness.97 If a vehicle is made lighter
without being made smaller, the net effect on occupant safety in multi-vehicle
crashes is unknown. In the single-vehicle crash, both vehicle size and weight
work to protect vehicle occupants.
If we consider ambitious MPG targets such as those in the Bryan bill, the most
likely, steady-state estimate of safety impacts is 1650 additional fatalities
and 8500 additional serious injuries per year, or about a 5% increase above
baseline assumptions. It is difficult to be optimistic that new safety rules
would offset this increase in occupant risk since few rules in the history of
NHTSA have been able to achieve a 5% improvement in vehicle occupant safety.
While these figures assume that the best predictor of the compliance behavior
of manufacturers is historical experience, the degree of uncertainty in these
estimates is large. Some of the projected increase in safety risk may be
compensated for by improved driver behavior and new safety standards, but there
is little reason to be confident that the predicted safety risks will be
eliminated.
In the final analysis, policy makers should recognize the adverse safety
consequences of new fuel economy legislation. These adverse impacts need to be
weighed against the societal benefits of improved fuel economy that are
anticipated. Policy makers should consider modifications to current legislative
proposals that might minimize adverse safety consequences without compromising
the benefits of fuel economy legislation. For example, more realistic MPG
targets and compliance schedules might be achieved by manufacturers without
significant safety risks.98 In the long run, policy makers should also consider
new policy options that can simultaneously save lives and oil. More
consideration of lower speed limits on two-lane rural highways99 and gradual
increases in gasoline taxes100 might be fruitful steps toward this end. While
gasoline prices would encourage a smaller new vehicle fleet, they would save
lives by reducing discretionary driving.101
---
Notes
* This report was supported by a grant from the Centers for Disease Control of
the U.S. Public Health Service to the Harvard Injury Control Center. The author
thanks the following individuals for providing helpful comments and
information: Robert Crandall, Leonard Evans, Michael Finkelstein, Dana
Gelb-Safran, David Greene, H.R. Hackney, Tara Hill, Hans Joksch, Charles
Kahane, Orron Kee, Leroy Lindgren, Adrian Lund, Terry Klein, David Kulp,
Patricia Levy, Brian O'Neill, Susan Partyka, Steve Plotkin, Deborah Servi,
Louis Walton, and Tom Wells. The author is responsible for all errors and
conclusions.
** Professor Graham received his B.A. (economics) from Wake Forest University,
his M.A. from Duke University, and his Ph.D. from Carnegie Mellon University.
He is the Director of the Center for Risk Analysis at the Harvard School of
Public Health and Director of the Harvard Injury Control Center.
1 Pub. L. 94-163, 89 Stat. 871, at 902. Provisions governing the average fuel
economy standards are codified at 15 U.S.C. Sec. 2002 (198).
2 See, e.g., 49 C.F.R. Part 525 (1991). See also, e.g., NHTSA, AUTOMOTIVE FUEL
ECONOMY PROGRAM (15th Ann. Rpt. to Congress, 1991).
3 Id.
4 OFFICE OF TECHNOLOGY ASSESSMENT (OTA), IMPROVING AUTOMOBILE FUEL ECONOMY: NEW
STANDARDS, NEW APPROACHES (1991).
5 Insurance Institute for Highway Safety, Where is Safety in the Fuel Economy
Debate? Status Report: Highway Loss Reduction, Sept. 8, 1990, at 1.
6 J. KIHLBERG, E. NARRAGON, B. CAMPBELL, AUTOMOBILE CRASH INJURY IN RELATION TO
CAR SIZE (Cornell Aero. Lab. Rpt. VJ-1823-R11, 1964); Campbell, Driver Injury
in Automobile Accidents Involving Certain Car Models, 2 J. SAFETY RESH. 207
(1970); B.CAMPBELL & D. REINFURT, RELATIONSHIP BETWEEN DRIVER CRASH INJURY
AND PASSENGER CAR WEIGHT (Hghwy. Safety Resh. Ctr., U. N. Car. 1973); J. O'DAY
& R. KAPLAN, HOW MUCH SAFER ARE YOU IN A LARGE CAR? 5 HIT Lab Reports, May
1975, at 1; O'Neill, Ginsburg & Robertson, The Effects of Vehicle Size on
Passenger Car Occupant Death Rates (Soc'y Auto. Eng. Tech. Paper Ser. 770808,
1977); H JOKSCH, INJURY CLAIM RISK IN 1977-79 CARS IN RELATION TO CAR WEIGHT
(Ctr. Envir. and Man 1982).
7 Jones & Whitfield, The Effects of Restraint Use and Mass in "Downsized"
Cars (Soc'y Auto. Eng. Tech. Paper Ser. 840199, 1984); Evans, Fatality Risk for
Belted Drivers Versus Car Mass, 17 ACCIDENT ANAL. & PREV. 251 (1985).
8 HIGHWAY LOSS DATA INSTITUTE, DRIVER INJURY EXPERIENCE IN 1990 MODELS EQUIPPED
WITH AIR BAGS OR AUTOMATIC BELTS (Insur. Spec. Rpt. A-38, 1991); P. ZADOR &
M. CICCONE, DRIVER FATALITIES IN FRONTAL IMPACTS: COMPARISONS BETWEEN CARS WITH
AIR BAGS AND MANUAL BELTS (Insur. Inst. Hghwy. Safety 1991).
9 Personal communication from S. Kazman of the Competitive Enterprise
Institute, Washington, D.C. (1991).
10 Crandall & Graham, The Effect of Fuel Economy Standards on Auto Safety,
32 J. LAW & ECON. 97 (1989).
11 Partyka, Registration-Based Fatality Rates by Car Size from 1978 through
1987, in PAPERS ON CAR SIZE: SAFETY AND TRENDS (NHTSA 1989); Partyka &
Boehly, Passenger Car Weight and Injury Severity in Single Vehicle Nonrollover
Crashes, in THE EFFECT OF CAR SIZE ON FATALITY AND INJURY RISK IN
SINGLE-VEHICLE CRASHES 14 (NHTSA 1990); Kahane, Effect of Car Size on the
Frequency and Severity of Rollover Crashes, in THE EFFECT OF CAR SIZE ON
FATALITY AND INJURY RISK IN SINGLE-VEHICLE CRASHES 28 (NHTSA 1990); T. KLEIN,
E. HERTZ, & S. BORENER, A COLLECTION OF RECENT ANALYSES OF VEHICLE WEIGHT
AND SAFETY (NHTSA 1991).
12 Insurance Institute for Highway Safety, Small-Car Deaths, Injuries Worst;
Models Vary Greatly, Status Report: Highway Loss Reduction, Dec. 30, 1982, at
1; Where is Safety, supra note 5; Insurance Institute for Highway Safety,
Comparison Shows Downsizing Plays a Dramatic Role in Occupant Death Rates,
Status Report: Highway Loss Reduction, Mar. 16, 1991, at 4.
13 NHTSA, supra note 2.
14 Testimony from J. D. Khazzoom on Reauthorization of the NHTSA before Senate
Consumer Subcomm., Comm. Commerce, Science & Transportation, (Apr. 11,
1991, transcript) OTA, supra note 4; GENERAL ACCOUNTING OFFICE (GAO), HIGHWAY
SAFETY: HAVE AUTOMOBILE WEIGHT REDUCTIONS INCREASED HIGHWAY FATALITIES?
(1991).
15 CENTER FOR AUTO SAFETY, THE SAFE ROAD TO FUEL ECONOMY (1991).
16 Partyka, supra note 11; Partyka & Boehly, supra note 11; Kahane, supra
note 11; HIGHWAY LOSS DATA INSTITUTE, 1988 MODELS: THREE YEAR RESULTS,
INSURANCE COLLISION REPORT (including appendix RO-90-2, 1990); Where is Safety,
supra note 5; Comparison Shows, supra note 12; Klein et al., supra note 11; G.
ERNST, E. BRUHNING, K. GLAESER & M. SCHMID, COMPATIBILITY PROBLEMS OF SMALL
AND LARGE CARS IN HEAD-ON COLLISIONS, (91-S1-0-12, Fed. Hghwy. Resh. Inst.,
Germany, 1991); Hackney, The Effects of FMVSS N. 208 and NCAP on Safety as
Determined from Crash Test Results, 13th Internat'l Tech. Conf. Experimental
Safety Vehicles, Paris, France, November 4-7, 1991; Evans, Mass Ratio and
Relative Driver Fatality Risk in Two-Vehicle Crashes, (GM Resh. Labs. GMR-7419,
1991); Evans & Frick, Car Size or Car Mass -- Which Has Greater Influence
on Fatality Risk? (GM Resh. Labs., GMR-7462, 1991); Evans & Frick, Driver
Fatality Risk in Two-Car Crashes -- Dependence on Masses of Driven and Striking
Car, 13th Internat'l Tech. Conf. Experimental Safety Vehicles 1991; Robertson,
How to Save Fuel and Reduce Injuries in Automobiles, 31 J. TRAUMA 107 (1991);
NATIONAL ACADEMY OF SCIENCES (NAS), AUTOMOTIVE FUEL ECONOMY: HOW FAR SHOULD WE
GO? (1992).
17 Crandall & Graham, supra note 10.
18 R. CRANDALL, H. GRUENSPECHT, T. KEELER, & L LAVE, REGULATING THE
AUTOMOBILE (1986); Crandall & Graham, New Fuel Economy Standards? The
American Enterprise, Mar./Apr. 1991, at 68; R. LEONE & T. PARKINSON,
CONSERVING ENERGY: IS THERE A BETTER WAY? (1990); CHARLES RIVER ASSOCIATES,
POLICY ALTERNATIVES FOR REDUCING PETROLEUM USE AND GREENHOUSE GAS EMISSIONS,
FINAL REPORT (1991).
19 Greene, CAFE or Price? An Analysis of the Effects of Federal Fuel Economy
Regulations and Gasoline Price on New Car MPG, 1978-89, 11 ENERGY J. 37 (1990);
Leone & Parkinson, supra note 18; OTA, supra note 4.
20 S. 279, 102d Cong., 1st Sess., introduced Jan. 29, 1991.
21 H.R. 446, 102d Cong., 1st Sess., introduced Jan. 4, 1991.
22 OTA, supra note 4.
23 Id.
24 Crandall & Graham, supra note 10.
25 GAO, supra note 14.
26 NHTSA, supra note 2.
27 D. Buist, Ford Motor Company, Testimony before NHTSA, Sept. 14, 1988.
28 NHTSA, supra note 2.
29 Id.
30 OTA, supra note 4.
31 Partyka & Boehly, supra note 11.
32 Hackney, supra note 16.
33 The classes between "mini" and "largest" are, in increasing size:
"subcompact," compact," "intermediate" and "full." This data is from Partyka
& Boehly, supra note 11.
34 J. HEDLUND, THE EFFECTS OF VEHICLE SIZE ON OCCUPANT FATALITIES (NHTSA
1982).
35 NHTSA, FATAL ACCIDENT REPORTING SYSTEM (1990).
36 Campbell & Reinfurt, supra note 6; Grime & Hutchinson, The Influence
of Vehicle Weight on the Risk of Injury to Drivers, PROCEEDINGS OF THE NINTH
TECHNICAL CONFERENCE ON EXPERIMENTAL SAFETY VEHICLES (NHTSA 1982); Evans, Mass
Ratio, supra note 16.
37 Dreyer, Richter & Zobel, Handling, Braking, and Crash Compatibility
Aspects of Small, Front-Wheel Drive Vehicles, (Soc'y Auto. Eng. Tech. Paper
Ser. 810792, 1981); Richter & Zobel, Aspects of the Passive Safety of Motor
Vehicles, Presented at 9th Intern'l Tech. Conf. Experimental Safety Vehicles,
1982.; GAO, supra note 14; OTA, supra note 4.
38 Evans & Frick, Driver Fatality, supra note 16.
39 NAS, supra note 16.
40 Klein et al., supra note 11.
41 Id.
42 Mela, How Safe Can We Be in Small Cars? PROCEEDINGS THIRD INTERN'L CONF.
AUTO. SAFETY (1974); Mela, A Statistical Relation Between Car Weight and
Injuries. Technical Note (NHTSA 1975); H. JOKSCH, ANALYSIS OF THE FUTURE
EFFECTS OF FUEL SHORTAGE AND INCREASED SMALL CAR USAGE UPON TRAFFIC DEATHS AND
INJURIES (NHTSA 1976).
43 O'Neill, Joksch & Haddon, Relationships Between Car Size, Car Weight,
and Crash Injuries in Car-to-Car Crashes, PROCEEDINGS THIRD INTERN'L CONF.
AUTO. SAFETY (1974).
44 Evans & Frick, Car Size, supra note 16.
45 Joksch, supra note 42; O'Neill et al., supra note 43; O'Neill et al., supra
note 6.
46 Joksch, Effect of Small Cars on Traffic Safety Projections (Soc'y Auto. Eng.
Tech. Paper Ser. 840878, 1984).
47 Evans & Frick, Car Size, supra note 16.
48 O'Day, Golomb & Cooley, A Statistical Description of Large and Small Car
Involvement in Accidents, 3 HIT LAB RPTS. 1 (1973); Campbell & Reinfurt,
supra note 6; Zaremba, Injuries to Unrestrained Occupants in Small Car-Small
Car and Large Car-Large Car Head-On Collisions, 12 ACCIDENT ANAL. & PREV.
11 (1980); Wasielewski & Evans, Do Drivers of Small Cars Take Less Risk in
Everyday Driving? 5 RISK ANAL. 25 (1985); Evans, supra note 16; Evans &
Frick, Driver Fatality, supra note 16; Ernst et al., supra note 16.
49 Evans, Driver Fatalities Versus Car Mass Using a New Exposure Approach, 16
ACCIDENT ANAL. & PREV. 19 (1984); Evans, Accident Involvement Rate and Car
Size, 16 ACCIDENT ANAL. & PREV. 387 (1984); Evans & Frick, Car Size,
supra note 16; Evans & Frick, Driver Fatality, supra note 16.
50 O'Neill et al., supra note 43; Sparrow, Accident Involvement and Injury
Rates for Small Cars in Japan, 17 ACCIDENT ANAL. & PREV. 409 (1985); GAO,
supra note 14.
51 OTA, supra note 4.
52 NAS, supra note 16.
53 Evans & Frick, Car Size, supra note 16.
54 Campbell & Reinfurt supra note 6; Stewart & Stutts, A Categorical
Analysis of the Relationship Between Vehicle Weight and Driver Injury in
Automobile Crashes, (Hghwy. Safety Resh Ctr., U. N. Car., 1978); Joksch &
Thoren, Car Size and Occupant Fatality Risk, Adjusted for Differences in
Drivers and Driving Conditions (Report to AAA Fndn. Traffic Safety 1984).
55 O'Day et al., supra note 48; Joksch, supra note 42; O'Neill et al., supra
note 6; H. JOKSCH, LIGHT-WEIGHT CAR SAFETY ANALYSIS, PHASE II, PART II:
OCCUPANT FATALITY AND INJURY RISK IN RELATION TO CAR WEIGHT (Ctr. Environ.
& Man 1983); Evans, Car Mass and Likelihood of Occupant Fatality (Soc'y
Auto. Eng. Tech. Paper Ser. 820807, 1982); Evans, Driver Fatalities, supra note
49; Evans, Accident Involvement, supra note 49.
56 Partyka & Boehly, note 11.
57 Klein et al., supra note 11.
58 Ivey, Smaller Cars and Highway Safety, Texas Transportation Researcher,
April 1981, at 5.
59 OTA, supra note 4.
60 Evans & Frick, Car Size, supra note 16; Evans & Frick, Driver
Fatality, supra note 16.
61 See supra note 35. This data is also from Partyka & Boehly, supra note
11.
62 Ivey, supra note 58; Richter & Zobel, supra note 37.
63 Sparrow & Whitford, The Coming Mini/Macro Car Crisis: Do We Need a New
Definition? 18A TRANS. RESH. 289 (1984); Sparrow, supra note 50.
64 Wolfe & O'Day, A Study of Vehicle Factors Related to Type and Severity
of Pedestrian Injury, (Hghwy Safety Resh. Inst., U. Mich. 1982); Evans, Driver
Fatalities, supra note 49; Evans, Accident Involvement, supra note 49; Evans,
Driver Behavior Revealed in Relations Involving Car Mass in HUMAN BEHAVIOR AND
TRAFFIC SAFETY 337 (L. Evans & R. Schwing eds.1985).
65 How Safe Are Small Cars? Consumer Reports, Apr. 1976, at 188.
66 Dutt, Reinfurt & Stutts, Accident Involvement and Crash Injury Rates: An
Investigation by Make, Model and Year of Car, 9 ACCIDENT ANAL. & PREV. 275
(1977).
67 Evans, Driver Fatalities, supra note 49; Evans, Accident Involvement, supra
note 49.
68 Williams & Carsten, Driver Age and Crash Involvement, 79 AM. J. PUB.
HEALTH 326 (1989).
69 D. REINFURT & B. CAMPBELL, MILEAGE CRASH RATES FOR CERTAIN CAR MAKE AND
MODEL YEAR COMBINATIONS: A PRELIMINARY STUDY (Hghwy. Safety Resh Ctr., U. N.
Car. 1974).
70 Evans, Driver Fatalities, supra note 49; Evans, Accident Involvement, supra
note 49; Joksch, Small Car Accident Involvement Study (Draft Final Report to
NHTSA October 30, 1985); Evans, Involvement Rate in Two-Car Crashes Versus
Driver Age and Car Mass of Each Involved Car, 17 ACCIDENT ANAL. & PREV. 155
(1985); Evans, Driver Age, Car Mass and Accident Exposure -- A Synthesis of
Available Data, 17 ACCIDENT ANAL. & PREV. 439 (1985); Sparrow, supra note
50.
71 Consumer Reports supra note 65; HIGHWAY LOSS DATA INSTITUTE, THE EFFECTS OF
CAR SIZE ON CRASH LOSSES, 1977-1980 MODELS (1981); Small-Car Deaths, supra note
12; HIGHWAY LOSS DATA INSTITUTE, 1988 MODELS: THREE YEAR RESULTS, INSURANCE
COLLISION REPORT (including appendix 1990).
72 Robertson, supra note 16.
73 Kahane, supra note 11.
74 Garrett, A Study of Rollover in Rural U.S. Automobile Accidents (Soc'y Auto.
Eng. Tech. Paper Ser. 680772, 1968); L. GRIFFIN, PROBABILITY OF OVERTURN IN
SINGLE VEHICLE ACCIDENTS AS A FUNCTION OF ROAD TYPE AND PASSENGER CAR CURB
WEIGHT (Tex. Trans. Inst. 1981); Partyka, supra note 11.
75 Kahane, supra note 11.
76 Joksch, supra note 55.
77 Difiglio, Duleep & Greene, Cost Effectiveness of Future Fuel Economy
Improvements, 11 ENERGY J. 65 (1990).
78 S. Plotkin, Testimony on Legislative Proposals to Increase Automotive Fuel
Economy and Promote Alternative Transportation Fuels, House Subcomm. Energy and
Power, Comm. Energy & Commerce (transcript, Apr. 17, 1991); OTA, supra note
4.
79 NAS, supra note 16.
80 2 FUEL ECONOMY 1 (1990).
81 Letter from D. Kulp to author, February 6, 1992.
82 NHTSA, NATIONAL ACCIDENT SAMPLING SYSTEM 1986 (1988).
83 NHTSA, supra note 34.
84 TRANSPORTATION RESEARCH BOARD, SAFETY RESEARCH FOR A CHANGING HIGHWAY
ENVIRONMENT, SPECIAL REPORT 229 (1991)
85 Bunch, Smaller Cars and Safety: The Effect of Downsizing on Crash Fatalities
in 1995, The HSRI Research Review, Nov./Dec. 1978, at 1; Joksch, supra note 46;
Lave, Conflicting Objectives in Regulating the Automobile, 212 SCIENCE 893
(1981).
86 Evans, Car Mass, supra note 55; Evans, Driver Behavior, supra note 63.
87 Wasielewski and Evans, supra note 48.
88 O'Neill, Williams, & Karpf, Passenger Car Size and Driver Belt Use, 73
AM. J. PUB. HEALTH 588 (1983).
89 Crandall & Graham, Automobile Safety Regulation and Offsetting Driver
Behavior: Some New Empirical Estimates, 74 AM. ECONOMIC REV. 328 (1984);
Crandall & Graham, supra note 10.
90 Khazzoom, supra note 14.
91 Comparison Shows, supra note 12.
92 Najjar, The Effects of Smaller Cars on Passenger Car Occupant Injuries in 3
COLL. TECH. STUDIES: ACCIDENT DATA ANAL. RESULTS & METHODOL. (NHTSA
1983).
93 Ctr. Auto Safety, supra note 15.
94 Trans. Resh. Bd., supra note 84.
95 J. CLAYBROOK, RETREAT FROM SAFETY (1984); Graham, Technology, Behavior, and
Safety: An Empirical Study of Automobile Occupant-Protection Regulation, 17
POL'Y SCI. 141 (1984).
96 J. GRAHAM, AUTO SAFETY: ASSESSING AMERICA'S PERFORMANCE (1989).
97 Joksch, supra note 46.
98 NAS, supra note 16.
99 TRANSPORTATION RESEARCH BOARD, 55: DECADE OF EXPERIENCE (1984).
100 Leigh & Frank, Tax Gasoline to Save Lives, 316 N. ENG. J. MED. 54
(1987) Leigh & Wilkinson, The Effect of Gasoline Taxes on Highway
Fatalities, 10 J. POL'Y ANAL. & MGMT. 474 (1991).
101 Leigh & Wilkinson, supra note 100.
Estimated Impact of 100 Pound Reduction in Vehicle Size/Weight on Occupant
Injury Risk by Crash Type [Omitted]
Steady-State Estimates of the Impacts of the Bryan Bill on Fatal and Nonfatal
Injuries [Omitted]
Risk Articles Index