SPECIAL ISSUE BUMPERS Insurance Institute for Highway Safety

Document Sample
SPECIAL ISSUE BUMPERS Insurance Institute for Highway Safety Powered By Docstoc
					DO NOT

              When the National Highway Traffic Safety Ad-
            ministration (NHTSA) gutted the 5 mph n<Hiamage
            bumper standard, reducing it to 2.5 mph, it ap-
            peared to be a fixed game from square one. The
            agency justified its decision to adopt 2.5 mph re-
            quirements with a cost·benefit analysis that ig-        ADVANCE
            nored the available real-world evidence. NHTSA
            preferred instead guesswork.                            TO BANK



                                                         SAVE $18   PARKING
2-1IHS Status Report,   ~'O{.   27, No.8, June 20, 1992

1111I!'                         .               -                                         Last month marked the 10th anniver-
                                                                                      sary 01 the bumper standard rollback. We
                                                              NHTSA's                 can now review the lo-year predictions
                                                                                      NIffSA made in 1982 to justify its action
                                                             Erroneous                and compare these with what actually
                                                                                      has happened. And what has happened
                                                             Predictions              is that NHTSA was wrong. What the
                                                                                      agency said would happen to save con-
                                                             NHTSA's                  sumers money by weakening bumper re-
                                                          "Most Probable"   NHTSA's   quirements hasn't transpired. For exam-
                                                            Predlclton       Range    ple, gasoline would be saved by making
                                                                                      bumpers lighter as well as weaker, NHTSA
                                                                                      said in 1982. But weaker bumpers aren't
                                                                                      as light as predicted and, because gaso-
                                                                                      line doesn't cost $3.47 per gallon as
                                                                                      NHTSA predicted, not as much money is
                                                                                      being saved in fuel costs.
                                                                                          In general terms, NHTSA rigged the
                                                                                      bumper rollback game by grossly under·
                                                                                      estimating the benefits 01 5 mph
                                                                                      bumpers and overestimating the savings
                                                                                      that would result from 2.5 mph bumpers.
                                                                                      This special Status Report looks back at
                                                                                      some of NHTSA's 1982 predictions and
                                                                                      compares them with what actually has
                                                                                      happened during the last decade.
                                                                                          uroneoUl Prediction' I: NHTSA
                                                                                      concluded in 1982 that 2.5 mph bumpers
                                                                                      would be 63-67 percent as effective as 5
                                                                                      mph bumpers in preventing damage over
                                                                                      the full range of impact types and severi·
                                                                                      ties in real, on·the-road crashes. This pre-
                                                                                      diction was pivotal because it served as
                                                                                      the basis for comparing the benefits of
                                                                                      various bumper requirements.
                                                                                          But a Volkswagen example presented
                                                                                      to NHTSA before the rollback illustrates
                                                                                      how the agency overstated the value 01
                                                                                      2.5 mph bumpers relative to 5 mph ones.
                                                                                      A 1981 Rabbit pickup truck eqUipped
                                                                                      with the equivalent of a 2.5 mph front
                                                                                      bumper sustained $364 damage (1981
                                                                                      dollars) in two Institute crash tests at 5
                                                                                      mph. A 1981 Rabbit sedan with the same
                                                                                      front-end design, except lor its 5 mph
                                                                                      bumper, sustained only $21 damage in

  Bottom line cost saving of                                                          the same two flat- and angle-barrier tests.
  2.5 mph bumpers, according                                                              NHTSA paid no attention to this exam-
  to NHTSA in 1982                                                                    ple. Nor did the agency conduct low·
                                                                                      speed crash tests of its own to determine
                                                                                      the relative effectiveness 01 5 and 2.5
                                                                                           IlHS Status Report, Vol, 27, No.8, June 20, 1992-3

mph bumpers. Instead, NHTSA derived
its values from arbitrary curves drawn
through three sets of data points repre-
senting three different bumper effective-
ness levels at three different impact
speeds. Such curves were nothing better
than guesses.
                                              " ..          ""*" COlts rtIIect
    Akey assumption - or, more correct-                 , . ptl1tlll'ld IIbor priceI                     •
ly, a key guess - leading NHTSA to its         '"''                                                                                       $I"

relative effectiveness estimates was that
a bumper loses all of its damage protec·       ... .•...-,. .• >.
                                                                                                              • •
                                                                                                                  •• •
                                                                                                              •.." I •
                                                                                                              .,.. . .
                                                 .. .. .-
tion properties at twice its no-damage                       • •
barrier impact design speed. That is,
NHTSA assumed a 5 mph bumper ceases                        •
                                                            • • • •• •
                                                                                       •                        .   .",        •
to provide protection from damage at 10
mph. But Institute crash tests demon-
strate that 5 mph no-<lamage bumpers re-
duce damage at speeds well above 10          AI/hough a correlation does exist between bumper weight and damage resistance, it's not a strong
mph. A1983 Plymouth Horizon with such        correlation. The amount of damage variability that's explained by bumper weight is slight. If au-
bumpers, for example, sustained about        tomakers choose to do so - a few have - they could equip their cars with bumpers that are both
$500 less damage in a 15 mph front-into-     lightweight to conseroe fuel and effective in preventing expensive damage in low-speed collisions.
barrier test than did an almost identical
 1983 Horizon with weaker bumpers                                                                             NHTSA claimed in 1982 that not
($3,317 versus $3,843 in 1983 dollars).                                                                  enough low·speed crashes occur to
                                                                                                         justify the weight, cos/, and complexi·
     NHTSA thus used effectiveness esti-                                                                 ty of 5 mph no-damage bumpers. In
 mates that would make 2,5 mph bumpers                                                                   fact. such low-speed crashes occur all
appear reasonably close to 5 mph ones                                                                    the time. Just how frequently was indi-
 in terms of damage resistance. The prob-                                                                cated in a /991InstilUte study of pas-
 lem is, they aren't close at all.                                                                       senger cars brought to 16 insurance
   Errooeous Predictloo *2: NHTSA                                                                        drive·in claims centers in four major
said 2.5 mph bumpers would incur $34-                                                                    metropolitan areas.
69 in additional repair and insurance                                                                         The suroey reveals that about 20
costs, compared with 5 mph bumpers,                                                                      percent of all claims for automobile
over the IG-year life of a car. Insurance                                                                damage involve parking lot collisions.
                                                                                                          Often one of the vehicles involved in
collision coverage results for General
                                                                                                          the colUsion is parked or standing
Motors models provide evidence from                                                                      stilL Such crashes happen IJy the thou·
on-the-road driving that weaker bumpers                                                                  sands every day on congested streets
cause much higher repair costs than                                                                       and in parking lots, and they're the
NHTSA predicted.                                                                                          kinds of impacts in which the differ·
    In 1984, three Buick models were                                                                      ence between 2,5 and 5" mph bumpers
eqUipped with weaker bumpers than their                                                                   can mean the difference between lots
otherwise nearly identical 1983 counter-                                                                  of costly damage to repair and no
parts. Similar models from the OIdsmtr                                                                    damage at al/.
bile division retained 5 mph bumpers in                                                                       Safety-related parts are often in-
both the 1983 and '84 model years. Insur-                                                                 volved. Damage to front bumpers was
                                                                                                          accompanied by damage to front lights
ance collision coverage losses, as mea-
                                                                                                          for 72 percent of {he cars brought to
sured by average loss payments per in-                                                                    the drive·in claims centers. "We need
sured vehicle year, averaged 21·35 per-                                                                   5 mph bumpers not only to save mono
cent higher for the 1984 Buicks compared                                                                  ey, • explains Institute President Brian
with '83 Buicks. On the other hand, insur-                                                                O'Neill, "but also to protect the snfety·
ance losses were only 4-8 percent higher                                                                  related parts of our cars like lights.•
4-IlHS SlaWS Rep"", Vol. 27, No.8, June 20, 1992

for the 1984 Oldsmobiles compared with             would weigh 15-33 pounds less than 5
their '83 counterparts. These differences
translate into an average 5135 increase in
collision coverage losses over the 10-
                                                   mph bumpers, but this weight saving
                                                   hasn't materialized. Weaker bumpers
                                                   have saved far less weight and, hence,
year life of acar (1983 dollars).                  less fuel than NHTSA predicted. For ex-     They're Even Worse
   The [ass figure of 5135 doesn't count
additional costs NHTSA did include in its
                                                   ample, the combined front and rear
                                                   bumper weights for 10 cars, all 1991·92
                                                                                               On Cars Sold Outside
$34-69 estimate - costs like increased             models, changed from 21 pounds less to      The U.S. and Canada
insurance deductibles consumers pay                26 pounds more than the bumper
first and added losses under property              weights of 198Q..83 counterpart models,         Bad as they are on cars sold in the
damage liability coverage. Total cost in-          which had 5 mph bumpers. The average        United States and Canada, bumpers are
creases for weaker bumpers aTe thus                was no weight saved at all for the newer    even flimsier on cars sold elsewhere.
substantially higher than estimated. The           models with the weaker bumpers.             Australia's new low-speed crash test
Institute submitted this information in               NHTSA didn't have this information       program provides some examples.
1989. Even though it invalidated NHTSA's           when it rolled back the federal bumper          ~Every year Australian motorists are
1982 prediction, the agency ignored it.            standard from 5to 2.5 mph in 1982. Then     being saddled with an accident repair
   Errooeoua Prediction 13: NHTSA                  again, the agency did have enough infor-    bill of hundreds of millions of dollars
estimated in 1982 that 2.5 mph bumpers             mation to know its        (cont'd on p.6)   for damage that need not occur," says
                                                                                               Australian Associated Motor Insurers
                                                                                               Limited (AAMI) after conducting auto-
                                                                                               mobile bumper crash tests in coopera-
                                                                                               tion with Monash University.
                                                                                                   "The results were appalling," says
                                                                                               Noel Murray, professor of structural en-
                                                                                               gineering at Monash who directed the
                                                                                               AAMJ tests. "None of the automobiles
                                                                                               we crashed withstood the tests without
                                                                                               substantial damage." The average re-
                                                                                               pair estimate worked out to $1,606
                                                                                               (1992 U.s. dollars) lor damage in each
                                                                                               of the 5 mph tests.
                                                                                                   Patterned after the Institute's lront-
                                                                                               into-flat-barrier and rear-into-pole tests,
                                                                                               AAMI's recent series of 5 mph tests in-
                                                                                               volved live 1991 and '92 model cars.
                                                                                               Three of the models - Ford Laser, Hold-
                                                                                               en Barina, and Toyota Corolla - have
                                                                                               counterparts that are sold in the United
                                                                                               States and have been tested by the Insti-
                                                                                               tute. The results of these crash tests
                                                                                               show repair costs that are higher in
                                                                                               every case and usually more than dou-
                                                                                               ble for the cars sold in Australia, com-
                                                                                               pared with those sold to buyers in the
                                                                                               United States.
                                                                                                   One reason bumpers are generally
                                                                                               worse on cars sold in countries other
                                                                                               than the United States and Canada is
                                                                                               that, despite the weak requirements of
                                                                                               the 2.5 mph federal bumper standard,
                                                                                         /IllS Status Report, Val. 27, No. ~ June 20, /992-5

                                                  IllImIge IlepeIr CoIls, 511PH Cr8sh T .....
                                                  AuIlnIIIIn CIrs and Equivlllenl U.S.1IocIeIs

some automakers elect to equip cars             ~The    tests are an incentive to im-
sold in this country with stronger, more     prove bumpers," explains Institute Pres-
damage-resistant bumpers than the            ident Brian O'Neill. "Insurance rates are
standard requires. Honda, for example,       another incentive. Collision coverage
does so. After the 1987 Accord and '88       premiums are based on insurance losses
Civic performed poorly in the Institute's    and, because so many collisions occur
5 mph crash tests, Honda improved the        at low speeds, a car with weak bumpers
bumpers on these cars. The result was        is likely to have much higher insurance
that the 1990 Civic OX was the best car      losses under collision coverage than a
the Institute tested that year.              car with stronger bumpers.·
   The following year, the '91 Accord            Cars sold in Australia, the United
was the best car tested. The bumpers on      States, and elsewhere come eqUipped
the '91 Accord performed better even         with generally flimsy bumpers for at
than those on the 1982 Accord, which         least one common reason - the ~very
met federal 5 mph no-damage require-         profitable and growing captive market in
ments. However, prospective car buyers       the supply of replacement parts. Manu-
cannot tell by looking whether a vehi-       facturers will always seek to preserve
cle's bumpers are as weak as the federal     and nurture this market," MMI points
standard allows or, like Iate-model HOIl-    out. The organization is asking the gov-
das, somewhat stronger. This is one of       ernment to enact a rule requiring more
the reasons lor the Institute's annual se-   damage-resistant bumpers on all cars
ries 01 crash tests at 5mph.                 sold in Australia.
6-IIHSSlatus Rep,,", Val. 27, No.8, June 20, /992

     Damage Repair Costs, 5MPH Crash Tests. 1980-83 Cars   Game Rigged for 2.5
                                                           (cont'd from p.4)   projected weight sav-
      With 5 MPH Bumpers Compared with 1991-92 Models      ing was way too high. A number of auto
                                                           manufacturers including Chrysler, Volks-
                                                           wagen, and Mitsubishi told NHTSA they
                                                           couldn't achieve the minimum saving 01
                                                           15 pounds the agency was predicting. If
                                                           this information had been lactored into
                                                           NHTSA's cost-benefit analysis - it wasn',
                                                           - the evidence would have shifted signif-
                                                           icantly toward keeping the 5 mph federal
                                                           nCHlamage bumper standard.
                                                               Erroneous Prediction t4: NHTSA
                                                           estimated in 1982 that consumers would
                                                           save $18-35 off the price of a car because
                                                           01 2.5 mph bumpers and another 56-20
                                                           because of weight saving elsewhere on
                                                           the vehicle. made possible by the weak-
                                                           er bumpers. However, at least partly be-
                                                           cause lolal weight saving hasn'l come
                                                           close 10 what NHTSA predicted, cost
                                                           saving hasn't matched the agency's pre-
                                                           diction either.
                                                               Car sticker prices were not reduced
                                                           dUring the 1983 model year when bump-
                                                           ers on several models were weakened.
                                                           Plus, comparison of current bumper re-
                                                           placement prices for a number of 198Q.83
                                                           models and 1991-92 models shows some
                                                           prices higher and others lower - hardly
                                                           a trend to support NHTSA's prediction
                                                           that weaker bumpers would cost less.
                                                               Er,oDeou PredlctlOD '5: NHTSA
                                                           estimated in 1982 that a gallon of gas~
                                                           line would be saved over Ihe I().year life
                                                           of an automobile for every pound of
                                                           weight saved. This would mean, the
                                                           agency said, consumers would save $28-
                                                           70, depending on actual weight saved,
                                                           over a car's life because of lighter 2.5
                                                           mph bumpers. However, because cars
                                                           manufactured since the 1983 model year
                                                           aren't nearly as much lighter as NHTSA
                                                           predicted, this benefit of 2.5 mph bump-
                                                           ers hasn't materialized.
                                                               Even if cars had been made as much
                                                           lighter as NIITSA predicted, the 128-70 lu·
                                                           el saving wouldn't have been realized be-
                                                           cause other assumptions on which
                                                           NHTSA based its fuel-saving prediction
                                                                                IIHS Status Reporl. Vol. 27. No.8, June 20. 1992-7

were far from the mark. The agency as-
sumed, for example, that gasoline prices
would reach $3.47 per gallon by 1991,
but gas is substantially cheaper. It cost
an average of $1.23 per gallon at the end
of 1991. Incorrect estimates of weight
and associated gasoline savings were
the biggest factors in skewing NHTSA's
1982 cost-benefit analysis in favor of 2.5
mph bumpers.
    Erroneous Prediction '6: NHTSA
estimated the consumer value of the lost
time and inconvenience associated with a
typical fender-bender crash at $26-50.
                                                         GAME IS
Taking into account how many crashes
NHTSA predicted would occur and its rel-
ative bumper effectiveness estimates, the
agency said in 1982 that 2.5 mph bumpers         Taken tngether, the erroneous predic·
would add $6 for lost time and inconve-       tions described in this Status Report un-
nience over the lo-year life of a car. But
these numbers ignored a 1981 survey of
                                              dercut NHTSNs 1982 decision to roll back
1,000 consumers conducted for the Insti-      bumper protection requirements to 2.5
tute by Opinion Research Corporation.         mph. The agency was wrong abnut the rei·
    When asked to tell how much money         ative effectiveness of 5 and 2.5 mph
it would be worth to avoid the inconve-       bumpers. it was wrong abnut the weight
nience associated with minor collisions       that wouid be saved by weakening
- waiting for police, missing work, being     bumpers. It was wrong about the fuel that
late, finding other transportation, getting   would be saved. And it was wrong about
repairs completed, etc. - 83 percent          the amount repair costs would increase.
said it would be worth $100 or more, not
                                                 "The agency knew all along its
the 126·50 NHTSA estimated in 1982.
Fifty-eighl percent said $200 or more, so     predictions were nothing more than
the agency obViously undervalued the          guesses," says Institute President Brian
benefit of strong, 5 mph bumpers in           O'Neill. "But the decision to weaken
avoiding such costs.                          bumper requirements wasn't based on
    A1992 survey of 400 prospective car       the evidence anyway. It was a political de-
buyers, conducted for the Institute in        cision. And it's one that has cost con-
March by ICR Survey Research Group,           sumers huge amounts of money for a
shows consumers haven't changed their         decade," O'Neill adds.
minds. They still want protective
bumpers. When asked whether they
plan, when buying a car, to consider the
capacity of bumpers to prevent damage
in low-speed collisions, 67 percent said                                                                      FREE
yes. Among women, three out of four                                          DO NOT

said they plan to consider how well the
bumpers prevent damage. When asked
whether they prefer stylish or protective
bumpers, if given a choice, 83 percent of
all respondents chose protection. Only                                      SAVE $18                     PARKING
15 percent opted for style.
                                                                     Vol. TI, No, 8, Jon. 20, 1992
Special Issue
This special issue of Status Report focuses on
NHTSA's 1982 rollback of the 5 mph bumper
standard. Other special issues have focused on
the following subjects:
1992 Low·Speed Crash Tests                        27:1, 1992
High Speed and SaJety                             26:8,1991
Death Rates by Car Make, Series                   26:4,1991
1991 Low·Speed Crash Tests                        26:2,1991
Fuel Economy and Safety                           25:8,1990
Antilock Brakes for Trucks                        25:5,1990
Speeds on Rural Interstates                       25:2,1990
Death Rates by Car Series                        24:11,1989
Designing Safer Vehicles                          24:8,1989
Truck Crash Congestion                           23:12,1988
Making Traffic Laws Work                          23:6,1988
Seat Belt and Helmet Laws                        22:13. 1987
NlffSA Safety Rules                               22~, 1987
Vehicle Size and Death Rates                      22:2, 1987
US. SaJety Acts                                  21:11,1986
Seat Bell Use Laws                               21"2, 1985
School Btls Safety                                ZO-S, 1985

1005 North G1tbe Road
ArllngtOll, VA 22201
(703) 247·1500 FAX (763) 247-1678

o.rtctOl' oIl'ublicatwns'Editor. Anne Flemlllg
~le Editor. Rea Kerr Howarth
V. rlttr: )\W ~uhJlann
Editorial AssIsllnt Carlene Hughes
ClIcuWioIL ShdIy Wontgcmery
DesIgn, ProdIxtJon: SheiIa.bcbon

The klsw'lIlJCt klslkute lor Highway Saldy 111/1 indqlm.
dent. lIOllj)rllfil, $dtrdk and ~ orpniulion. It
II dednled to mkJCq tilt kJs5C$ - deJlhs. Wljurlts, Mel
propert) clllnIte - rtSU!tlllg frolll mshes on the nalion's
Iu&bWlys. The Institute Is supported by tilt Amtrlcin D.
SlnIICe HigIno"l}' SUely AssociIlion. the Aaaerbn In5urM
tbgbwly s.!dy AIlmce, the National Assocl&tlon 01 ....
peodmt Insurm Safety As5ociitlon. and I number 0I1ndl-
VtduIIlnsurlllCe COIIlpanies.

Conttnts may be republlsht'd whole, 01' in pm. wltb Ittn.
llutJon.l! you are not now ~ng SlallG Repott but would
like to. cootllCt the CommunlClltions Department

ISSN 0018·986X

Shared By: