GAO-05-370, Vehicle Safety: Opportunities Exist to Enhance NHTSA's New Car Assessment Program


This is the accessible text file for GAO report number GAO-05-370 
entitled 'Vehicle Safety: Opportunities Exist to Enhance NHTSA's New 
Car Assessment Program' which was released on April 29, 2005.

This text file was formatted by the U.S. Government Accountability 
Office (GAO) to be accessible to users with visual impairments, as part 
of a longer term project to improve GAO products' accessibility. Every 
attempt has been made to maintain the structural and data integrity of 
the original printed product. Accessibility features, such as text 
descriptions of tables, consecutively numbered footnotes placed at the 
end of the file, and the text of agency comment letters, are provided 
but may not exactly duplicate the presentation or format of the printed 
version. The portable document format (PDF) file is an exact electronic 
replica of the printed version. We welcome your feedback. Please E-mail 
your comments regarding the contents or accessibility features of this 
document to Webmaster@gao.gov.

This is a work of the U.S. government and is not subject to copyright 
protection in the United States. It may be reproduced and distributed 
in its entirety without further permission from GAO. Because this work 
may contain copyrighted images or other material, permission from the 
copyright holder may be necessary if you wish to reproduce this 
material separately.

Report to Congressional Committees:

April 2005:

Vehicle Safety:

Opportunities Exist to Enhance NHTSA's New Car Assessment Program:

[Hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-05-370]:

GAO Highlights:

Highlights of GAO-05-370, a report to congressional committees.

Why GAO Did This Study:

In 2003, 42,643 people were killed and more than 2.8 million people 
were injured in motor vehicle crashes. Efforts to reduce fatalities on 
the nation’s roadways include the National Highway Transportation 
Safety Administration’s (NHTSA) New Car Assessment Program. Under this 
program, NHTSA conducts vehicle crash and rollover tests to encourage 
manufacturers to make safety improvements to new vehicles and provide 
the public with information on the relative safety of vehicles. GAO 
examined (1) how NHTSA’s New Car Assessment Program crash tests 
vehicles, rates their safety, and reports the results to the public; 
(2) how NHTSA’s program compares to other programs that crash test 
vehicles and report results to the public; and (3) the impact of the 
program and opportunities to enhance its effectiveness.

What GAO Found:

NHTSA conducts three types of tests in the New Car Assessment 
Program—full frontal and angled side crash tests and a rollover test. 
Each year, NHTSA tests new vehicles that are expected to have high 
sales volume, have been redesigned with structural changes, or have 
improved safety equipment. Based on test results, vehicles receive 
ratings from one to five stars, with five stars being the best, to 
indicate the vehicles’ relative crashworthiness and which are less 
likely to roll over. NHTSA makes ratings available to the public on the 
Internet and through a brochure. Other publications, such as Consumer 
Reports, use NHTSA’s test results in their safety assessments.

Examples of NHTSA’s New Car Assessment Program Tests

[See PDF for Image]

[End of Figure] 

GAO identified four other programs—the Insurance Institute for Highway 
Safety’s program and the New Car Assessment Programs in Australia, 
Europe, and Japan—that crash test vehicles and report the results to 
the public. They share the goals of encouraging manufacturers to 
improve vehicle safety and providing safety information to consumers. 
These programs conduct different types of frontal and side crash tests, 
and some perform other tests, such as pedestrian tests, that are not 
conducted under the U.S. program. Only the U.S. program conducts a 
rollover test. The other programs measure test results differently and 
include more potential injuries to occupants in ratings. They also 
reported their test results differently, with all summarizing at least 
some of the scores or combining them into an overall crashworthiness 
rating to make comparisons easier. 

NHTSA’s New Car Assessment Program has been successful in encouraging 
manufacturers to make safer vehicles and providing information to 
consumers. However, the program is at a crossroads where it will need 
to change to maintain its relevance. The usefulness of the current 
tests has been eroded by the growing number of larger pickups, 
minivans, and sport utility vehicles in the vehicle fleet since the 
program began. In addition, NCAP scores have increased to the point 
where there is little difference in vehicle ratings. As a result, the 
program provides little incentive for manufacturers to further improve 
safety, and consumers can see few differences among new vehicles. 
Opportunities to enhance the program include developing approaches to 
better measure the interaction of large and small vehicles and occupant 
protection in rollovers, rating technologies that help prevent crashes, 
and using different injury measures to rate the crash results. NHTSA 
also has opportunities to enhance the presentation and timeliness of 
the information provided to consumers.

What GAO Recommends:

GAO recommends that NHTSA examine the direction of the New Car 
Assessment Program to ensure that it maintains its relevance in 
improving vehicle safety, including identifying tests that best address 
the fatalities occurring on the nation’s roads. GAO also recommends 
that NHTSA enhance the presentation and timeliness of the information 
provided to the public. NHTSA generally agreed with GAO’s findings.

[Hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-05-370].

To view the full product, including the scope and methodology, click on 
the link above. For more information, contact JayEtta Hecker at (202) 
512-2834 or [Hyperlink, heckerj@gao.gov].

[End of Section]

Contents:

Letter:

Results in Brief:

Background:

NCAP Crash Tests Vehicles, Rates Their Safety, and Reports the Results 
to the Public:

U.S. NCAP Differs from Other Crash Programs in Testing, Rating, 
Reporting, and Government Involvement:

NCAP Has Contributed to Making Vehicles Safer, but Changes are Needed 
to Maintain Its Relevance:

Conclusions:

Recommendations for Executive Action:

Agency Comments and Our Evaluation:

Appendixes:

Appendix I: Scope and Methodology:

Appendix II: Basis for NHTSA's Crashworthiness Ratings:

Appendix III: Basis for NHTSA's Rollover Rating:

Appendix IV: Insurance Institute for Highway Safety:

Appendix V: Australian New Car Assessment Program:

Appendix VI: European New Car Assessment Programme:

Appendix VII: Japan New Car Assessment Program:

Appendix VIII: Different Types of Tests Used by the Programs:

Appendix IX: GAO Contacts and Staff Acknowledgments:

GAO Contacts:

Acknowledgments:

Tables:

Table 1: List of Organizations Contacted:

Table 2: Abbreviated Injury Scale (AIS):

Table 3: Scoring Basis for Australia NCAP Frontal and Side Star Ratings:

Table 4: Australia NCAP Point System for the Pedestrian Test:

Table 5: Scoring Basis for Australia NCAP Pedestrian Rating:

Table 6: Euro NCAP Pedestrian Test Assessment Criteria:

Table 7: Euro NCAP Front and Side-Impact Star Rating System:

Table 8: Scoring Basis for Euro NCAP Pedestrian Ratings:

Table 9: Scoring Basis for Euro NCAP Child Protection Star Ratings:

Table 10: Japan NCAP Vehicle Types Used for Pedestrian Test:

Table 11: Description of Different Types of Tests Used by the Programs:

Figures:

Figure 1: Three Types of Tests--Frontal, Side, and Rollover--Conducted 
by NCAP:

Figure 2: Improvement of Average Star Ratings for Frontal and Side NCAP 
Tests:

Figure 3: Time Line of NCAP:

Figure 4: Full Frontal Crash Test Conducted under NCAP:

Figure 5: Angled Side Crash Test Conducted under NCAP:

Figure 6: Rollover Test Conducted under NCAP:

Figure 7: Frontal Star Rating and the Corresponding Chance of Serious 
Injury to the Head and Chest:

Figure 8: Side Star Rating and the Corresponding Chance of Serious 
Injury to the Chest:

Figure 9: Calculation of a Vehicle's Top-Heaviness:

Figure 10: Rollover Star Rating and the Corresponding Risk of Rollover:

Figure 11: Ratings for a 2004 Passenger Vehicle as It Appears on 
NHTSA's Web Site:

Figure 12: Details of Frontal, Side, and Rollover Star Ratings for a 
Passenger Vehicle as They Appear on NHTSA's Web Site:

Figure 13: Vehicle Safety Tests Conducted by Five Testing Programs:

Figure 14: Offset Frontal Crash Test:

Figure 15: Perpendicular Side Impact Crash Test:

Figure 16: Comparison of Barriers Used in the NCAPs' Side Test (left) 
and the Insurance Institute Side Test (right):

Figure 17: Side-Impact Crash Test with SUV-like Barrier:

Figure 18: Side Pole Crash Test:

Figure 19: Head Form into Hood for the Pedestrian Test:

Figure 20: Child Restraint Test:

Figure 21: Dummy Injury Diagrams of Driver and Passenger in Frontal 
Test and Driver in Side Test:

Figure 22: Improvement of Average Star Ratings for Frontal and Side 
NCAP Tests:

Figure 23: Frequency of Four-and Five-Star Ratings for Frontal and Side 
Crash Tests in 2004:

Figure 24: Example of Height Mismatch Between Vehicles of Different 
Type and Weight:

Figure 25: Example of the Damage Caused by an SUV Striking the Side of 
a Small Vehicle:

Figure 26: Example of Head Movement during the Insurance Institute Test 
with the SUV-like Barrier:

Figure 27: Example of Improved Matching of Frontal Vehicle Structures:

Figure 28: Example of Australia's NCAP Safety Rating Information:

Figure 29: Example of Japan's NCAP Safety Rating Information:

Figure 30: Example of NHTSA's Rollover Rating for a Pickup Truck with 
Bar Chart Showing How the Vehicle Performed within Its Class:

Figure 31: Examples of Euro NCAP Public Displays of Crash Vehicles:

Figure 32: A Crash Test Dummy Fitted with Load Sensors and 
Accelerometers:

Figure 33: Injury Curve for HIC:

Figure 34: Injury Curve for Chest G:

Figure 35: Probability Equation Used to Produce Star Ratings for 
Frontal Crashes:

Figure 36: Thoracic Trauma Index Curve:

Figure 37: Static Stability Factor of a Passenger Vehicle Compared with 
an SUV:

Figure 38: Programmable Steering Controller:

Figure 39: Diagram of the Dynamic Test, Showing Steering Parameters:

Figure 40: Sensors to Detect Wheel-lift:

Figure 41: NCAP Logistic Model Used to Determine Rollover Ratings:

Figure 42: Guidelines for Rating Occupant Compartment Intrusion 
Measured in Centimeters:

Figure 43: Insurance Institute Rating of a Mid-size Passenger Car:

Figure 44: Insurance Institute Crash Ratings Provided to Public in its 
Status Report Publication:

Figure 45: Example of Australian NCAP Rating of a 2004 Mid-size 
Passenger Car:

Figure 46: Australian NCAP July 2004 Crash Test Update Brochure:

Figure 47: European NCAP Rating of a 2004 Mid-size Passenger Car:

Figure 48: Dummy Injury Diagrams of Driver and Passenger in Frontal 
Test, and Driver in Side Test:

Figure 49: Test Vehicle Undergoing Brake Tests Under Dry and Wet 
Conditions:

Figure 50: Pedestrian Head Impact Test and Target Area:

Figure 51: Japan NCAP Rating of a 2003 Mid-sized Passenger Car:

Figure 52: Example of Japan NCAP Detailed Full Frontal Data Available 
for a 2003 Test:

Figure 53: Key to Japan NCAP's Door Openability Ratings:

Figure 54: Key to Japan NCAP's Rescueability Ratings:

Abbreviations:

AAAM: Association for the Advancement of Automotive Medicine:

ADAC: General German Automobile Association (Allegmeiner Deutscher 
Automobil-Club e V):

AIS: Abbreviated Injury Scale:

c.g.: center of gravity:

CFR: Code of Federal Regulations:

DOT: Department of Transportation:

ESC: Electronic Stability Control:

Euro NCAP: European New Car Assessment Programme:

HIC: Head Injury Criterion:

kg: kilogram:

km/h: kilometers per hour:

mm: millimeter:

mph: miles per hour:

NASVA: National Agency for Automotive Safety and Victims' Aid:

NBC: National Broadcasting Company:

NCAP: New Car Assessment Program:

NHTSA: National Highway Traffic Safety Administration:

SID: side-impact dummy:

SUV: sport utility vehicle:

SSF: Static Stability Factor:

TREAD Act: Transportation, Recall Enhancement, Accountability, and 
Documentation Act:

TSRE AB: Traffic Safety Research and Engineering AB:

TTI: Thoracic Trauma Index

Letter:

April 29, 2005:

The Honorable Christopher Bond:
Chairman:
The Honorable Patty Murray:
Ranking Member:
Subcommittee on Transportation, Treasury, the Judiciary, Housing and 
Urban Development, and Related Agencies:
Committee on Appropriations:
United States Senate:

The Honorable Joe Knollenberg:
Chairman:
The Honorable John W. Olver:
Ranking Member:
Subcommittee on Transportation, Treasury, and Housing and Urban 
Development, the Judiciary, District of Columbia, and Independent 
Agencies:
Committee on Appropriations:
House of Representatives:

Motor vehicle travel is the primary means of transportation in the 
United States. Yet for all its advantages, deaths and injuries 
resulting from motor vehicle crashes are the leading cause of death for 
all persons from 3 through 33 years old. In 2003, 42,643 people were 
killed and more than 2.8 million people were injured in motor vehicle 
crashes. Frontal crashes caused the largest portion of occupant deaths 
(about 41 percent), followed by rollovers and side impact crashes (30 
percent and 22 percent, respectively). In addition to the loss of life, 
motor vehicle crashes have a high economic cost, which the National 
Highway Traffic Safety Administration (NHTSA) estimated at over $230 
billion in 2000, the most recent year for which cost estimates were 
available.[Footnote 1]

Efforts to reduce fatality rates in motor vehicle crashes have resulted 
in some improvement. The fatality rate per 100 million miles of travel 
in 2003 was at a historic low of 1.48, down from 1.75 per 100 million 
miles of travel in 1993. The Department of Transportation (DOT) 
attributes this change to several factors, including increased safety 
belt use, reduction of alcohol-related deaths, and vehicle safety 
programs, including Federal Motor Vehicle Safety Standards and the New 
Car Assessment Program (NCAP), both run by NHTSA. Under NCAP, which 
began in 1978, NHTSA conducts frontal and side crash tests and rollover 
tests of new cars, light trucks, vans, and sport utility vehicles (SUV) 
and reports the results to the public. The specific goals of the 
program are to encourage market forces that prompt vehicle 
manufacturers to make safety improvements to new vehicles and provide 
the public with objective information on the relative safety 
performance of vehicles.

Federal Motor Vehicle Safety Standards are regulations that establish 
minimum performance levels that manufacturers must self-certify to in 
order to sell vehicles in the United States. Under the safety 
standards, NHTSA requires vehicles to pass a number of performance 
tests to ensure that the minimum safety level is met. The NCAP frontal 
and side crash tests are based on two of the crash tests carried out 
under the safety standards. However, the NCAP tests are conducted at 5 
miles per hour faster so that the differentiation between vehicles 
becomes more apparent. When considering changes to NCAP, NHTSA 
generally follows the rulemaking process, which includes seeking 
informal comments on proposed changes before they become effective.

The Senate Appropriations Committee Report accompanying the Department 
of Transportation appropriations bill for fiscal year 2004 (S. 1589) 
directed us to conduct a study of the New Car Assessment Program. This 
report examines (1) how NHTSA's New Car Assessment Program tests 
vehicles, rates their safety, and reports the results to the public; 
(2) how NCAP compares to other programs that test vehicles and report 
results to the public; and (3) the impact NCAP has had and the 
opportunities that exist to enhance its effectiveness.

To understand NHTSA's basis for testing and rating vehicles, we 
reviewed laws, regulations, and program documentation. We also 
conducted interviews with NHTSA officials, crash test contractors, 
vehicle manufacturers, trade associations, public interest groups, and 
independent researchers. We observed various vehicle crash tests and 
documented how the results were converted into star ratings. To 
document how NHTSA reports the results to the public, we consulted 
NHTSA officials, the Internet, and other vehicle safety information 
sources, such as Consumer Reports. To compare NHTSA's program with 
other crash test and rating programs, domestic and foreign, we 
interviewed officials of the Insurance Institute for Highway Safety and 
the New Car Assessment Programs of Australia, Europe, and 
Japan.[Footnote 2] We also interviewed vehicle safety experts and 
officials of foreign government entities, foreign vehicle 
manufacturers, and foreign consumer magazines such as Which?. To 
identify the impact of NCAP and opportunities for improvement, we 
analyzed changes in NCAP scores over time and obtained views from 
experts in the auto and insurance industries, public interest groups, 
and academia. We determined that NCAP data were sufficiently reliable 
for the purposes of this report. In addition, we analyzed how other 
organizations tested vehicles, rated the crash tests, and reported 
their results to the public to identify practices from other programs 
that may have potential application to the U.S. program. We conducted 
our work from March 2004 through April 2005 in accordance with 
generally accepted government auditing standards.

Results in Brief:

Under NCAP, NHTSA conducts three types of tests on vehicles--a full 
frontal crash test, a side crash test, and a rollover test, as shown in 
figure 1.[Footnote 3]

Figure 1: Three Types of Tests--Frontal, Side, and Rollover--Conducted 
by NCAP:

[See PDF for image]

[End of figure]

Each year, NHTSA tests a number of new vehicles that are predicted to 
have high sales volume, have been redesigned with structural changes, 
or have improved safety equipment. This practice is designed to ensure 
that NHTSA rates vehicles that consumers are buying. NHTSA develops 
five separate ratings based on the three tests and assigns one to five 
stars to indicate which vehicles are more crashworthy in frontal and 
side crashes and which vehicles are less likely to roll over. NCAP 
ratings, designed to help consumers decide which vehicle to purchase, 
are available to the public on the Internet and through the NHTSA 
Buying a Safer Car brochure. NCAP crash results are also incorporated 
in different vehicle safety ratings developed by others, such as 
Consumer Reports and The Car Book, both of which get more extensive 
distribution than direct NHTSA reporting.

We identified four other programs that crash test vehicles and report 
the results to the public--the Insurance Institute for Highway Safety's 
program and NCAP programs in Australia, Europe, and Japan. Like the 
U.S. NCAP, all these programs shared similar goals--providing relative 
safety information to consumers and encouraging manufacturers to 
improve vehicle safety. The three foreign NCAPs conduct different types 
of frontal and side crash tests and conduct pedestrian and child 
restraint systems tests that are not conducted under the U.S. program. 
The Insurance Institute also conducts different frontal and side crash 
tests than NHTSA's NCAP. Only the U.S. program conducts a rollover 
test. Each vehicle testing program also measures test results 
differently than the United States For example, in addition to the data 
provided by the crash test dummies in the vehicles, inspectors in other 
programs examine vehicles after crash tests to determine if there was 
intrusion into the passenger compartment or other abnormalities and 
adjust the test score accordingly. These other programs also report 
their testing results to the public in a different manner. While the 
U.S. NCAP reports results for each crash dummy by their seating 
locations in the crash test, all of the other programs summarize at 
least some of the scores or combined them into an overall 
crashworthiness rating in an effort to make it easier for the public to 
understand the results.

NCAP has contributed to making vehicles safer, but the program is at a 
crossroads where it will need to change in order to maintain its 
relevance. As shown in figure 2, vehicle safety as measured by NHTSA 
star ratings has improved since the program began.

Figure 2: Improvement of Average Star Ratings for Frontal and Side NCAP 
Tests:

[See PDF for image]

Note: Data include only the vehicles that were crash tested for each 
model year.

[End of figure]

The usefulness of the current tests has been eroded by changes in the 
vehicle fleet that have occurred since the program began. Today there 
are many more large pickups, minivans, and SUVs than existed 27 years 
ago, and this has created new safety hazards from the incompatibility 
between large and small vehicles and rollover crashes, which are not 
fully addressed by current NCAP tests. In addition, because most 
vehicles now receive four-or five-star ratings, NCAP tests provide 
little incentive for automakers to continue to improve vehicle safety 
and little differentiation among vehicle ratings for consumers. Lastly, 
NHTSA is upgrading its frontal and side crash tests in the Federal 
Motor Vehicle Safety Standards, which will make current NCAP tests less 
meaningful. For example, NHTSA is increasing the speed of the frontal 
safety standards to the same speed as the NCAP test, eliminating the 
difference between the frontal NCAP and safety standard tests. 
Opportunities to enhance the program include developing approaches to 
better measure the effects of crashes between large and small vehicles 
and occupant protection in rollovers, rating technologies that help 
prevent crashes from occurring, and using different measures to rate 
the crash results. NHTSA also has opportunities to enhance the 
timeliness of the tests and the presentation of the information 
provided to consumers.

We are making recommendations to the Secretary of Transportation to 
ensure that NCAP maintains its relevance in improving vehicle safety 
and to enhance the presentation and timeliness of the information 
provided to the public. We received oral comments from NHTSA on a draft 
of this report. In general, NHTSA agreed with the report's findings. We 
are also making a version of this report available at [Hyperlink, 
http://www.gao.gov], which includes video clips of crash tests that are 
conducted by NHTSA and others.

Background:

Motor vehicle crashes are complex events resulting from several 
factors, including driver behavior, the driving environment, and the 
vehicle.[Footnote 4] Vehicle design can affect safety through 
crashworthiness--that is, by providing occupants protection during a 
crash--and through crash avoidance--that is, by helping the driver to 
avoid a crash or recover from a driving error. Vehicle characteristics 
such as size, weight, and the type of restraint system affect 
crashworthiness because they play a large role in determining the 
likelihood and extent of occupant injury from a crash. Vehicle 
characteristics such as vehicle stability and braking performance are 
examples of crash avoidance features in that they aid the driver in 
preventing a crash from occurring.

The New Car Assessment Program (NCAP) was established in response to a 
requirement in the Motor Vehicle Information and Cost Savings Act of 
1972 to provide consumers with a measure of the relative 
crashworthiness of passenger vehicles.[Footnote 5] NCAP's goals are to 
improve occupant safety by providing market incentives for vehicle 
manufacturers to voluntarily design vehicles with improved 
crashworthiness and provide independent safety information to aid 
consumers in making comparative vehicle purchase decisions. NHTSA has 
pursued these goals by conducting frontal and side crash tests and a 
rollover test, assigning star ratings, and reporting the results to the 
public.[Footnote 6] In fiscal year 2004, NCAP conducted 85 crash tests 
and 36 rollover tests, with a budget of $7.7 million.[Footnote 7]

NHTSA also administers the Federal Motor Vehicle Safety 
Standards.[Footnote 8] All motor vehicles sold in the United States for 
use on the nation's highways must meet minimum safety requirements as 
required by the standards. The standards prescribe a minimum 
performance level for crashworthiness that vehicles must meet in a 
number of different crash tests. Auto manufacturers self-certify that 
their vehicles meet these minimum standards. To test compliance with 
some of these standards, NHTSA conducts 30 miles per hour (mph) frontal 
impact tests and 33.5 mph side impact tests for belted 
occupants.[Footnote 9]

The Federal Motor Vehicle Safety Standards tests serve as a foundation 
for NCAP testing. The test protocols for NCAP's frontal and side crash 
tests are the same as the safety standards, except that the NCAP tests 
are conducted at 5 mph faster. NHTSA's policy, although not required by 
law, has been to make changes to the safety standards before 
considering changes to NCAP. When considering changes to NCAP, NHTSA 
generally follows the informal rulemaking process, which includes 
seeking comments on proposed changes.

NCAP provides consumers with information regarding the crashworthiness 
of new cars beyond the applicable Federal Motor Vehicle Safety 
Standards with which all vehicles sold in the United States must 
comply. There are no minimum performance levels for the NCAP tests. 
NHTSA tests as many vehicles as possible under NCAP to provide 
consumers with sufficient independent information to make vehicle 
comparisons. In contrast, NHTSA relies on auto manufacturers to self-
certify compliance with the Federal Motor Vehicle Safety Standards and 
only conducts a limited number of tests to ensure manufacturer 
compliance.[Footnote 10]

NHTSA conducted the first NCAP crash tests in 1978 on model year 1979 
vehicles, measuring only the crashworthiness of passenger cars in 
frontal crashes. Since then, there have been a number of vehicle tests 
added to NCAP, as shown in figure 3. For model year 1983, NHTSA 
expanded NCAP to include light trucks, vans, and SUVs. In 1996, NHTSA 
first began the side-impact NCAP test for model year 1997 vehicles. 
NHTSA expanded the side-impact NCAP test to include light trucks, vans, 
and SUVs for model year 1999. NHTSA began to rate vehicles for their 
rollover risk beginning with the 2001 model year. NHTSA initially rated 
the risk of vehicle rollover by measuring the top-heaviness of a 
vehicle and comparing this measurement to the top-heaviness of vehicles 
involved in single-vehicle crashes, as reflected in crash data. As 
required by the November 2000 Transportation, Recall Enhancement, 
Accountability and Documentation (TREAD) Act, NHTSA began dynamic 
rollover testing on model year 2004 vehicles to supplement the 
measurement of a vehicle's top-heaviness in determining a vehicle's 
rollover risk.[Footnote 11]

Figure 3: Time Line of NCAP:

[See PDF for image]

[A] Indicates vehicle model year.

[B] NHTSA has not always used the star rating system to communicate a 
vehicle's crashworthiness as measured by NCAP. Previously, NHTSA 
published the numerical injury scores indicating the likelihood of 
head, chest, and upper leg injuries to the vehicle occupants. NHTSA 
devised the star rating system after the Senate and Conference 
Appropriations Reports for fiscal year 1992 requested that NHTSA 
improve methods of informing consumers of the comparative safety of 
passenger vehicles as measured by NCAP.

[End of figure]

NCAP Crash Tests Vehicles, Rates Their Safety, and Reports the Results 
to the Public:

NHTSA conducts three types of tests in NCAP: a full frontal crash test, 
an angled side crash test, and a rollover test.[Footnote 12] NCAP 
ratings, designed to aid consumers in deciding which vehicle to 
purchase, are available to the public on the Internet and through 
NHTSA's Buying a Safer Car brochure. NCAP crash results are also used 
in developing vehicle safety ratings by other organizations, such as 
Consumer Reports and The Car Book.

NCAP Conducts Three Tests--Full Frontal, Side, and Rollover:

Every year NHTSA tests new vehicles that are predicted to have high 
sales volume, have been redesigned with structural changes, or have 
improved safety equipment. NHTSA purchases vehicles--the base model 
with standard equipment--for frontal and side crash tests directly from 
dealerships across the country, just as the consumer would. The 
vehicles are provided to five contractors that conduct the crash tests. 
NCAP crash-test ratings only apply to belted occupants, as the crash 
test dummies used in NCAP tests are secured with the vehicle's safety 
belts. According to NHTSA officials, NCAP crash-test ratings are 
available on about 85 percent of the new vehicles sold because ratings 
for some models that have had no significant safety or structural 
changes are carried over from year to year. For the rollover tests, 
which are nondestructive, NHTSA leases new vehicles, which are tested 
at one contractor location. Rollover risk ratings are available for 
about 75 percent of new vehicles sold, according to NHTSA officials.

Full Frontal Crash Test:

The full frontal crash test is the equivalent of two identical 
vehicles, both traveling at 35 mph, crashing into each other head-
on.[Footnote 13] The test vehicle is attached to a cable and towed 
along a track at 35 mph so that the entire front end of the vehicle 
engages a fixed rigid barrier, as shown in figure 4. This type of crash 
test produces high level occupant deceleration, making this test 
demanding of the vehicle's restraint system.

Figure 4: Full Frontal Crash Test Conducted under NCAP:

[See PDF for image]

[End of figure]

Click the following link to watch a video of a full frontal crash test 
conducted by NHTSA NCAP at 35 mph: [Hyperlink, http://www.gao.gov/
media/video/d05370v1.mpg]:

Because the full frontal crash test is equivalent to two identical 
vehicles moving toward each other at 35 mph, the crash test results can 
only be compared to other vehicles in the same class and with a weight 
that is plus or minus 250 pounds of the test vehicle. The test 
protocols for the full frontal NCAP test are the same as the full 
frontal belted test in the Federal Motor Vehicle Safety Standards, with 
the exception of the test speed--the NCAP test is conducted at 35 mph, 
5 mph faster than the standard test.[Footnote 14]

Angled Side Crash Test:

The angled side crash test simulates an intersection collision in which 
one moving vehicle strikes another moving vehicle.[Footnote 15] The 
test vehicle is positioned such that the driver's side forms a 63 
degree angle with the test track. On the other end of the test track is 
a chassis with a barrier also turned at a 63 degree angle.[Footnote 16] 
The barrier is made of a deformable material to replicate the front of 
another vehicle and is attached to a cable that tows it down a track 
into the test vehicle at 38.5 mph. Both the barrier face and the 
driver's side of the vehicle are parallel, so that the entire face of 
the barrier impacts the side of the vehicle, as shown in figure 5.

Figure 5: Angled Side Crash Test Conducted under NCAP:

[See PDF for image]

[End of figure]

Click the following link to watch a video of an angled side crash test 
conducted by NHTSA NCAP at 38.5 mph: [Hyperlink, http://www.gao.gov/
media/video/d05370v2.mpg]:

Because all vehicles are hit with the same force by the same moving 
barrier, test results can be compared across weight classes. The 
barrier used in this test weighs approximately 3,015 pounds, and the 
top of the deformable face is approximately 32 inches from the ground. 
The side NCAP test is similar to the Federal Motor Vehicle Safety 
Standards test, with the exception that the side NCAP test is conducted 
at 38.5 mph, or 5 mph faster than the safety standard test.[Footnote 17]

Rollover Test:

The dynamic rollover test simulates a driver making a high-speed 
collision avoidance maneuver--steering sharply in one direction, then 
sharply in the other direction--within about 1 second. NHTSA has 
focused its rollover test primarily on pickups and SUVs because cars 
are not susceptible to tipping up in this test. The rollover test is 
actually a series of four runs, two left/right tests and two right/left 
tests, at two different steering wheel angles and different speeds. 
Before the test, the vehicle is loaded to represent five passengers and 
a full tank of gas. During the test, the steering wheel is turned 
sharply in one direction at a high speed and then turned sharply in the 
opposite direction at a greater steering angle.[Footnote 18] The first 
run of each test is conducted at 35 mph. Subsequent runs are conducted 
at about 40 mph, 45 mph, 47.5 mph and 50 mph, until the vehicle fails 
or "tips up" as defined by test procedures or attains a speed of 50 mph 
on the last run of each test without tipping up. Tipping up is defined 
as both wheels on one side of the vehicle lifting off the ground more 
than 2 inches simultaneously, which most commonly occurs during the 
second turn, as exhibited in figure 6. Outriggers are attached to the 
vehicle to prevent it from tipping all the way over and injuring the 
test driver.

Figure 6: Rollover Test Conducted under NCAP:

[See PDF for image]

[End of figure]

Click the following link to watch a video of a dynamic rollover test 
conducted by NHTSA NCAP at 48 mph: [Hyperlink, http://www.gao.gov/
media/video/d05370v3.mpg]:

NHTSA Rates Vehicles by Assigning Up to Five Stars to Communicate the 
Results of Its Tests:

NHTSA separately rates the frontal, side, and rollover tests. It 
assigns one (worst) to five (best) stars to communicate the results of 
the three tests to aid consumers in their vehicle purchase decisions. 
Each star in the frontal and side ratings corresponds to a diminishing 
probability of a potentially life-threatening injury, whereas each star 
in the rollover rating corresponds to a reduced likelihood of vehicle 
rollover. The rollover rating does not represent the chance of a 
potentially life-threatening injury should a rollover crash occur.

Frontal and Side Crashworthiness Ratings:

Frontal and side star ratings represent the chances of a person wearing 
a safety belt incurring an injury serious enough to require immediate 
hospitalization or to be life threatening in the event of a crash. 
Frontal star ratings indicate the combined chance of a serious head and 
chest injury[Footnote 19] to the driver and right front seat passenger, 
as shown in figure 7.

Figure 7: Frontal Star Rating and the Corresponding Chance of Serious 
Injury to the Head and Chest:

[See PDF for image]

[End of figure]

Side star ratings indicate the chance of a serious chest injury to the 
driver and the rear seat driver's side passenger, as shown in figure 
8.[Footnote 20] NHTSA reports two separate star ratings for the frontal 
and side crash test, according to the occupant position.

Figure 8: Side Star Rating and the Corresponding Chance of Serious 
Injury to the Chest:

[See PDF for image]

[End of figure]

In the side and frontal test, NHTSA uses crash test dummies that 
represent an average-sized adult male. Each dummy is secured with the 
vehicle's safety belts prior to the crash test.[Footnote 21] The 
dummies are affixed with instruments that measure the force of impact 
experienced in different parts of the body during the crash. While only 
forces to the head and chest are used to calculate the frontal star 
ratings, impacts to each dummy's neck, pelvis, legs, and feet are also 
measured.[Footnote 22] For the frontal rating, NHTSA calculates the 
chance of serious injury to the head and chest by linking measured 
forces on the dummies' heads and chests during the crash test to 
information about human injury.[Footnote 23] For the side rating, NHTSA 
calculates the chance of serious injury to the chest by linking 
measured forces on the dummies' ribs and lower spine during the crash 
test and information about human injury. Forces to the head and pelvis 
are also measured but are not included in side star ratings.

Rollover Rating:

NHTSA's rollover star ratings represent the propensity of a vehicle to 
roll over but do not address the probability of a severe injury in a 
rollover crash. Knowing a vehicle's propensity to roll is important 
because rollovers are the most deadly crashes. While totaling just over 
2 percent of police reported crashes, rollovers account for almost one-
third of all passenger vehicle occupant fatalities. The crash avoidance 
rollover rating is based primarily on the measure of a vehicle's top-
heaviness, as shown in figure 9, and, to a lesser extent, the results 
of the dynamic test.[Footnote 24]

Figure 9: Calculation of a Vehicle's Top-Heaviness:

[See PDF for image]

[A] The Static Stability Factor (SSF) is a vehicle's track width 
divided by two times its center of gravity height.

[End of figure]

NHTSA uses the measure of a vehicle's top-heaviness to predict the 
likelihood of a vehicle rolling over under the circumstances that occur 
most often--when a vehicle leaves the roadway and the vehicle's wheels 
hit a curb, soft shoulder, or other roadway object, causing it to roll 
over. These "tripped" rollovers account for about 95 percent of all 
rollover crashes. NHTSA's dynamic rollover test does not correspond to 
these types of rollovers because it does not involve the vehicle 
hitting a tripping mechanism, such as a curb or soft shoulder. As such, 
NHTSA's dynamic rollover test does not affect the star rating 
significantly, resulting in no more than a half-star difference in a 
vehicle's rollover rating. NHTSA primarily selects top-heavy vehicles, 
such as light trucks, small vans, and SUVs for the rollover 
test.[Footnote 25] NHTSA assigns one to five stars to reflect the 
chance of rollover, as shown in figure 10.[Footnote 26]

Figure 10: Rollover Star Rating and the Corresponding Risk of Rollover:

[See PDF for image]

[End of figure]

NHTSA Reports the Results to the Public Through the Internet and the 
Buying a Safer Car Brochure:

NHTSA distributes NCAP safety ratings and information about a vehicle's 
safety features through its Web site, press releases, and the Buying a 
Safer Car brochure. NHTSA primarily relies on the Web site to educate 
consumers about vehicle safety; in 2004 there were about 4.3 million 
visits to the NCAP Web site. The Web site was last redesigned in August 
2004 and provides information about crash test ratings from model year 
1990 to the present.[Footnote 27] To view a vehicle's ratings, users 
can search using parameters such as vehicle class, year, make, and 
model. Once a vehicle class and year are selected, the list of vehicles 
comes up with the star rating information, as shown in figure 11.

Figure 11: Ratings for a 2004 Passenger Vehicle as It Appears on 
NHTSA's Web Site:

[See PDF for image]

[End of figure]

Users can get more detailed information about the vehicle's star rating 
by selecting a specific vehicle, as shown in figure 12.

Figure 12: Details of Frontal, Side, and Rollover Star Ratings for a 
Passenger Vehicle as They Appear on NHTSA's Web Site:

[See PDF for image]

[End of figure]

In addition to the Web site, NCAP's star ratings and a list of 
vehicles' safety features are available in the Buying a Safer Car 
brochure. The American Automobile Association primarily distributes the 
brochure, and it is also available at NHTSA's regional offices, state 
highway safety offices, and libraries. For vehicle model year 2004, 
NHTSA published 25,000 copies of the Buying a Safer Car brochure. For 
vehicle model year 2005, NHTSA published a first printing of the 
brochure in December 2004. In addition, it plans to print a second 
brochure in spring 2005. While the 2004 edition does not have all the 
test results for model year 2005, it has a large number of carryover 
vehicles from model year 2004 plus some early 2005 tests.

Other sources of vehicle safety information that use data from NCAP 
crash tests include Consumer Reports and The Car Book. Consumer Reports 
takes into consideration a vehicle's performance in NHTSA NCAP tests 
and tests conducted by the Insurance Institute for Highway Safety 
(Insurance Institute) to determine an overall crash-protection rating. 
Instead of printing stars, Consumer Reports uses a circle rating 
scheme. Consumer Reports publishes this crash-protection rating, as 
well as individual NHTSA and Insurance Institute front and side crash 
test results, in its monthly magazine, in all of its newsstand-only new-
car publications, and on its Web site. Consumer Reports magazine has 
about 4 million subscribers, but representatives told us they inform in 
excess of 13.5 million people monthly as a result of pass-along 
readership. The Web site has an additional 1.8 million 
subscribers.[Footnote 28]

Published annually, The Car Book provides consumers with a broad range 
of information about new vehicles, listed alphabetically by model. 
Information such as fuel economy, repair costs, and front and side 
crash tests are included in the book. The Car Book takes the NCAP raw 
test results and converts them into a numerical rating scheme, 10 being 
best and 1 being worst. In addition to the information by vehicle 
model, The Car Book also presents detailed safety information based on 
the safety features of each car and the government's rollover ratings. 
Since first being published privately for the 1983 vehicle model year, 
The Car Book has sold over 1.5 million copies.[Footnote 29]

U.S. NCAP Differs from Other Crash Programs in Testing, Rating, 
Reporting, and Government Involvement:

We identified four other programs that crash test vehicles and report 
the results to the public--the Insurance Institute for Highway Safety 
(Insurance Institute) program in the United States and NCAP programs in 
Australia, Europe, and Japan.[Footnote 30] All of the programs shared 
the U.S. NCAP goals of providing manufacturers with an incentive to 
produce safer vehicles and providing consumers with comparative safety 
information on the vehicles they plan to purchase. We found differences 
in the types of tests conducted, how the crash tests were evaluated, 
and how the test results were shared with the public. In addition, we 
found that each program had varied levels of government and industry 
involvement.

Vehicle Testing Programs Conduct Different Tests:

Each of the organizations we examined conducts a variety of frontal, 
side, and other tests designed to measure various elements of vehicle 
safety. Figure 13 shows the tests performed across the U.S. NCAP and 
other four programs. (See appendixes II through VIII for additional 
discussion on each program and the tests conducted.)

Figure 13: Vehicle Safety Tests Conducted by Five Testing Programs:

[See PDF for image]

[End of figure]

AThe Insurance Institute conducts a perpendicular side-impact crash 
test with a SUV-like barrier.

Frontal Crash Tests:

The five programs we examined use two crash tests to represent frontal 
crashes--full frontal and offset crash tests. The U.S. and Japan NCAPs 
conduct full frontal tests, which involve crashing the test vehicle's 
entire front end into a solid barrier. The offset frontal test involves 
crashing the test vehicle traveling at 40 mph (64 kilometers per hour-
-km/h) into a deformable barrier with about 40 percent of the vehicle's 
overall width on the driver's side actually impacting the barrier, as 
shown in figure 14. All programs, except the U.S. NCAP, conduct the 
offset frontal test.

Figure 14: Offset Frontal Crash Test:

[See PDF for image]

Note: This test is conducted on the driver side of the vehicle, whether 
it is right-hand drive or left-hand drive. In the photo shown, the 
driver is on the right side of the vehicle.

[End of figure]

Click the following link to watch a video of an offset frontal crash 
test conducted by Australia NCAP at 40 mph: [Hyperlink, 
http://www.gao.gov/media/video/d05370v4.mpg]:

The full frontal and offset frontal tests measure different 
characteristics of vehicle crashworthiness. The full frontal test 
focuses on measuring the ability of the vehicles' restraint systems to 
protect the occupants. The offset frontal test assesses a vehicle's 
structural integrity and its ability to manage the crash energy 
generated from a crash entirely on one side of the vehicle. Officials 
from the programs using the offset test told us they believe it is more 
representative of real world crashes because most frontal crashes 
involve vehicles hitting only a portion of their front ends.

Side Crash Tests:

Three types of side-impact tests are conducted among the programs we 
examined--the angled side test, the perpendicular side test, and the 
pole side test. Only the U.S. NCAP performs the angled side 
test.[Footnote 31] All of the other testing programs conduct a 
perpendicular side tests. This test involves crashing a moving 
deformable barrier traveling at about 31 mph (50 km/h) into a 
stationary vehicle at a 90 degree angle centered on the driver's 
seating position. Figure 15 illustrates how the perpendicular test is 
performed.

Figure 15: Perpendicular Side Impact Crash Test:

[See PDF for image]

Note: This test is conducted on the driver side of the vehicle, whether 
it is right-hand drive or left-hand drive. In the photo shown, the 
driver is on the right side of the vehicle.

[End of figure]

Click the following link to watch a video of a perpendicular side 
impact crash test conducted by Euro NCAP at 31 mph: [Hyperlink, 
http://www.gao.gov/media/video/d05370v5.mpg]:

Other differences between the side tests were the height, shape, and 
weight of the barriers and the crash dummies used. For example, the 
U.S. NCAP and the three foreign programs performed their side tests 
using a moving deformable barrier with a front end simulating a 
passenger car, while the Insurance Institute's barrier simulates the 
front end of a typical pickup truck or SUV. In addition, the Insurance 
Institute barrier weighs about 3,300 pounds (1,500 kilograms--kg) 
compared to 3,015 pounds (1,367 kg) for the U.S. barrier and 2,095 
pounds (950 kg) for the Australian, European, and Japanese barriers. 
Also, the Australia, Europe, Japan, and U.S. side tests used 50TH 
percentile adult male dummies and the Insurance Institute used 
5THpercentile adult female dummies.[Footnote 32]

Insurance Institute officials told us they found that in serious real-
world side-impact collisions, occupants' heads are often struck by 
intruding vehicles, especially in the side collisions involving pickup 
trucks or SUVs with high front hoods. As a result, in 2003 when they 
began their side impact test, they developed the barrier to simulate 
these types of vehicles, while using dummies that represented smaller 
occupants. They said that the test challenges the automobile industry 
to provide additional occupant protection specifically for the head 
region. Figure 16 shows the difference in the size and height of the 
barriers, while figure 17 shows the crash test.

Figure 16: Comparison of Barriers Used in the NCAPs' Side Test (left) 
and the Insurance Institute Side Test (right):

[See PDF for image]

[End of figure]

Figure 17: Side-Impact Crash Test with SUV-like Barrier:

[See PDF for image]

[End of figure]

Click the following link to watch a video of a side-impact crash test 
with an SUV-like barrier conducted by the Insurance Institute for 
Highway Safety at 31 mph: [Hyperlink, http://www.gao.gov/media/video/
d05370v6.mpg]:

The Australia NCAP and European NCAP (Euro NCAP) also include optional 
pole side tests. The pole side test involves a side impact to a vehicle 
placed on a platform and propelled at about 29 km/h (about 18 mph) into 
a stationary cylindrical pole. The pole test is an optional extra test, 
available at the manufacturer's cost. This option is only available if 
a vehicle has head-protecting side air bags and receives the highest 
score in the side-impact test. If the vehicle performs well in the pole 
test, the vehicle can receive a higher overall score. Officials in 
Europe said this test is important, for example, because in Germany 
over half of the serious to fatal highway injuries occur when a vehicle 
crashes into a pole or a tree. The test is designed to encourage auto 
manufacturers to equip vehicles with head protection devices. Officials 
in Australia stated they are considering replacing the perpendicular 
side test with a pole side test to better test the increasing number of 
SUVs on their roadways. They said that SUVs are higher off the ground 
and heavier than most passenger cars. As a result, SUVs would always 
score higher under the current side-impact test because the barrier 
often impacts below the hip point on the dummy and would register 
little injury data. The pole test will impact all vehicles, including 
SUVs, the same way regardless of height and weight. NHTSA officials 
told us that while they have no plans at this time to include this test 
in NCAP, they plan to investigate revisions to the side NCAP once the 
pole test requirements for the Federal Motor Vehicle Safety Standards 
are resolved and finalized.[Footnote 33] Figure 18 illustrates how the 
pole test is performed.

Figure 18: Side Pole Crash Test:

[See PDF for image]

Note: This test is conducted on the driver side of the vehicle, whether 
it is right-hand drive or left-hand drive. In the photo shown, the 
driver is on the left side of the vehicle.

[End of figure]

Click the following link to watch a video of a side pole crash test 
conducted by Euro NCAP at about 18mph: [Hyperlink, 
http://www.gao.gov/media/video/d05370v7.mpg]:

Other Safety Tests:

In addition to the frontal and side crash tests, other safety tests are 
conducted in the various programs. These include vehicle rollover, 
pedestrian protection, and child restraint tests. The U.S. NCAP is the 
only program to conduct a vehicle rollover test.[Footnote 34] Officials 
of the other NCAPs told us they do not conduct this test because 
rollover has not been a major problem in their countries due to their 
smaller-sized vehicle fleet. However, Australian NCAP officials told us 
they have noted a growth in the size of their vehicle fleets, and they 
will be evaluating the usefulness of adding a rollover test to their 
programs.

The NCAPs in Australia, Europe, and Japan also conduct pedestrian 
tests, which are used to assess the risk to pedestrians if struck by 
the front of a car. The pedestrian test involves projecting adult and 
child-sized dummy parts (such as heads) at specified areas of the front 
of a vehicle to replicate a car-to-pedestrian collision. Officials in 
these programs said they included this test because pedestrian 
fatalities in some of their countries were quite high. For example, in 
2003 pedestrians accounted for nearly 30 percent of the annual traffic 
fatalities in Japan, 20 percent in Europe (nearly 30 percent in the 
United Kingdom alone), and 14 percent in Australia. In contrast, in the 
United States, approximately 5,000 pedestrians were killed in motor 
vehicle crashes in 2003, accounting for 13 percent of the annual 
traffic fatalities. Figure 19 illustrates how the pedestrian protection 
test is performed.

Figure 19: Head Form into Hood for the Pedestrian Test:

[See PDF for image]

[End of figure]

Click the following link to watch a video of a pedestrian test, where a 
head form is propelled into a vehicle hood, conducted by Euro NCAP: 
[Hyperlink, http://www.gao.gov/media/video/d05370v8.mpg]:

The NCAPs in Europe and Japan also conduct child restraint tests to 
evaluate child protection, although these tests are not directly 
related to crashworthiness. In Europe, two different child-size dummies 
are placed in child seats of the auto manufacturer's choice during the 
frontal and side crash tests, as shown in figure 20. In Japan, two 
child-size dummies are placed in child seats installed in the rear 
passenger seats of a test vehicle that has been stripped down to its 
body frame. The test vehicle is placed on a sled and subjected to a 
shock identical to the test speed used in the full frontal crash test. 
Japan NCAP also separately assesses the ease of correctly using child 
seats. NHTSA officials told us that the U.S. NCAP is conducting a pilot 
test to determine whether or not the addition of child safety seats 
into the frontal NCAP would provide meaningful consumer information. 
NHTSA also provides ratings on child safety seat ease of use.

Figure 20: Child Restraint Test:

[See PDF for image]

[End of figure]

Vehicle Testing Programs Rate Safety Differently:

Each vehicle testing organization used crash dummy readings as a 
principal part of its rating process.[Footnote 35] However, we found 
some differences in other aspects of the organizations' rating 
processes. For example, all programs except NHTSA supplement the dummy 
measures with inspector observations or measurements of the post-crash 
vehicles. In addition, in Europe and Australia, rating scores can be 
modified depending on the existence or absence of certain safety 
features. Further, each program except the Insurance Institute uses 
stars to convey the test results, and some programs combine individual 
ratings into summary ratings in an effort to make it easier for the 
public to understand crash test results.

Organizations Use Different Body Region Measurements and Types of 
Dummies to Develop Ratings:

The four organizations we reviewed used more dummy measures in 
calculating a vehicle's safety rating than U.S. NCAP. The U.S. NCAP 
uses head and chest crash dummy readings in frontal crashes and chest 
and lower spine readings for side crashes, then converts them to a 
probability for serious injury, which in turn is converted into a star 
rating.[Footnote 36] NHTSA officials said they use these measures 
because they are the most important indicators of serious or fatal 
injury in frontal and side crashes.[Footnote 37] In addition to the 
U.S. NCAP measures, the Insurance Institute uses measurements of the 
neck, left leg and foot, and right leg and foot for its frontal crash 
analysis and measurements of the head, neck, pelvis, and left leg for 
its side crash analysis. Australia and Euro NCAP use the neck, knee, 
femur, pelvis, and leg and foot for frontal tests and head, abdomen, 
and pelvis for side tests. Japan uses neck, femur, and tibia 
measurements for its frontal crash analysis and head, abdomen, and 
pelvis measurements for its side crash analysis. The other 
organizations use some of these additional measures to capture what in 
some cases may not necessarily be life-threatening injuries, such as 
those to the victim's legs. As discussed earlier, the U.S. NCAP 
measures the impact of crashes on many of the same body regions but 
does not use them to calculate safety ratings.

In addition to differences in the body areas being measured, some 
programs use different dummies in their side-impact tests. For the 
frontal tests, the U.S. NCAP and other organizations use dummies that 
represent an average-size adult male who is 5 feet 9 inches tall and 
weighs about 170 pounds. While this size dummy is used by most programs 
for the side-impact tests, there are differences in the dummy types and 
the instrumentation it contains.[Footnote 38] In addition, in its side-
impact tests, the Insurance Institute uses a smaller female dummy 
(about 5 feet tall and weighing about 110 pounds). Insurance Institute 
officials said they chose this dummy because there is evidence that 
females are more at risk in side collisions. It hopes this test will 
encourage manufacturers to install side curtain air bags that are 
designed to extend low enough to protect smaller passengers. Although 
NHTSA's proposed changes to the Federal Motor Vehicle Safety Standards 
would add a side-impact pole test using the average-size male and the 
smaller female dummies, NHTSA officials said that at this time they 
have no plans to alter the sizes or types of crash dummies they use but 
plan to investigate revisions to the side NCAP once the pole test 
requirements for the safety standards are resolved and finalized.

Using Inspector Observations to Supplement Dummy Results:

Another distinction between the U.S. program and other programs is the 
use of observations to modify test results. All programs except the 
U.S. NCAP observe or measure changes to various parts of the occupant 
compartment after the frontal crash test to identify potential safety 
concerns. For example, the Euro NCAP measures the intrusion of the 
steering column and lower leg area into the occupant compartment. Euro 
NCAP officials noted that while an intrusion may not have affected the 
dummy in the test, the potential for serious injury to vehicle 
occupants in real-world crashes causes them to lower the safety rating. 
Japan's NCAP also measures intrusion into the passenger compartment, 
but rather than relying on observation, Japan has established fixed 
measures that if exceeded will result in a lower score in a particular 
area.

The U.S. NCAP does not use observations to modify test scores. 
According to a NHTSA official, these observations add subjectivity to 
the rating assessments and are not based on criteria that can be 
repeated and substantiated. Many of the automobile manufacturers we 
contacted stated that using observations adds a subjective element to 
the test that is difficult for them to replicate. Additionally, some 
pointed out that in some cases different inspectors could reach 
different conclusions.

Using a Modifier System to Adjust Scores:

Another basic difference in scoring vehicles is the use of a modifier 
system in Europe and Australia. This system adjusts the score generated 
from the dummy injury data where injuries to occupants can be expected 
to be worse than indicated by the dummy readings or the vehicle 
deformation data alone. For example, a frontal test modifier might 
result in points being deducted if the dummy's head hit the steering 
wheel in a vehicle without an air bag.

The system in Europe and Australia also adjusts points based on the 
existence or absence of various safety features on the test vehicles. 
For example, a test vehicle can get extra points if it has a safety 
belt reminder system that meets their NCAP specifications. Officials 
said they use this approach to encourage manufactures to install new 
safety features sooner than might otherwise occur.

Officials from several organizations and automobile manufacturers 
operating under the Europe and Australia programs expressed concerns 
that some of the modifiers might not have a direct impact on occupant 
safety and could artificially increase scores. They noted, for example, 
that in some countries safety belt usage exceeds 90 percent and that 
giving extra points for a feature to encourage safety belt use may not 
really add to safety. In addition, some automobile manufacturers 
identified concerns with how items included in the modifier system are 
developed and measured. They said that in some cases they have received 
just 6 months notice of changes. They said that such changes can be 
expensive and that they need to be notified sooner, so they have time 
to make changes to comply with new measures.

Use of Stars as a Measure of Safety:

Except for the Insurance Institute, all programs used stars to convey 
test results. Officials from the NCAPs noted that star ratings are well 
understood by the public. For example, NHTSA officials said they used 
focus groups in 1993 to examine various options to communicate crash 
test results to the public, and the five-star rating was found 
preferable. In addition, officials in the other programs told us they 
followed the U.S. NCAP's use of star ratings. None of the programs has 
plans to change its rating measures.

There have been some concerns expressed about the use of stars. For 
example, a 1996 study by the National Academy of Sciences noted that 
stars are inherently positive symbols and the public may not understand 
the distinctions between the different levels of stars.[Footnote 39] In 
addition, officials of a consumer group noted that most people would 
associate the star rating with hotels and that staying in a three-star 
hotel would be quite acceptable to most people. In discussing its use 
of Good, Acceptable, Marginal, and Poor, the Insurance Institute said 
it considered these types of qualitative measures as being clearer to 
the general public.

Developing Summary Ratings:

Australia, Europe, and Japan NCAPs provide summary ratings, while the 
U.S. NCAP provides only individual ratings for each seating position 
that is included in the test for the frontal and side crash tests. For 
example, Australia and Euro NCAPs provide overall ratings that combine 
the frontal and side crash tests. Japan's NCAP combines frontal and 
side crash tests to provide overall ratings for the driver and 
passenger of a vehicle. Australian and European officials explained 
that they believed potential vehicle purchasers can be confused by the 
large amount of detail available on the test results and that 
summarizing results makes the ratings more useful. They noted they make 
the actual injury readings available for those interested in that level 
of detail. In addition, while the Insurance Institute does not combine 
individual ratings, it does identify "Best Pick-Frontal" and "Best 
Pick-Side" to assist consumers. Similarly, officials with 
publications like Consumer Reports and The Car Book told us they have 
found it helpful to provide consumers with summarized rating 
information. NHTSA officials noted that overall or summary ratings 
might hide or mask deficiencies in some areas of the tests. For 
example, they said that if a vehicle were to get a very high frontal 
rating and a very low side rating, merging the results could give 
consumers a misleading impression of the overall safety of that vehicle.

Programs Used a Variety Approaches to Inform Consumers of Safety 
Results:

The crash testing programs we examined used a variety of approaches to 
share safety results with the public. Across all the programs, the 
Internet was the most relied-upon source for getting information to 
consumers, with each organization providing details of its test 
results. Safety pamphlets were used by all programs to supplement the 
safety information presented on their Web sites. Some programs also 
work with the news media to increase awareness of test results.

Internet Used to Convey Results:

Each organization made the results of its testing program available to 
the public on the Internet. In general, the public can access the 
results of individual tests, including the actual numeric dummy 
readings. To help the public understand these results, each Web site 
uses charts, tables, and graphics. For example, in addition to 
providing star ratings, the Euro NCAP also uses color-coded dummy 
injury diagrams to display how the specific body regions perform in the 
frontal, side, pole, and pedestrian tests. The color-coded indicators 
are: Good (Green), Adequate (Yellow), Marginal (Orange), Weak (Red), 
and Poor (Brown). The color used is based on the points awarded for 
that body region, as shown in figure 21.

Figure 21: Dummy Injury Diagrams of Driver and Passenger in Frontal 
Test and Driver in Side Test:

[See PDF for image]

[End of figure]

Publications Used to Share Test Results:

Each testing organization publishes the results of its testing 
programs. The U.S. NCAP publishes the Buying a Safer Car booklet, which 
provides new and carryover crash test ratings. The Insurance Institute 
publishes a Status Report newsletter about 10 times a year, which 
contains new crash test ratings as well as other highway safety 
information. It can be obtained in hard copy through subscription, as 
well as downloaded from the Insurance Institute's Web site. Australia 
publishes a Crash Test Update brochure twice a year, which provides new 
crash test results. According to Euro NCAP officials, Euro NCAP divides 
its tests into two test phases and releases the results twice a year--
in November and June. The results are also published by What Car? (a 
British car magazine), Which? Car (a magazine owned and produced by 
British consumer associations), and the General German Automobile 
Association (ADAC) magazine. Other consumer magazines in Europe also 
provide crash test information. Lastly, Japan annually publishes the 
Choosing a Safer Car booklet, which provides new and carryover crash 
test results. The Japan NCAP also publishes summary brochures of test 
results.

Working with the News Media to Share Test Results:

Like the U.S. NCAP, the Insurance Institute and the Australia and Euro 
NCAPs worked with the news media to inform consumers about the results 
of the vehicle safety tests. For example, each program issued press 
releases to convey the results of safety research and crash tests. In 
addition, the Insurance Institute has worked with television 
broadcasts, such as the prime time news magazine program Dateline NBC, 
to raise the public's awareness of how vehicles perform in the 
program's crash tests. Insurance Institute officials grant interviews 
explaining the results of the tests and use broadcast-quality film and 
lighting to record the crash tests and make them available for 
television broadcasts. According to Japan NCAP officials, they work 
with television shows to help produce news segments that highlight 
changes in test procedures and recent test results. Further, according 
to Euro NCAP officials, in addition to other activities, Europe 
promotes consumer education by using crashed vehicles as public 
displays in prominent places in Europe during press conferences. The 
events are designed to attract news media and public attention in an 
attempt to increase public interest in and knowledge about car safety.

Government and Industry Involvement Varies among the Crash Testing 
Programs:

The level of government and industry involvement varies among the crash 
test programs. For example, the U.S. NCAP, which is operated and funded 
solely by the U.S. DOT, has traditionally based its U.S. NCAP on the 
Federal Motor Vehicle Safety Standards as a matter of agency policy and 
follows an informal rulemaking process where industry and other 
interested parties can submit comments once NHTSA issues a notice of 
proposed rulemaking. The Insurance Institute, which is funded by 
private insurance companies, has no such process and can make an 
internal decision to modify tests at any time. For example, according 
to Insurance Institute officials, when they began their side-impact 
tests, they developed a crash test barrier to represent the risk of 
severe head injuries in side impacts by SUVs and pickups. The Insurance 
Institute officials said they did not involve automobile manufacturers 
in the decision-making process but informed them as well as NHTSA 
before implementing the change.

The Australian NCAP was developed and is dominated mainly by private 
motor clubs but includes government transportation departments in six 
Australian states and territories, the New Zealand government, and 
consumer groups. The national Australian government sets minimum safety 
standards for vehicles but is not involved in funding or managing NCAP. 
Similarly, the Euro NCAP is sponsored by the governments of Great 
Britain, Sweden, Germany, France, and the Netherlands, as well as a 
number of motor clubs and consumer organizations. According to Euro 
NCAP officials, each sponsoring member agrees to perform or sponsor a 
number of crash tests and participates in making the decisions related 
to the program. In Australia and Europe, NCAP officials told us that by 
not being exclusively controlled by government, they have flexibility 
when modifying their programs. They said that as a result they can make 
changes quicker because they do not have to follow governmental 
procedures.

According to NCAP officials, the decision processes for Australia and 
Europe involve the use of committees and working groups to examine 
issues and make recommendations for change. The automobile industry and 
public safety organizations may be involved in providing research or 
opinions, but the committees are free to make decisions they believe 
are appropriate. When these committees make recommendations, the full 
governing body votes to accept or reject the changes. The government 
partners have a vote in the process but cannot veto the result. In 
Australia, according to NCAP and government officials, automobile 
manufacturers were initially reluctant to engage in meaningful dialogue 
with the officials of the Australia NCAP. However, more recently, 
Australia NCAP officials have consulted with manufacturers prior to 
making changes in the program and have received positive responses. On 
the other hand, the Euro NCAP allows industry representatives to 
participate in the discussions of the subgroups of its two technical 
working groups--primary safety and secondary safety.[Footnote 40] Also, 
the technical working groups and automobile manufacturers engage in 
direct dialogue in industry liaison meetings.

According to NCAP officials, Japan's NCAP is funded by the government 
but administered by an independent, government-appointed committee. The 
committee includes members who are experts from automobile research 
institutes, academics, journalists, and representatives of the Japanese 
automobile industry and the automobile importers association. This 
government/industry committee manages the program and must approve 
changes submitted by program officials. The committee reaches its 
decisions through consensus. Although the government ministry that 
oversees the program may override the committee's decisions, this has 
never occurred.

NCAP Has Contributed to Making Vehicles Safer, but Changes are Needed 
to Maintain Its Relevance:

NCAP has been successful in encouraging manufacturers to produce safer 
vehicles and providing consumers with comparative safety information. 
However, the program is at a crossroads where it will need to change to 
maintain its relevance. The usefulness of the current tests has been 
eroded by changes in the vehicle fleet that have occurred since the 
program began. Today there are many more large pickups, minivans, and 
SUVs than existed 27 years ago and new safety hazards have resulted 
from the incompatibility between large and small vehicles and rollover 
crashes, which are not fully addressed by current NCAP tests. In 
addition, because most vehicles now receive four-or five-star ratings, 
the NCAP tests provide little incentive for automakers to continue to 
improve vehicle safety and little differentiation among vehicle ratings 
for consumers. Lastly, NHTSA is upgrading its frontal and side crash 
tests in the Federal Motor Vehicle Safety Standards, which will make 
current NCAP tests less meaningful. Opportunities to enhance the 
program include developing approaches to better measure the effects of 
crashes between large and small vehicles and occupant protection in 
rollovers, rating technologies that help prevent crashes from 
occurring, and using different measures to rate the crash results. 
NHTSA also has opportunities to enhance the presentation and timeliness 
of information provided to consumers.

NCAP Has Encouraged Improvement in Vehicle Safety and Provided the 
Public with Vehicle Safety Information:

NCAP testing has contributed to more crashworthy passenger vehicles and 
NHTSA has informed the public of test results. As shown in figure 22, 
there has been a substantial increase in the average star rating of 
vehicles since testing began. In 2004, tested vehicles averaged about 
4.6 stars for the driver in frontal crash tests, about 4.4 stars for 
the passenger in frontal crash tests, about 4.4 stars for the driver in 
side crash tests, and about 4.3 stars for the rear passenger in side 
crash tests.

Figure 22: Improvement of Average Star Ratings for Frontal and Side 
NCAP Tests:

[See PDF for image]

Note: Data include only the vehicles that were crash tested for each 
model year.

[End of figure]

The improved ratings indicate that manufacturers have taken NCAP 
seriously and designed and built vehicles that do well on NCAP tests. 
Automakers told us that vehicle safety and NCAP test results have 
become an important marketing tool. As a result, many auto 
manufacturers advertise five-star ratings in government crash tests in 
their television, radio, and print ads.

NHTSA has informed the public of the NCAP test results through its Web 
site and by publishing a safety brochure. In addition, according to 
NHTSA officials, the NCAP Web site has been redesigned in an effort to 
make it more user-friendly. More importantly, NCAP crash test results 
are used by popular publications that influence large segments of the 
car-buying public. Both Consumer Reports and The Car Book use NCAP test 
results as part of their vehicle safety ratings.

Without Change, NCAP's Relevance Will Likely Diminish:

While NCAP has been successful in encouraging manufacturers to make 
safer vehicles, it will need to change to remain relevant. There have 
been significant changes in the makeup of the nation's vehicle fleet, a 
growing similarity of crash test ratings, and upgrades in the safety 
standard tests for frontal and side crashworthiness. Without addressing 
these changes, NCAP provides little incentive to manufacturers to 
continue to improve safety and may provide consumers with only limited 
comparative information on vehicle safety.

Since NHTSA began NCAP testing in 1979, there have been dramatic 
changes in the vehicle fleet. Vehicles such as pickups, minivans, and 
SUVs have transformed the fleet once dominated by passenger cars. There 
are now more than 85 million pickups, minivans, and SUVs on the road, 
representing about 37 percent of the vehicle fleet. The change in 
vehicle fleet presents new safety challenges that NCAP's testing does 
not fully address--vehicle incompatibility and rollover. The issue of 
incompatibility emerges when a large vehicle such as a pickup, minivan, 
or SUV crashes into a smaller, lighter vehicle because the larger 
vehicle can inflict serious damage that is particularly dangerous to 
the occupants of the smaller vehicle. The current NCAP frontal and side 
tests do not account for vehicles of different size, weight, and 
geometry crashing into one another. Significant differences in ratings 
can result when tests are designed to address these vehicle 
differences, as evidenced by comparing the Insurance Institute side 
test results with NCAP results. The Insurance Institute, which uses a 
higher SUV-like barrier, gave 27 vehicles its lowest rating (Poor) in 
side-impact tests, primarily because there were no side air bags in the 
vehicle. NHTSA, which uses a low barrier and, unlike the Insurance 
Institute, does not include head measures in its star calculations, 
gave 21 of these same 27 vehicles (77 percent) four-or five-star safety 
ratings. Also, with the increase in pickups, minivans, and SUVs in the 
nation's fleet, vehicle rollover has become a more important issue; in 
2003, rollovers accounted for over 10,000 fatalities, or more than 30 
percent of all passenger vehicle occupant fatalities. However, the NCAP 
rollover test only measures the likelihood that a vehicle will roll 
over and does not assess the safety afforded to occupants should a 
rollover occur.

NCAP frontal and side crash test results have improved to a point where 
there is little difference among most vehicles' ratings. In 2004, 
NHTSA: provided the public with NCAP rating information for 234 
vehicles.[Footnote 41] Most of these vehicle ratings were four or five 
stars for drivers and passengers in frontal and side crash tests, as 
shown in figure 23.

Figure 23: Frequency of Four-and Five-Star Ratings for Frontal and Side 
Crash Tests in 2004:

[See PDF for image]

Note: For the driver and front passenger in the frontal test, star 
ratings were available for 234 vehicles in 2004. For the driver in the 
side test, star ratings were available for 210 vehicles in 2004. For 
the rear passenger driver's side, in the side test, star ratings were 
available for 186 vehicles in 2004.

[End of figure]

The vehicles crash tested more recently have done even better. Of the 
49 frontal and 18 side crash tests conducted in 2004, over 95 percent 
received a four-or five-star rating. As a result, NCAP's ability to 
challenge auto manufacturers to continue improving vehicle safety has 
eroded. Also, with almost all scores being about the same, consumers do 
not have comparative safety information that differentiates 
significantly among vehicles.

Lastly, NHTSA is upgrading the frontal and side tests under the Federal 
Motor Vehicle Safety Standards, which make current NCAP testing less 
meaningful.[Footnote 42] For frontal tests, safety standards will 
require that for vehicles built after September 1, 2007, manufacturers 
must certify the crashworthiness of their vehicles at 35 mph (instead 
of the current 30 mph).[Footnote 43] This change will eliminate the 
speed difference between the frontal NCAP and the frontal belted safety 
standard tests. Because of this change, NHTSA has begun to examine 
alternatives to its current frontal crash test program and hopes to 
finalize any changes to the NCAP frontal test in 2006. Similarly, NHTSA 
announced in May 2004 that it is proposing to add a 20 mph side pole 
crash test to the Federal Motor Vehicle Safety Standards. This test 
will use a more technically advanced average-size male dummy than is 
currently used in the NCAP tests and a dummy that represents a small 
female. According to NHTSA officials, the new test and advanced dummy 
will enable them to confidently measure compliance with head injury 
standards and challenge automakers to provide adequate head protection 
to vehicle occupants in side impact crashes. However, neither this test 
nor the new dummies are currently part of NCAP. NHTSA officials said 
they plan to begin examining alternatives to the side crash test at the 
end of 2005.

Opportunities to Enhance NCAP Testing:

NHTSA could explore several opportunities to enhance NCAP and ensure 
its relevance. These opportunities include (1) addressing changes to 
the vehicle fleet, particularly as it relates to vehicle 
incompatibility and rollover; (2) developing approaches for NCAP to 
encourage improved safety from emerging technology that helps drivers 
avoid crashes; and (3) examining the various testing procedures and 
measures that are available and in use by other organizations and 
determining their applicability to NCAP.

Addressing the Increased Fatality Risks Created by Changes in the 
Vehicle Fleet:

Vehicle Incompatibility:

When pickups, minivans, and SUVs collide with smaller passenger cars, 
the mismatch of the vehicles' weight, height, and geometry are 
considerable, as shown in figure 24. In terms of the weight 
differences, subcompact cars may weigh as little as 1,500 pounds while 
the large SUV may exceed 6,000 pounds.

Figure 24: Example of Height Mismatch Between Vehicles of Different 
Type and Weight:

[See PDF for image]

[End of figure]

Because of the higher ground clearance of large pickups and SUVs, their 
bumpers may skip over the crash structures of passenger cars, raising 
the likelihood that an occupant of the car will be killed or seriously 
injured.[Footnote 44] A 2003 NHTSA study found that in frontal 
collisions involving a car and a light truck or van, there were almost 
four times the number of fatalities in the car than in the light truck 
or van.[Footnote 45] The success of NCAP and the other testing programs 
may have indirectly contributed to this problem. According to some 
experts, to improve crashworthiness scores of large vehicles, vehicle 
manufacturers have increased the rigidity of the structure that absorbs 
and manages the substantial forces in the crash tests.[Footnote 46] As 
a result, the structure of large vehicles has had to become more 
substantial and stiffer than that of smaller vehicles because the 
larger vehicles must absorb more energy in the crash test due to their 
greater weight.

NHTSA's NCAP frontal tests could potentially be modified to measure and 
rate vehicle incompatibility. Some experts, NHTSA officials, and 
vehicle manufacturers told us that there are a number of approaches 
being investigated that could help to address vehicle incompatibility. 
For example, some researchers are examining the use of sensors in test 
walls; crashing a moving deformable barrier into the front of the test 
vehicles, instead of propelling the test vehicle into a solid wall; or 
crashing test vehicles into a solid wall at varying speeds, depending 
on the size of the vehicle, to equate the crash to hitting a 
standardized vehicle. The hypothesis is that information obtained by 
measuring how vehicles strike the crash test barrier could be used to 
estimate the relative damage that a vehicle would cause in collisions 
with another vehicle and could be used to rate the aggressiveness of 
vehicles. Using a moving barrier for frontal crash tests would make 
test results comparable across weight classes, as is the case with the 
current side-impact rating, because all vehicles would be struck by the 
same size barrier. Using variable speeds based on vehicle weight would 
also allow ratings of small and large vehicles to be compared. Each of 
these alternatives requires further development and testing to assess 
the overall safety implications, including the potential for reducing 
fatalities in passenger cars when struck by larger vehicles, the 
potential for diminished occupant protection for large vehicles in 
single vehicle crashes, and consideration of potential costs.[Footnote 
47] Ratings based on these tests could provide manufacturers with 
incentives to address incompatibilities between large and small 
vehicles and provide consumers with information on the potential safety 
hazards associated with vehicle incompatibility.

The problem of vehicle incompatibility is even worse in side crashes. 
When a large vehicle like an SUV crashes into the side of a small 
vehicle, the larger vehicle may miss the door sill of the vehicle, 
causing most of the energy to be directed to the door and window areas, 
as shown in figure 25. In such cases, the injuries can be exacerbated 
when there is no side head protection, leaving the window as the only 
barrier between the occupant's head and the impacting vehicle. Head 
injuries are a major cause of fatalities in side collisions, 
particularly in crashes where a single vehicle strikes a tree or 
utility pole and in intersection crashes where smaller, lighter 
vehicles are hit in the side by larger, heavier vehicles. NHTSA has 
estimated that in serious side-impact crashes involving one or more 
fatalities in 2002, nearly 60 percent of those killed suffered brain 
injuries.

Figure 25: Example of the Damage Caused by an SUV Striking the Side of 
a Small Vehicle:

[See PDF for image]

[End of figure]

There are also possibilities for modifying the NCAP side test to help 
address vehicle incompatibility. For example, NHTSA could examine the 
barrier that is being used to ensure that it best represents today's 
vehicles. NHTSA's current side-impact barrier is about the size and 
weight of a compact car. As a result, when this barrier hits the test 
vehicle, it will almost always hit the bottom sill of the door, which 
is designed to manage much of the crash energy.[Footnote 48] To address 
the disparity in height between passenger cars and SUVs, the Insurance 
Institute uses a side-impact test barrier that is larger and higher 
than NCAP's barrier, as shown in figure 26. According to Insurance 
Institute officials, they designed this barrier to represent an SUV so 
their test could more accurately reflect the increased risk for 
occupants in smaller vehicles. They said that it has encouraged 
manufacturers to install side curtain air bags. Using this higher 
barrier has resulted in different scores than NHTSA's NCAP. For 
example, the Insurance Institute has given 27 vehicles its lowest 
rating (Poor) in side-impact tests, while NHTSA, which uses a low 
barrier and does not include head measures in its star calculations, 
gave 21 of these 27 vehicles (77 percent) four-or five-star safety 
ratings.

Figure 26: Example of Head Movement during the Insurance Institute Test 
with the SUV-like Barrier:

[See PDF for image]

[End of figure]

Click the following link to watch a video of an interior view of the 
side impact crash test with an SUV-like barrier conducted by the 
Insurance Institute for Highway Safety at 31 mph: [Hyperlink, 
http://www.gao.gov/media/video/d05370v9.mpg]:

Officials from a number of automobile makers told us that vehicle 
compatibility is an important safety issue, and they are working to 
enhance occupant protection in front and side crashes, outside of NHTSA 
safety standards or NCAP testing. Several automakers voluntarily 
entered into an agreement with the Insurance Institute to work 
collaboratively to have all of their vehicles meet new safety criteria 
that require large vehicles to match the height of the fronts of small 
vehicles by September 2009, as shown in figure 27. According to 
Alliance of Automobile Manufacturers members, better matching of 
structural components may enhance the ability to absorb crash forces, 
thereby reducing occupant fatalities by an estimated 16 to 28 percent. 
The agreement also specified that by September 1, 2007, at least 50 
percent of these automakers' vehicles offered in the United States will 
meet enhanced side-impact protection with features such as side air 
bags, air curtain bags, and revised side-impact structures. By 
September 2009 all vehicles of these manufacturers are to meet the new 
side criteria.

In commenting on a draft of this report, NHTSA officials noted that in 
order for 50 percent of the vehicles to meet the voluntary side 
requirements by September 1, 2007, manufacturers can certify by using 
either the existing Federal Motor Vehicle Safety Standard pole test or 
the Insurance Institutes side impact test. They noted that in September 
2009, the pole test will no longer be an option and that, therefore, it 
is very possible that large vehicles, such as pickups, minivans, and 
SUVs, would be able to pass the test without incorporation of enhanced 
side-impact features such as side air bags or curtains for the 
following reasons:

* Manufacturers may not need to subject large vehicles to the pole test 
by September 1, 2007, if 50 percent of its fleet is comprised of 
smaller passenger cars.

* Larger vehicles will sustain a lower velocity change than smaller 
vehicles when struck by the Insurance Institute barrier.

* The higher ride height of large vehicles could keep the dummy's head 
from striking the top of the Insurance Institute barrier.

Figure 27: Example of Improved Matching of Frontal Vehicle Structures:

[See PDF for image]

[End of figure]

Occupant Protection in Rollover Crashes:

Given the changes in the vehicle fleet, fatalities due to rollover 
crashes have continued to increase. Rollovers are dangerous incidents 
and have a higher fatality rate than other kinds of crashes. Just over 
2 percent of all police-reported crashes that occurred in 2003 were 
rollovers, but they accounted for over 10,000 highway fatalities, or 
more than 30 percent of all passenger vehicle occupant deaths. All 
types of vehicles can roll over. However, taller, narrower vehicles 
such as pickups, minivans, and SUVs have higher centers of gravity and 
thus are more susceptible to roll over if involved in a single-vehicle 
crash. NHTSA reported that 61 percent of fatalities in SUVs and 45 
percent of fatalities in pickups in 2002 were the results of rollover 
crashes.[Footnote 49] NCAP's rollover testing does not rate the chance 
of a potentially life-threatening injury should a rollover crash occur; 
it only measures the risk of rollover.

Although NHTSA has not incorporated occupant protection in rollovers 
into NCAP, officials said they have been examining occupant protection 
in rollover crashes, focusing on reducing occupant ejection and 
increasing roof strength through regulation. According to NHTSA 
officials, the most deadly rollovers occur when unbelted occupants are 
completely ejected from the vehicle though doors, windows, and sun 
roofs and when the roof crushes into the occupant compartment, causing 
serious, if not deadly, head, neck, and spinal cord injuries. NHTSA has 
proposed changes to the Federal Motor Vehicle Safety Standards that 
would upgrade the door lock requirements to help prevent vehicle 
occupant ejection and increase roof strength. They are also considering 
other ways to prevent ejection, specifically looking at the potential 
of side curtain air bags to prevent ejection through vehicle windows.

NHTSA's NCAP rollover testing could be modified to better measure and 
rate the risks of serious injury associated with a rollover crash. 
NHTSA officials and others said that they have not been able to develop 
a repeatable crash test in which the vehicle rolls over and dummies 
would be used to measure injuries. However, in the absence of such a 
rollover crash test, NCAP could examine various aspects of the vehicle 
which are known to affect occupant safety in rollover, such as rating 
the roof strength of vehicles. For example, officials from a consumer 
group told us that NHTSA could conduct dynamic tests on roof strength 
and point to a 2002 Society of Automotive Engineers paper that attests 
that such drop tests for roof strength are repeatable. They also said 
that there has been other promising research that would measure roof 
crush in dynamic tests. However, including such tests in NCAP would 
require further development and funding considerations.

Incorporating Active Safety Technologies into NCAP:

NCAP also has an opportunity to begin assessing new technology that 
could help prevent crashes. Vehicle manufacturers and others have been 
developing and testing new active safety systems that hold promise for 
reducing traffic fatalities by helping drivers avoid crashes 
altogether. These active safety systems include improving vehicle 
handling and braking in emergency situations, providing warning alerts 
for potential collisions or straying out of roadway lanes, and 
providing distance alerts when driving too close to another vehicle. A 
2004 NHTSA study estimated that the incorporation of electronic 
stability control systems[Footnote 50] could reduce certain crashes by 
about 67 percent.[Footnote 51] Similarly, the Insurance Institute 
reported that electronic stability control can reduce the risk of 
involvement in single vehicle crashes by more than 50 percent.[Footnote 
52] Some experts suggested that NCAP might be used to encourage and 
speed the adoption of active safety systems into the vehicle fleet.

Some elements of active safety systems are included in some current 
tests. While the rollover test is not designed to measure the 
effectiveness of electronic stability control systems, vehicles 
equipped with this technology would be expected to perform better in 
the rollover test because the vehicle would be less likely to tip up. 
In addition, brake tests are conducted as part of Japan's NCAP, with 
the results provided as a separate safety rating. The Euro NCAP has 
also established committees to identify potential active safety systems 
to include in their program, as well as the testing protocols that 
would be used.

While using NCAP to further test and rate active safety systems could 
encourage their adoption in the marketplace, there are challenges to 
overcome. According to NHTSA officials, NHTSA would first need to 
identify those active safety systems that could be effective in 
preventing crashes. They said this would be difficult because they 
would have to determine how well a system helps drivers avoid crashes. 
Also, determining the testing methodology would be challenging because 
the effectiveness of some active systems could be affected by factors 
such as driver behavior and the physical characteristics of the road, 
such as the dampness of the pavement.

Officials from various automobile manufacturers told us that they are 
developing many new active safety systems with the objective of helping 
drivers avoid crashes. They pointed out that while NCAP could be used 
to encourage them to market such systems, they would have concerns 
regarding which systems to include in NHTSA's program and how the 
system would be rated. In addition, they noted that because of 
competitive forces, active safety advances could be available sooner 
than NHTSA is capable of deciding to include them and developing an 
acceptable approach for testing and rating them. Officials from 
automakers said they are willing to share their research and work in 
cooperation with NHTSA to develop tests or measurements that could help 
NCAP address these issues.

Using Additional Test Measures and Different Size Dummies:

NHTSA could provide consumers with more safety information by using 
additional test measures and different crash dummies. All of the other 
organizations we contacted used more dummy measures to calculate 
vehicles' safety ratings than U.S. NCAP used. To determine the star 
ratings, NHTSA uses head and chest readings from the frontal NCAP test 
and chest and lower spine readings for side-impact tests. Other 
organizations use measurements that included such areas as the head, 
neck, chest, leg, and foot for frontal test ratings and the head, neck, 
chest, pelvis, and leg for side test ratings.[Footnote 53]

The concern with using few dummy readings is that the safety rating 
might not include important safety considerations. While NHTSA uses 
head and chest readings for frontal ratings and chest and lower spine 
readings for the side ratings, it measures other items during crash 
tests and may identify them as "Safety Concerns" on its Web site if 
they exceed certain values. We identified over 140 Safety Concerns on 
NHTSA's Web site since vehicle model year 1990--36 of these were for 
vehicles that received four-or five-star ratings.[Footnote 54] The 
Safety Concerns included high femur readings in frontal crashes, which 
could mean there was a high likelihood of thigh injury; high head 
acceleration readings in side crashes, which could indicate a high 
likelihood of serious head trauma; and doors opening during side crash 
tests, which could increase the likelihood of occupant 
ejection.[Footnote 55] Having a Safety Concern noted for vehicles with 
a four-or five-star rating presents conflicting information that could 
be confusing to consumers.

As NHTSA makes changes to its testing program, it has the opportunity 
to reexamine the size and type of dummies it uses in crashes in 
addition to the body areas of the dummies being measured. At present, 
NHTSA's dummies equate to an average-size adult male who is about 5 
feet 9 inches tall and weighs about 170 pounds. Most of the other 
organizations use this size dummy in their crash tests, and vehicle 
manufacturers work to maximize the safety systems for an occupant with 
these characteristics. However, not all vehicle occupants are the same 
size, and optimizing the restraint system for the average male would 
not necessarily be optimum for others who may be smaller, shorter, 
taller, or heavier.[Footnote 56] Also, children and the elderly may 
react differently to crash forces than the average-size male. 
Recognizing this, the Insurance Institute uses a smaller female dummy 
(about 5 feet tall and weighing about 110 pounds) in the driver and 
rear seat of the side-impact test. Insurance Institute officials said 
they made this change to encourage manufacturers to install side 
curtain air bags that would extend low enough to protect the heads of 
smaller passengers. In addition, in its proposed side-impact pole 
standards test, NHTSA specifies using a 50tTHpercentile male and a 
5tTHpercentile female to address the issue of different size drivers 
and passengers.

The U.S. NCAP officials said that at this time they are waiting on the 
resolution to the proposed safety standard changes that would add a 
side pole test before deciding on altering the size or type of crash 
dummies they use. While generating additional information on which to 
base safety ratings, altering the size of the dummy in the NCAP tests 
could provide challenges for automobile manufacturers because they 
would have to conduct more internal tests. Officials from many vehicle 
manufacturers said they must already conduct hundreds of crash tests 
each year to ensure that they meet the variety of tests and dummies 
used in NHTSA's standards, U.S. NCAP, and tests conducted by the other 
testing organizations.[Footnote 57]

Opportunities to Enhance the Presentation and Timeliness of NCAP 
Results:

NCAP has the opportunity to enhance its program by changing the way it 
reports test results. Specifically, it could provide summary ratings, 
present information in a comparative manner, increase public awareness, 
and make results available earlier in the model year.

Providing Summary Ratings:

According to some safety experts, NHTSA could improve its program by 
developing an overall safety rating rather than reporting four separate 
ratings for crash tests.[Footnote 58] Consumer Reports, The Car Book, 
the Insurance Institute, and all of the other NCAPs provide more 
summary information for consumers than NHTSA. Further, a 1996 National 
Academy of Sciences study that examined NCAP recommended that NHTSA 
provide an overall rating to provide consumers with an overview of a 
vehicle's safety. However, the study also recommended that NHTSA make 
the detailed test results available for those consumers who wish to 
examine them more fully.[Footnote 59]

NHTSA and Insurance Institute officials said they did not develop an 
overall crashworthiness rating because combining ratings are 
technically difficult and could obscure low ratings in one test area 
that would be revealed when test results are reported separately. 
Insurance Institute officials added that consumers can evaluate the 
different ratings to determine those that are most applicable to their 
situations. They said a person who is primarily the sole occupant of a 
vehicle might not be as concerned with the passenger safety rating as 
someone who routinely carries passengers.

NHTSA officials said that they will continue investigating the 
feasibility of creating an overall safety rating for vehicles. However, 
they said that they would like to incorporate additional elements into 
such a rating. For example, they said that it is important to develop a 
rating that considers more than just the frontal and side-impact test 
results, such as the rollover results and vehicle compatibility, which 
can have a large bearing on the overall safety of vehicles. In their 
view, without the elements that address rollover and compatibility, 
consumers might get the wrong impression of the relative safety of 
vehicles. Officials said they have not yet developed a method to 
incorporate the rollover rating into an overall rating and have not 
identified measures to reflect vehicle compatibility, although they 
have long recognized compatibility as an issue. They could not estimate 
how long it would take to address the problem of adding the rollover 
rating to a combined rating but said they would pursue developing a 
summary safety rating for vehicles after they decide how to measure 
vehicle compatibility.

Comparative Safety Information Could Benefit the Public:

Each testing organization uses a different presentation approach for 
reporting its test results, with some providing additional information 
to the public. The U.S. NCAP provides separate star ratings for the 
four dummy positions in the two crash tests and the rollover test. The 
only ratings the U.S. NCAP presents in a comparative manner are the 
rollover ratings, which compare vehicle performance within a class of 
vehicles, such as pickup trucks. In contrast, Australia's and Japan's 
NCAPs provide more comparative information by supplementing their star 
ratings by adding bar charts that show how well the vehicle performed 
in the tests, as shown in figures 28 and 29. The Australia publication 
shows that although two vehicles received three stars, one of them 
performed better than the other. The Japan NCAP rating shows that the 
vehicle received five stars for overall driver safety but that the 
passenger score was higher than that of the driver.

Figure 28: Example of Australia's NCAP Safety Rating Information:

[See PDF for image]

[End of figure]

Figure 29: Example of Japan's NCAP Safety Rating Information:

[See PDF for image]

[End of figure]

Similarly, Consumer Reports provides summary safety ratings for 
accident avoidance and crash protection and uses a bar chart to present 
its overall safety score. Consumer Reports also lists vehicles in 
ranked order rather than alphabetically, provides comments to highlight 
particular aspects of each vehicle's performance, and uses qualitative 
descriptions--Excellent, Very Good, Good, Fair, and Poor--to help 
inform its readers of safety results. Consumer Reports officials said 
that the overall rating provides an overview of the vehicle's safety, 
and the two summary categories of accident avoidance and crash 
protection provide additional information that consumers may want.

NHTSA recently began using a rating system for its rollover assessment 
that indicates, along with the star rating, the percentage of 
likelihood that a vehicle may roll over. NHTSA's rollover information 
provides an extra level of detail of vehicle performance than the 
information provided for the frontal and side collision tests. The 
rollover results are ranked according to performance and, as 
illustrated in figure 30, show how well each vehicle performed within 
the range of performance of its vehicle class, such as passenger cars, 
pickups, vans, and SUVs.

Figure 30: Example of NHTSA's Rollover Rating for a Pickup Truck with 
Bar Chart Showing How the Vehicle Performed within Its Class:

[See PDF for image]

[End of figure]

Increasing Public Awareness of Results:

NHTSA could look to other programs for innovative ways to garner more 
interest in crash test results. Like other testing organizations, NHTSA 
uses the Internet, brochures, and press releases to inform the public 
of NCAP ratings. However, other organizations use additional approaches 
to inform the public of their program and test results. For example, 
the Japan Automobile Federation creates public awareness of the program 
with a portable sled in which the general public can experience a 
simulated collision at 5 kilometers per hour and have a protective air 
bag deploy. The Euro NCAP also stages a public display of crash 
vehicles. They try to select areas where media and public interest 
would be high. Recent events were held in Wenceslas Square, Prague; 
Athens; and London. Figure 31 shows two events, one in London and 
another in Prague.

Figure 31: Examples of Euro NCAP Public Displays of Crash Vehicles:

[See PDF for image]

[End of figure]

There have also been proposals to increase public awareness of NCAP 
results by requiring their inclusion on new car stickers. For example, 
S. 1072, a bill introduced in the 108th Congress to reauthorize funds 
for federal aid highways, highway safety programs, and transit 
programs, included a provision that would require automakers to include 
NCAP test results on new car stickers. Officials from consumer advocate 
groups told us that they support such an approach because consumers 
would have information available at the time of their purchase 
decisions. Officials from automakers said that there are a number of 
challenges that would need to be overcome if such an approach were 
taken, including scheduling tests to ensure that results are available 
in time for the information to be included on new car stickers.

NCAP Ratings Could Be Available Earlier in the Car Model Year:

NHTSA could conduct vehicle tests earlier and release NCAP ratings 
sooner in the model year, which would make the results more useful for 
consumers. NCAP ratings are often released late in the model year, 
after many of the vehicles have already been purchased. In May 2003, 
long after the beginning of model year 2003, NHTSA released the results 
of some model year 2002 vehicle tests. NHTSA published its Buying a 
Safer Car brochure for 2004 in February 2004, about 6 months after the 
vehicles were available for sale and before all of the tests were 
completed for the 2004 models. To the extent that test results are 
available sooner, more car buyers could have safety information to help 
make their purchase decisions. For example, by the time NHTSA released 
the Buying a Safer Car brochure in February 2004, according to industry 
sales statistics, about 7.7 million, or over 46 percent, of new cars 
and trucks had been purchased in the United States.[Footnote 60] For 
model year 2005, NHTSA attempted to address the issue of getting timely 
information to consumers by publishing an early edition of its Buying a 
Safer Car brochure in December 2004. This publication included test 
results for some 2005 models. In addition, towards of the end 2004, 
NHTSA began posting the results to its Web site as soon as the quality 
control process was completed. NHTSA officials plan to publish an 
updated version in spring 2005, after additional testing has been 
completed.

There are several factors that affect the timing of the testing and the 
release of NCAP ratings. First, NHTSA obtains vehicles for NCAP testing 
directly from the dealerships and leasing companies to ensure that each 
vehicle is representative of that make and model. Under this approach, 
testing cannot begin until after vehicles are available for purchase by 
the public--the model year begins in September for many companies. In 
addition, NHTSA does not receive its funding until after the fiscal 
year begins on October 1ST of each year. Further, due to the number of 
vehicles to be included, vehicle testing is spread out over a period of 
months. As a result, testing can extend from October though April. 
Until recently, NHTSA did not make ratings available to the public as 
soon as the results were known but waited until all testing of a 
vehicle category was finished before issuing a press release announcing 
the test results. Beginning with model year 2005 tests, NHTSA began 
posting the test results to its Web site after the quality control 
process was complete. Press releases continue to be generated after 
each batch of tests is completed. NHTSA officials said that by 
releasing the results this way, consumers have comparative information 
on all vehicles of one type at the same time.

One testing organization has addressed some of the timeliness issues. 
Euro NCAP obtains some vehicles directly from the manufacturers prior 
to distribution to dealerships. This enables them to begin testing 
before the vehicles are available to the public.[Footnote 61] In 
addition, the Euro NCAP divides its program into two testing and 
information releases each year--one in November and one in June--to 
speed the information to the public.

Conclusions:

While NHTSA's New Car Assessment Program has contributed to making 
safer vehicles, it is at a crossroads where it will need to change to 
remain relevant. The usefulness of the current testing has been eroded 
by changes in the vehicle fleet that have occurred since the program 
began. The growing number of large pickups, minivans, and SUVs in the 
nation's vehicle fleet is creating different safety risks, particularly 
with regard to the incompatibility of large and small vehicles and 
vehicle rollover, which NCAP does not fully address. In addition, the 
very success of the program has brought it to a point where it is not 
clear that the program's goals can continue to be met. Because almost 
all vehicles today receive four-and five-star frontal and side-impact 
safety ratings, NCAP provides little incentive for manufacturers to 
further improve the safety of their vehicles and does not provide 
consumers with information that differentiates the safety of one 
vehicle compared to another. Further, the planned changes to the safety 
standards for frontal and side crashworthiness may make current NCAP 
tests less meaningful.

While we believe there are opportunities to enhance NCAP by developing 
approaches to better measure the interaction of large and small 
vehicles and occupant protection in rollovers, rating technologies that 
help prevent crashes from occurring, and using different injury 
measures to rate the crash results, there are challenges that must be 
considered and addressed before changes can be implemented. However, 
without changing its testing, NCAP provides little incentive for 
manufacturers to improve vehicle safety. In addition, NHTSA will need 
to enhance the timeliness of testing and presentation of the New Car 
Assessment Program information. For example, by the time NHTSA finished 
its testing and published the test results for model year 2004 
vehicles, about 7.7 million, or over 46 percent of new vehicles had 
already been purchased. To enhance the information available to 
consumers, NHTSA can provide summary ratings, present information in a 
comparative manner, increase public awareness, and conduct tests 
earlier in the car model year. Given the substantial numbers of traffic 
deaths and injuries suffered on the nation's roads each year, efforts 
to improve vehicle safety seem warranted.

Recommendations for Executive Action:

We recommend that the Secretary of Transportation direct the 
Administrator, National Highway Traffic Safety Administration, to 
examine the future direction of the New Car Assessment Program to 
maximize its value in providing an incentive for manufacturers to 
improve vehicle safety and informing the public about the relative 
safety of vehicles.

This examination should include:

* identifying and evaluating NCAP tests that should help prevent 
fatalities on the nation's roadways, which should include developing 
measures for rating vehicle incompatibility in front and side-impact 
tests and occupant protection in rollover crashes;

* developing approaches to incorporate active safety systems ratings as 
a part of NCAP; and:

* analyzing alternative testing methodologies and dummies to provide a 
robust and accurate measure of the likelihood of serious injuries to a 
wide range of vehicle occupants.

In addition, we recommend that steps be taken to provide the public 
with improved NCAP safety information in a more timely manner. In doing 
so it may be necessary to examine how other organizations inform the 
public and develop summary ratings, whether vehicles could be obtained 
more efficiently for testing, how budgeted funds are managed during the 
year, and how efficiently NCAP times the crash tests conducted by its 
contractors.

Agency Comments and Our Evaluation:

We provided a copy of the draft report to the Department of 
Transportation for its review and comment. In commenting on the report, 
the Senior Associate Administrator for Vehicle Safety commented that 
NHTSA was pleased that the report concluded that NCAP has been 
successful in encouraging manufacturers to make safer vehicles and 
providing vehicle safety information to consumers. While NHTSA 
generally agreed with the report findings, including recognition that 
there are opportunities to enhance NCAP, the official emphasized that 
NCAP was just one of the many interrelated methods, including Federal 
Motor Vehicle Safety Standards and traffic injury control programs, the 
agency uses to achieve its mission of saving lives, preventing 
injuries, and reducing vehicle-related crashes. The official said that 
NHTSA has been consistently working to address the challenges 
associated with enhancing this complex technical program while ensuring 
that the testing and results reported to consumers are accurate and 
reliable. The official explained that this requires NHTSA to ensure 
that any changes to NCAP, or for that matter to the Federal Motor 
Vehicle Safety Standards, are based on sound science and careful 
analysis of supporting data. The official cited a number of recent 
efforts that NHTSA said demonstrate the careful and systematic approach 
the agency uses when considering changes to the program. These include 
pilot studies with child restraint systems to determine the feasibility 
of incorporating them into NCAP, seeking public comments for revising 
frontal NCAP collision testing, and working to ensure that advanced 
safety technologies are publicized so that consumers can factor them 
into the vehicle purchase decision-making process.

The NHTSA official also said that the agency recognizes that vehicle 
rollover and compatibility issues cause a significant portion of the 
fatal and serious motor vehicle occupant injuries on our nation's 
highways, and NHTSA has made these areas two of its highest priorities. 
In June 2003, NHTSA published initiatives for public comment to address 
both of these areas. The NHTSA official said the agency is continuing 
its efforts to identify effective vehicle metrics and countermeasures 
to address these issues, since they are necessary in order for NCAP to 
provide meaningful consumer information that can be linked to safety 
improvements in the vehicle.

We recognize that NCAP is one of a number of efforts that NHTSA uses in 
an attempt to reduce highway crashes, serious injuries, and fatalities. 
In addition, we support NHTSA's view that changes to the NCAP program 
should be based on sound science and careful analysis of supporting 
data. We encourage NHTSA to take timely action to address the issues 
raised in this report. NCAP has helped make vehicles safer, but there 
are opportunities to improve the program and ultimately help save more 
lives. The risks associated with vehicle incompatibility and rollover 
and the potential benefits to be gained from active safety systems 
heighten the importance of addressing these issues as promptly as 
possible. In addition, analyzing alternative testing methodologies and 
dummies could lead to more robust and accurate measures of the 
likelihood of serious injury to a wide range of vehicle occupants. 
Lastly, NHTSA has the opportunity to improve the timeliness and 
presentation of the NCAP results, which could help consumers make 
informed decisions when they purchase cars.

NHTSA also provided technical clarifications to our report, which we 
incorporated as appropriate.

We are sending copies of this report to appropriate congressional 
committees and the Secretary of Transportation. We will also make 
copies available to others upon request. In addition, this report will 
be available at no charge on GAO's Web site at [Hyperlink, 
http://www.gao.gov]. We are also making available a version of this 
report that includes video clips of some of the crash tests conducted 
by NHTSA and others.

If you or your staffs have any questions regarding the contents of this 
report, please contact me at (202) 512-2834 or [Hyperlink, 
heckerj@gao.gov]. Individuals making key contributions to this report 
are listed in appendix IX.

Signed by:

JayEtta Z. Hecker:
Director, Physical Infrastructure Issues:

[End of section]

Appendixes:

Appendix I: Scope and Methodology:

To determine how NHTSA's New Car Assessment Program tests vehicles, 
rates their safety, and reports the results to the public, we reviewed 
Federal Motor Vehicle Safety Standards (CFR Title 49: Chapter V, Part 
571); the Motor Vehicle Information and Cost Savings Act of 1972 
(Public Law 92-513); the Transportation, Recall Enhancement, 
Accountability and Documentation (TREAD) Act; and other documents 
pertaining to NCAP regulations. We also searched NHTSA's docket and 
NCAP documentation. In addition, we conducted interviews with NHTSA 
officials responsible for operating the Federal Motor Vehicle Safety 
Standards regulatory program and the New Car Assessment Program. We 
visited and interviewed officials from the Federal Highway 
Administration and the National Crash Analysis Center. During visits to 
all five of the contractors that perform regulatory and NCAP crash 
tests--including Karco Engineering, LLC, in Adelanto, California; MGA 
Research Corporation in Burlington, Wisconsin; Medical College of 
Wisconsin in Milwaukee, Wisconsin; General Dynamics--Advanced 
Information Systems in Buffalo, New York; and the Transportation 
Research Center, Inc., in East Liberty, Ohio--we interviewed officials 
and engineers performing tests and observed various crash tests. We 
documented the procedures for obtaining the data, how results were 
recorded, and the conversion into star ratings. We determined that NCAP 
data were sufficiently reliable for the purpose of this report. In 
addition, we reviewed literature pertaining to vehicle safety issues 
and documents published by the Transportation Research Board.

To compare NHTSA's New Car Assessment Program with other programs that 
test vehicles and report vehicle safety results to the public, we 
researched literature and interviewed NHTSA officials to identify three 
foreign New Car Assessment Programs (in Australia, Europe, and Japan) 
and the Insurance Institute for Highway Safety as a domestic program. 
We also identified publishers of Consumer Reports and The Car Book as 
organizations that used NHTSA's NCAP data to derive their own vehicle 
safety ratings. We identified a program in Korea but did not include 
this program in our review because it began operating in 1999 and had 
not tested a significant number of vehicles.

We obtained information on these programs by reviewing their literature 
and their Web sites. We also interviewed officials and visited the test 
facilities of the Insurance Institute and the NCAPs in Australia, 
Europe, and Japan. We visited the Insurance Institute of Highway 
Safety's Vehicle Research Center and observed a crash test. We also 
examined international crash test and rating programs, including the 
Australia, Euro, and Japan NCAPs. For Australia's NCAP, we visited 
Australia and conducted interviews with government officials associated 
with the respective New Car Assessment Program and vehicle safety 
policy. For Euro NCAP, we visited Belgium, Germany, Sweden, and the 
United Kingdom, where we conducted interviews with the European 
Commission, and the government officials associated with the respective 
New Car Assessment Programs and vehicle safety policies in Germany, 
Sweden, and the United Kingdom. For Japan's NCAP, we visited Japan and 
interviewed government officials associated with the respective New Car 
Assessment Program and vehicle safety policy. While in these countries, 
we also interviewed auto associations, consumer advocacy groups, and 
vehicle safety experts. We identified and selected these auto 
associations, consumer advocacy groups, and vehicle safety experts by 
reviewing studies and conference papers, talking to program officials 
and other experts, and reviewing materials on Web sites. We interviewed 
auto manufacturers in these countries, including BMW, Honda, Mercedes, 
Nissan, Toyota, and Volvo. We reviewed New Car Assessment Program 
regulations, testing protocols, and program documentation. See table 1 
for a list of domestic and international organizations contacted.

To determine whether opportunities exist for NCAP to enhance its 
vehicle safety testing and reporting, we obtained views from experts in 
vehicle safety and the auto and insurance industries. In selecting 
vehicle safety experts, we examined studies and conference papers, 
considered referrals from other experts, and consulted the National 
Academy of Sciences. We interviewed officials of the Association for 
the Advancement of Automotive Medicine and Applied Research Associates. 
We visited and interviewed automobile manufacturers in the United 
States, including General Motors, Ford Motor Company, Daimler-Chrysler, 
and American Honda Motor Company. We interviewed trade associations 
including the Alliance of Automobile Manufacturers and the Association 
of International Automobile Manufacturers. We interviewed consumer 
advocacy groups, including Consumers Union, Public Citizen, the AAA 
Foundation for Traffic Safety, Advocates for Highway Safety and Auto 
Safety, and the National Safety Council. We reviewed relevant research 
on consumer information regarding vehicle safety from the 
Transportation Research Board.

Table 1: List of Organizations Contacted:

U.S. NCAP.

Federal agencies: 
Department of Transportation:
NHTSA.

Federal agencies: 
Department of Transportation:
George Washington University's National Crash Analysis Center.

Federal agencies: 
Department of Transportation:
Federal Highway Administration.

Consumer Information Organizations:
Consumer Reports:

Consumer Information Organizations:
The Center for Auto Safety (The Car Book).

Crash Test Organizations:
Insurance Institute for Highway Safety.

Automobile Manufacturers:
American Honda Motor Company.

Automobile Manufacturers:
Daimler-Chrysler Corporation.

Automobile Manufacturers:
Ford Motor Company.

Automobile Manufacturers:
General Motors Corporation.

Industry associations and experts:
AAA Foundation for Traffic Safety.

Industry associations and experts:
Advocates for Highway and Auto Safety.

Industry associations and experts:
Alliance of Automobile Manufacturers.

Industry associations and experts:
American Association of State Highway and Transportation Officials.

Industry associations and experts:
Applied Research Associates.

Industry associations and experts:
Association for the Advancement of Automotive Medicine.

Industry associations and experts:
Association of International Automobile Manufacturers.

Industry associations and experts:
Consumers Union.

Industry associations and experts:
Japan Automobile Standards Internationalization Center.

Industry associations and experts:
National Safety Council.

Industry associations and experts:
Public Citizen.

Industry associations and experts:
Transportation Research Board.

Industry associations and experts:
Vehicle Certification Agency North America.

Contractors:
General Dynamics--Advanced Information Systems.

Contractors:
Karco Engineering, LLC.

Contractors:
Medical College of Wisconsin.

Contractors:
MGA Research Corporation.

Contractors:
Transportation Research Center.

Australia NCAP:
Australia Automobile Association (administrator for Australia NCAP).

Australia NCAP:
Australian Consumers' Association (CHOICE magazine).

Australia NCAP:
Department of Transport and Regional Services--Vehicle Safety Standards.

Australia NCAP:
Crashlab--New South Wales Road and Traffic Authority (Australia NCAP 
test facility).

Australia NCAP:
Federal Chamber of Automotive Industries.

Australia NCAP:
Monash University Accident Research Center.

Australia NCAP:
National Roads and Motorists Association Motoring and Services.

Australia NCAP:
Royal Automobile Club of Victoria.

Australia NCAP:
Victoria Road and Traffic Authority.

Belgium:
European Auto Manufacturers Association.

Belgium:
European Commission, Directorate General for Energy and Transport.

Belgium:
European Transport Safety Council.

Germany:
General German Automobile Club (ADAC).

Germany:
BMW AG.

Germany:
Daimler Chrysler AG, Mercedes-Benz Technology Center.

Germany:
Federal Ministry for Traffic, Building, and Housing.

Germany:
TÜV Automotive Group.

United Kingdom:
Automotive Safety Centre at the University of Birmingham.

United Kingdom:
Department for Transport.

United Kingdom:
Euro NCAP.

United Kingdom:
International Consumer Research and Testing (Which? magazine).

United Kingdom:
Thatcham (Motor Insurance Repair Research Centre).

United Kingdom:
Transportation Research Laboratory.

Sweden:
Swedish Road Administration.

Sweden:
Swedish National Road and Transport Research Institute.

Sweden:
Volvo.

Japan NCAP:
Honda Research and Development Corporation.

Japan NCAP:
Japan Auto Research Institute (Japan NCAP test facility).

Japan NCAP:
Japan Automobile Federation.

Japan NCAP:
Japan Automobile Importers Association.

Japan NCAP:
Japan Automobile Manufacturers Association.

Japan NCAP:
Japan Ministry of Land Infrastructure and Transport.

Japan NCAP:
Mitsubishi Motors Corporation.

Japan NCAP:
National Agency for Automotive Safety and Victim's Aid (Japan NCAP).

Japan NCAP:
Nissan Motor Corporation.

Japan NCAP:
Toyota Motor Corporation.

Source: GAO.

[End of table]

We conducted our work from March 2004 through April 2005 in accordance 
with generally accepted government auditing standards.

[End of section]

Appendix II: Basis for NHTSA's Crashworthiness Ratings:

To rate a vehicle's crashworthiness, NHTSA combines information about 
(1) the forces that would injure a human during a crash and (2) the 
effects of those forces on areas of the human body. The forces that 
would injure a human during a crash are measured by anthropomorphic 
test devices, commonly referred to as crash test dummies, which serve 
as proxies for human vehicle occupants. These dummies are fitted with 
accelerometers and load sensors that measure the forces of impact on 
particular areas of the body, as shown in figure 32.

Figure 32: A Crash Test Dummy Fitted with Load Sensors and 
Accelerometers:

[See PDF for image]

[End of figure]

Because the current dummy technology has yet to replicate a human with 
the same biological matter or physiology, dummies cannot exhibit 
injuries following a crash as a human would. Therefore, the effects of 
the forces on particular areas of the human body, as measured by the 
dummies, have been developed by researchers who have gathered 
information by applying varying forces to biological specimens and by 
using a scale developed by the Association for the Advancement of 
Automotive Medicine (AAAM). This scale, the Abbreviated Injury Scale 
(AIS), ranks injuries, from minor through currently untreatable, for 
particular areas of the body and assigns a number from 1 through 6 to 
each rank, as shown in table 2. The AIS is used to provide a simple 
numerical method for ranking and comparing injuries by severity.

Table 2: Abbreviated Injury Scale (AIS):

AIS code: 1;
Description of injury: Minor.

AIS code: 2;
Description of injury: Moderate.

AIS code: 3;
Description of injury: Serious.

AIS code: 4;
Description of injury: Severe.

AIS code: 5;
Description of injury: Critical.

AIS code: 6;
Description of injury: Currently untreatable.

Source: Copyright © 2005 Association for the Advancement of Automotive 
Medicine, reprinted with permission. All rights reserved.

[End of table]

AIS values in NCAP are injury probability values derived from 
measurements of dummy response taken from specific characteristics 
(e.g., size, shape, mass, stiffness, or energy dissipation) that 
simulate corresponding human responses (e.g., acceleration, velocity, 
or articulation). These dummy responses are correlated with both 
experimental biomechanical research as well as with real world crash 
injury investigation. Researchers have used a statistical procedure to 
relate the levels of injury to the forces that caused them. This 
procedure produces theoretical injury curves, which NHTSA uses as the 
basis for safety ratings.

NHTSA develops crashworthiness ratings, expressed in stars, for both 
frontal and side crashes. To develop the NCAP ratings for frontal 
crashes, NHTSA measures forces to the head and chest. Specifically, the 
injury criteria for the frontal star rating are the head, as measured 
by a composite of acceleration values known as the Head Injury 
Criterion (HIC), and the chest, as measured by a chest deceleration 
value known as chest Gs. Each of these two measures has its own injury 
risk curve that has been fixed at AIS level 4 or greater--that is, a 
severe, critical, or currently untreatable injury, as shown in figures 
33 and 34. Using the mathematical functions that describe each of these 
injury risk curves, NHTSA transforms the HIC and chest G measures from 
the frontal NCAP test into probabilities of head and chest injuries of 
AIS level 4 or greater. The lower the HIC and chest G measures, the 
less risk of receiving a severe, critical, or currently untreatable 
injury to the head and chest in a full frontal crash.

Figure 33: Injury Curve for HIC:

[See PDF for image]

[End of figure]

Figure 34: Injury Curve for Chest G:

[See PDF for image]

[End of figure]

To convert the probability of severe injury for particular HIC and 
Chest G scores into a star rating for the frontal NCAP test, NHTSA adds 
the probability of severe injury to the head and chest and then 
subtracts the product, shown below in figure 35. NHTSA concluded that a 
combined effect of injury to the head and chest should be used since it 
is well documented that an individual who suffers multiple injuries has 
a higher risk of death. NHTSA calculates the probability of severe 
injury to the head and chest for both the driver and the front 
passenger dummies in the frontal NCAP test.

Figure 35: Probability Equation Used to Produce Star Ratings for 
Frontal Crashes:

[See PDF for image]

[End of figure]

To develop the NCAP ratings for side crashes, NHTSA measures forces to 
the ribs and lower spine. Specifically, the injury criteria for the 
side star rating are the greater acceleration of the upper or lower 
ribs and the acceleration of the lower spine. NHTSA averages these 
accelerations to generate a measurement known as the Thoracic Trauma 
Index (TTI). The TTI also has an injury curve that has been fixed at 
the AIS level of 4 or greater, as shown in figure 36. The lower the TTI 
measure, the lower the risk of receiving a severe, critical, or 
currently untreatable injury to the thorax and upper abdomen in a side 
crash.

Figure 36: Thoracic Trauma Index Curve:

[See PDF for image]

[End of figure]

The Thoracic Trauma Index score and its associated probability of 
receiving an AIS level 4 or greater injury is the sole basis for the 
side NCAP star rating. NHTSA calculates probability of severe injury to 
the thorax and upper abdomen for both the front and rear dummies on the 
driver's side.

Using the probability of injury calculated from the frontal and side 
NCAP tests, NTHSA assigns a vehicle a rating of one (the worst) to five 
(the best) stars for each of the dummy occupants in each of the 
crashworthiness tests. The star ratings for the frontal and side tests 
correspond to the percentage chance of serious injury for each of these 
tests.

The numerical boundaries between each star rating are determined by 
NHTSA. The frontal NCAP star boundaries are roughly twice as large as 
the side NCAP star boundaries because NHTSA uses a combined probability 
of injury to generate star ratings for the frontal NCAP test and uses 
only one probability of injury to generate star ratings for the side 
NCAP test. In addition, the forces and associated probabilities at the 
boundary between two and three stars for both the frontal and side NCAP 
tests are roughly equal to the relevant force thresholds for compliance 
with two Federal Motor Vehicle Safety Standards--numbers 208 and 214, 
respectively.

[End of section]

Appendix III: Basis for NHTSA's Rollover Rating:

To indicate the likelihood of a vehicle's rolling over in a single-
vehicle crash, NHTSA combines the risk of rollover in a single vehicle 
crash indicated by a measure of the vehicle's top-heaviness, called the 
Static Stability Factor (SSF), with the results of a dynamic rollover 
test to produce a star rollover rating. The SSF is an indicator for the 
most frequent type of rollover, called a "tripped rollover," which 
occurs when a vehicle leaves the roadway and its wheels are tripped by 
a curb, soft shoulder, or other roadway object, causing the vehicle to 
roll over. About 95 percent of rollovers are tripped. Because the SSF 
is an indicator of the most frequent type of rollover, it plays a 
significantly larger role in a vehicle's star rating than do the 
results of the dynamic rollover test. The dynamic rollover test 
determines how susceptible a vehicle is to an on-road "untripped" 
rollover--a type that accounts for less than 5 percent of rollovers. 
Because untripped rollovers are so infrequent, the rollover test does 
not affect the vehicle's star rating significantly, resulting in a 
difference of no more than half a star in the rating.

Static Stability Factor:

The SSF is a calculation of a vehicle's top-heaviness, defined as one-
half of the vehicle's track width divided by the height of the center 
of gravity (c.g.) A higher SSF value equates to a more stable, less top-
heavy vehicle. SSF values across all vehicle types range from around 
1.0 to 1.5. Most passenger cars have values in the 1.3 to 1.5 range, as 
shown in figure 37. Higher riding SUVs, pickups, and vans usually have 
values in the 1.0 to 1.3 range, also shown in figure 37. Many of the 
higher riding vehicles of previous model years are being redesigned to 
ride lower on a wider track to improve their rollover resistance and 
obtain a higher SSF rating.

Figure 37: Static Stability Factor of a Passenger Vehicle Compared with 
an SUV:

[See PDF for image]

[End of figure]

After determining the SSF, NHTSA selects certain vehicles for the 
dynamic rollover test. Not all passenger cars selected for NCAP testing 
undergo the dynamic test. Thus far, for most passenger cars, NHTSA has 
imputed or assigned a no-tip result for the dynamic test based on the 
testing of other passenger cars that are more top heavy (according to 
the SSF score) but did not tip up during the dynamic test. NHTSA 
periodically tests passenger cars to validate the imputed results.

Dynamic Rollover Test:

In the dynamic rollover test, a driver sits in the vehicle and conducts 
the test by applying the accelerator and initiating commands for the 
programmable steering controller, which actually maneuvers the 
vehicles, as shown in figure 38. The general steering parameters are 
270 degrees (about a three-quarters turn) for the initial turn and 540 
degrees (about one and one-half turn) for the correction turn, as shown 
in figure 39. Outriggers are attached to the vehicle to prevent the 
vehicle from tipping all the way over.

Figure 38: Programmable Steering Controller:

[See PDF for image]

[End of figure]

Figure 39: Diagram of the Dynamic Test, Showing Steering Parameters:

[See PDF for image]

[End of figure]

The result of the dynamic rollover test is either "tip-up" or "no tip-
up." To receive a "no tip-up" result, a vehicle must reach a speed of 
50 miles per hour (mph) on four dynamic test runs--two from left to 
right and two from right to left--without the inside wheels on either 
side of the vehicle simultaneously lifting at least 2 inches off the 
surface, and it must do this at two different steering wheel angles. 
Sensors are used to detect wheel-lift, as shown in figure 40. For the 
first run of each test, the speed is 35 mph, and subsequent runs are 
conducted at about 40 mph, 45 mph, 47.5 mph, and 50 mph, until the 
vehicle tips up or attains an entrance speed of 50 mph on the last run 
of each test without tipping up. The same series of tests is repeated 
at a different steering wheel angle.

Figure 40: Sensors to Detect Wheel-lift:

[See PDF for image]

[End of figure]

NHTSA first began to rate vehicles' rollover avoidance in model year 
2001, using the SSF alone to determine the star rating. At that time, 
NHTSA used a statistical procedure to determine how the SSF affects the 
risk of rollover.[Footnote 62] Physics theory would suggest that 
vehicles with a low SSF--vehicles that are more top-heavy--are more 
likely to roll over than vehicles with a high SSF. NHTSA's empirical 
model confirmed this theory, showing that the lower the SSF, the more 
likely a vehicle is to roll over in a single-vehicle crash. For the 
first 3 years that NHTSA rated rollover risk, it used a linear model 
that examined accident report data at the state level. Following the 
passage of the TREAD Act, which required NHTSA to include a dynamic 
rollover test in NCAP, and the publication of a National Academy of 
Sciences report, which recommended that NHTSA use a nonlinear model to 
predict rollover risk, NHTSA altered its method of calculating rollover 
risk.[Footnote 63] NHTSA now links the SSF and the risk of rollover 
using a nonlinear model.[Footnote 64] In addition, NHTSA includes the 
results of the dynamic test--that is, whether a vehicle tips or not--in 
this new model, as shown in figure 41.

Figure 41: NCAP Logistic Model Used to Determine Rollover Ratings:

[See PDF for image]

[End of figure]

A vehicle's rollover rating is an estimate of its risk of rolling over 
in a single-vehicle crash, not a prediction of the likelihood of a 
rollover crash.

[End of section]

Appendix IV: Insurance Institute for Highway Safety:

The Insurance Institute for Highway Safety is a nonprofit research and 
communications organization funded by the U.S. auto insurance industry. 
The Insurance Institute has been conducting vehicle safety research 
since 1969, and in 1992 it opened the Vehicle Research Center to 
conduct vehicle crash tests. The Insurance Institute began crash 
testing and rating vehicles for frontal collisions in 1995 and for side 
collisions in 2003.[Footnote 65] The center conducts the Insurance 
Institute's vehicle-related research, which includes controlled tests 
of vehicles and their components using instrumented crash tests, as 
well as studies of real collisions. Insurance Institute officials told 
us that scrutinizing the outcomes of both controlled tests and on-the-
road crashes gives researchers--and ultimately the public--a better 
idea of how and why vehicle occupants are injured in crashes. This 
research, in turn, leads to vehicle designs that reduce injuries. The 
Insurance Institute buys the vehicles for crash tests directly from 
dealers. It also chooses vehicles for testing to represent both a range 
of manufacturers and the largest portions of new car sales, in an 
effort to cover as much of the marketplace as possible. The Insurance 
Institute tests vehicles in categories, such as small cars, minivans, 
and midsize SUVs.

Testing Conducted:

The Insurance Institute conducts two types of crash tests--an offset 
frontal test and a perpendicular side test. The offset frontal test is 
conducted at about 40 mph to simulate a typical head-on collision of 
two vehicles. The offset frontal test evaluates the potential for 
injuries caused to occupants by intrusion into the occupant 
compartment. The Insurance Institute uses a frontal impact dummy, 
called the 50THpercentile Hybrid III dummy, in its frontal crash tests. 
This dummy represents a man of average size, 5 feet 9 inches tall and 
weighing about 170 pounds. Such dummies were designed to measure the 
risk of injury to the head, neck, chest, and lower extremities in a 
frontal crash.

The Insurance Institute's perpendicular side test measures the impact 
of a moving deformable barrier striking the driver's side of a 
passenger vehicle at 31 mph. The barrier weighs 3,300 pounds and has a 
front end shaped to simulate the typical front end of a pickup truck or 
SUV. Two instrumented 5TH percentile side-impact dummies (SID-IIs), 
representing small females or 12-year-old adolescents who are 5 feet 
tall and weigh about 110 pounds, are positioned in the driver's seat 
and in the rear seat behind the driver to measure the impact of the 
vehicle crash. The SID-IIs dummies were designed to measure 
acceleration of the spine and ribs plus compression of the rib cage in 
a side crash.[Footnote 66] They are also equipped with unique load 
cells, which measure the force of the impact applied to the dummies 
during the crash.

Scoring Tests:

To evaluate a vehicle's performance in the frontal crash test and 
develop an overall rating for the frontal test, the Insurance Institute 
uses three types of measures: (1) structural performance, the amount 
and pattern of intrusion into the occupant compartment during the 
offset test; (2) injuries measured by a Hybrid III dummy positioned in 
the driver's seat; and (3) dummy kinematics, or the dummy's movements 
during the test, as determined through an analysis of a slow-motion 
film. The structural performance assessment indicates how well the 
front-end crush zone managed the crash energy and how well the safety 
cage limited intrusion into the driver space. Figure 42 shows the 
intrusion levels on which a vehicle's structural performance is rated. 
Injury measures are used to determine the likelihood of injury to 
various regions of the driver's body. The measures recorded from the 
head, neck, chest, legs, and feet of the dummy indicate the level of 
stress/strain on that part of the body. Thus, greater numbers mean 
larger stresses and strains and a greater risk of injury. Because 
significant risk of injury can result from undesirable dummy kinematics 
in the absence of high injury measures, such as partial ejection from 
the occupant compartment through a window, a slow-motion film is used 
during the crash test. An analysis of this slow-motion film helps 
evaluate the interactions of the restraint system's components--
including the safety belts, air bags, steering columns, head 
restraints, and other components--to control the dummy's movement.

Figure 42: Guidelines for Rating Occupant Compartment Intrusion 
Measured in Centimeters:

[See PDF for image]

[End of figure]

A vehicle's overall frontal rating depends on the effectiveness of its 
structure, or safety cage, in protecting the occupant compartment, the 
risk of injury measured for an average-size male, and the effectiveness 
of the restraint system in controlling occupants' movements. The 
structural performance and injury assessments are the major components 
of each vehicle's overall frontal rating; the dummy kinematics 
(movement) contributes less to the rating.

A vehicle's side crash test performance and overall rating are based on 
(1) the injury measures recorded on the two instrumented SID-IIs 
dummies positioned in the driver's seat and in the rear seat behind the 
driver, (2) an assessment of head-protection countermeasures, and (3) 
the vehicle's structural performance during the impact. The injury 
measures are used to determine the likelihood that the driver, the 
passenger, or both would have sustained serious injury to various body 
regions. Measures are recorded from the head, neck, chest, abdomen, 
pelvis, and leg. These injury measures, especially from the head and 
neck and from the torso (chest and abdomen), are the major components 
of the vehicle's overall rating. To supplement head injury measures, 
the movements and contacts of the dummies' heads during the crash are 
evaluated. High head injury measures typically are recorded when the 
moving deformable barrier hits a dummy's head during impact. Moreover, 
a "near miss" or a grazing contact also indicates a potential for 
serious injury in a real-world crash because small differences in an 
occupant's height or seating position, compared with a dummy's, could 
result in a hard contact and high risk of serious head injury. The 
vehicle's structural performance is based on measurements of intrusion 
into the occupant compartment around the B-pillar (between the doors). 
This assessment indicates how well the vehicle's side structure 
resisted intrusion into the driver's and rear-seat passenger space. 
Some intrusion into the occupant compartment is inevitable in serious 
side crashes.

The overall side rating depends on the risk of injury measured for 
small female occupants mainly to the head and neck and torso (chest and 
abdomen); the effectiveness of the occupant compartment in protecting 
the head; and the vehicle's structure performance during the impact. 
The overall side rating for any body region, based on the injury 
measures recorded on the two SID-IIs dummies, is the lowest rating 
scored for any injury within that region.

Sharing Results with the Public:

The Insurance Institute's rating system provides qualitative ratings of 
Good, Acceptable, Marginal, and Poor. The Insurance Institute provides 
one rating for the frontal test and one rating for the side test. 
Vehicle rating information is available on the Insurance Institute's 
Web site, through press releases, and through television coverage. 
Figure 43 shows how the Insurance Institute communicated its ratings to 
consumers on the Internet. In addition to the ratings for frontal and 
side crashes, the Insurance Institute provided the results of various 
tests, such as those of the vehicle's structural performance and of 
injuries to various body regions.

Figure 43: Insurance Institute Rating of a Mid-size Passenger Car:

[See PDF for image]

[End of figure]

Figure 44 shows how the Insurance Institute presented its ratings to 
consumers in its Status Report. The print version is available only to 
subscribers, and some of the publications can be downloaded from the 
Insurance Institute's Web site.

Figure 44: Insurance Institute Crash Ratings Provided to Public in its 
Status Report Publication:

[See PDF for image]

Note: The names of vehicles that received the crash ratings above were 
purposely removed.

[End of figure]

News magazine television shows, such as Dateline NBC, periodically use 
Insurance Institute crash test results and interview representatives, 
including the president or chief operating officer, as report segments 
for their programs.

[End of section]

Appendix V: Australian New Car Assessment Program:

The Australian New Car Assessment Program (NCAP) provides information 
for consumers on the safety performance of new vehicles sold in 
Australia and New Zealand. The main purposes of the program are to 
provide new vehicle buyers with independent advice on vehicle occupant 
protection and to develop strategies for vehicle manufacturers to 
increase the level of passive safety in their vehicles.[Footnote 67] 
The program is funded by a consortium of the state government transport 
departments of New South Wales, Queensland, Victoria, South Australia, 
Tasmania, and Western Australia; automobile clubs through the 
Australian Automobile Association and New Zealand Automobile 
Association; the Land Transport Safety Authority of New Zealand; and 
the FIA Foundation for the Automobile and Society.[Footnote 68] The 
Australia Commonwealth Department of Transport and Regional Services 
has established minimum safety standards for vehicles sold in Australia 
and has conducted joint research projects with NCAP but has not 
contributed to the support of the crash test program.

The Australia NCAP buys the vehicles that it crash tests directly from 
dealers, as would any consumer. The program selects vehicles on the 
basis of (1) actual or projected sales, to target vehicles that are 
most popular; (2) vehicle model, to account for standard or deluxe 
models, which may contain more expensive passive safety features such 
as air bags and advanced restraint systems; (3) new and popular body 
designs, to select the body design that is most popular or to allow for 
direct comparisons across different makes and models; (4) market 
segment, to target individual segments of the market to allow 
comparisons of results; and (5) vehicle price. Using these selection 
criteria, the Australia NCAP covers more than 70 percent of the new 
vehicle fleet by volume. The program also uses European NCAP (Euro 
NCAP) crash test results. However, the Euro NCAP results are intended 
to be used as a guide only, because the structure and equipment of the 
European specification model may differ materially from the model of 
the same name sold in Australia or New Zealand. The Australia NCAP 
tests and reports on vehicles in seven categories--small, medium, and 
large passenger cars; luxury cars; four-wheel drive vehicles (SUVs); 
multipurpose utility vehicles (small trucks); and sports cars.

Testing Conducted:

The Australia NCAP's testing has evolved over time. Established in 
1992, the Australia NCAP was originally modeled on the U.S. program and 
began rating vehicles in 1993. Initially, it conducted only a full 
frontal crash test, but it added an offset frontal test in 1994. In 
1999, the Australia NCAP harmonized its tests and assessment procedures 
with the Euro NCAP through a memorandum of understanding. By 
harmonizing, it discontinued the full frontal crash test and began 
conducting the perpendicular side-impact test and pedestrian test. 
Australia NCAP officials have been considering eliminating the 
perpendicular side-impact test in favor of a pole test that they 
believe will more accurately test vehicles of all sizes for occupant 
protection.

In 2004, the Australia NCAP performed three crash tests and a 
pedestrian protection test. The three crash tests include the 40 
percent offset frontal, the perpendicular side-impact, and the side-
impact pole tests. The offset frontal test involves pulling a test 
vehicle traveling at 40 mph (64 km/h) and crashing it into an offset 
deformable aluminum barrier. The deformable barrier has a crushable 
aluminum honeycomb face attached to a solid barrier. The deformable 
structure resembles the front-end characteristics of another vehicle. 
Two instrumented 50TH percentile Hybrid III dummies (weighing about 194 
pounds each) are used to collect data during the crash and are placed 
in the front driver's and front passenger seats. Two child dummies, 
representing a 3-year-old and a 1-1/2-year-old child, are placed in the 
rear seats in appropriate restraints. While Australia NCAP does not use 
the measurements from the child dummies in its crash test rating, the 
dummies are included in the tests to maintain alignment with Euro NCAP 
testing.

The perpendicular side-impact test involves pulling a barrier with a 
deformable face at about 31 mph (50 km/h) and crashing it into a 
stationary test vehicle at a 90 degree angle centered on the driver's 
seating position. The moving deformable barrier has a mass of 2,095 
pounds (950 kg) compared with 3,015 pounds (1,367 kg) for the U.S. 
barrier. One instrumented 50TH percentile EuroSID-II dummy (weighing 
about 176 pounds) is used to collect data during the crash and is 
placed in the front driver seat. As in the frontal test, to maintain 
alignment with Euro NCAP's testing, the two child dummies are placed in 
the rear seats in appropriate restraints.

The pole side-impact test involves propelling a vehicle placed on a 
platform at 18 mph (29 km/h) into a cylindrical pole. The pole has a 
diameter of about 10 inches, or about 254 millimeters (mm), and its 
vertical axis is aligned with the front seat dummy's head. One 
instrumented 50TH percentile EuroSID-II dummy is used to collect data 
during the crash and is placed in the front driver's seat.

The pedestrian protection test evaluates the interaction of dummy parts 
and the bumper, hood, and windshield area of a vehicle. Adult and child-
size dummy parts are propelled at specified areas of the hood and front 
bumper of a vehicle to simulate a 25 mph (40 km/h) car-to-pedestrian 
collision. The test simulates the impact of a lower leg against a 
bumper, a thigh against the lower edge of the hood, and an adult and a 
child head against the upper portion of the hood.

Scoring Tests:

Frontal tests in the Australia NCAP are scored on the basis of three 
types of observations--dummy measurements, a vehicle's structural 
performance, and a post-crash inspection of the vehicle. The injury 
measurements are recorded from two Hybrid III dummies positioned in the 
front driver's seat and front passenger seat. The injury assessment 
evaluates four body regions: (1) head and neck; (2) chest; (3) knee, 
femur, and pelvis; and (4) legs and feet. Structural performance is 
based on measurements indicating the amount and pattern of intrusion 
into the occupant compartment during the test. Dummy injury 
measurements and vehicle deformation can be compared with predicted 
values. Evidence of structural collapse can be determined by a post-
crash inspection and by viewing a high-speed video recorded from 
various angles during the crash test. The post-crash inspection and 
video allow trained inspectors to assess dummy kinematics, evaluate the 
evidence of interior contacts, and inspect safety belts, seats, and air 
bags to ensure they operated as intended.[Footnote 69] For example, 
according to Australia NCAP officials, air bag performance could be 
compromised by the dynamics of a crash in ways that might not be 
evident from a post-crash inspection but could be revealed through 
careful analysis of the video.

Each body region receives a score based on the dummy measurements, the 
vehicle deformation data, and the findings of the post-crash inspection 
(using modifiers). For example, excessive rearward movement of the 
steering wheel could lower the head score by a point to reflect 
identified risks. Other modifiers include lack of air bag stability, 
steering column movement, A-pillar movement, structural integrity, 
hazardous structures in the knee impact area, and brake pedal 
movement.[Footnote 70]

For the side-impact and pole tests, the scores are based on injury 
measurements recorded on one EuroSID-II dummy positioned in the front 
driver's seat. The injury assessment evaluates four body regions: the 
head, ribs, abdomen, and pelvis. A post-crash inspection and high-speed 
video are also used to evaluate structural collapse.

A summary star rating shows the protection level indicated by the front 
and side-impact tests together. The summary score for the two tests is 
based on the point scores achieved in each test. Sixteen points can be 
achieved in the frontal test and 18 points in the side tests, for a 
maximum of 34 points. Two of the 18 points available in the side test 
come from the optional pole test, which assesses only one body region-
-the head. Each of the four body regions in the frontal test could 
receive a maximum score of 4 points, for a cumulative score of 16 
points. Similarly, the four body regions in the side-impact test could 
receive a maximum score of 4 points, for a cumulative score of 16 
points. If a vehicle has head-protecting side air bags, the 
manufacturer of the vehicle has the option of accepting a side impact 
pole test, through which 2 bonus points can be earned.

The offset and side-impact scores are added together to produce an 
overall score with a maximum of 32 points. In addition, if a pole side 
test is conducted and shows good head protection, then 2 extra points 
can be earned, and up to 3 more points can be earned for having a 
safety belt reminder system. The points are translated into stars, as 
shown in table 3.

Table 3: Scoring Basis for Australia NCAP Frontal and Side Star Ratings:

Star rating: 1;
Minimum score in offset test: -;
Minimum score in side impact test: -;
Minimum combined score (including pole test): 0.5.

Star rating: 2;
Minimum score in offset test: 1.5;
Minimum score in side impact test: 1.5;
Minimum combined score (including pole test): 8.5.

Star rating: 3;
Minimum score in offset test: 4.5;
Minimum score in side impact test: 4.5;
Minimum combined score (including pole test): 16.5.

Star rating: 4;
Minimum score in offset test: 8.5;
Minimum score in side impact test: 8.5;
Minimum combined score (including pole test): 24.5.

Star rating: 5;
Minimum score in offset test: 12.5;
Minimum score in side impact test: 12.5;
Minimum combined score (including pole test): 32.5[A].

Source: Australian NCAP, Australian Automobile Association.

[A] To earn five stars a vehicle must achieve at least 1 point in the 
optional pole test (maximum 2 points). This is an Australia NCAP 
requirement.

[End of table]

If the injury score for the head, chest, abdomen, or pelvis is 0, then 
there is a high risk of a life-threatening injury. A warning note is 
added to the overall rating to highlight concern that there is a 
serious risk of injury in at least one vulnerable body region. The 
regions are the head or chest for the frontal impact test and the head, 
chest, abdomen, or pelvis for the side-impact test.

For the pedestrian test, the scores are based on adult and child-size 
dummy parts (head and lower limbs) used to assess the severity of 
impact. The two different size dummy heads are tested at six areas of 
the hood, and the lower limbs for an adult and child are tested at 
three areas, for a total of 18 impacts tested for each vehicle. Based 
on the injury measurements recorded from the dummy parts, each impact 
can receive up to 2 points, and the maximum number of points that can 
be received is 36, as shown in table 4.

Table 4: Australia NCAP Point System for the Pedestrian Test:

Type of test: Child head;
Impact locations: 6;
Maximum: Points: 12.

Type of test: Adult head;
Impact locations: 6;
Maximum: Points: 12.

Type of test: Upper leg;
Impact locations: 3;
Maximum: Points: 6.

Type of test: Lower leg;
Impact locations: 3;
Maximum: Points: 6.

Type of test: Total;
Impact locations: 18;
Maximum: Points: 36.

Source: Euro NCAP.

[End of table]

A separate rating of one to four stars shows the level of pedestrian 
protection. The score reflects the results of the 18 impacts of the 
dummy parts against the specified areas of the bumper and hood. These 
results are summed to provide an overall score. The pedestrian 
protection star rating for a vehicle is based on the number of points 
received, or a maximum of 36 points. The points are translated into 
stars, as shown in table 5.

Table 5: Scoring Basis for Australia NCAP Pedestrian Rating:

Star rating: 0;
Points: 0.

Star rating: 1;
Points: 1-9.

Star rating: 2;
Points: 10-18.

Star rating: 3;
Points: 19-27.

Star rating: 4;
Points: 28-36.

Source: Euro NCAP.

[End of table]

Sharing Results with the Public:

The Australia NCAP's reporting of results to the public has evolved 
over time. Initially, the program reported the raw test results for the 
head, chest, and legs. The program also portrayed the risk of injury in 
each area as high (red), medium (yellow), or low (green) and 
graphically represented the risk on an outline of a human figure in 
each area. When the offset frontal crash test was added in 1994, its 
results were reported in the same way. Also in 1994, the program began 
publishing tables comparing the results of the vehicles tested.

In 1995, the Australia NCAP began summarizing full frontal and offset 
frontal head, chest, and leg test results by using bar charts to 
represent the percentage of risk of a life-threatening injury to 
drivers and to passengers. In 1996, the program began differentiating 
between upper and lower leg injuries, reported the results separately, 
and adopted the Insurance Institute for Highway Safety rating scale of 
Good, Acceptable, Marginal, and Poor. However, the program combined the 
scores for the full frontal driver and passenger tests with the score 
for the offset frontal driver test to arrive at an overall vehicle 
rating. According to Australia NCAP officials, subsequent research with 
focus groups supported the decision because the results indicated that 
consumers wanted the safety information in a simplified, summary form. 
In November 1999, to align with the Euro NCAP, the Australia NCAP first 
used a five-star system to report crash test performance. This system 
provided an overall rating along with a bar chart that enabled 
consumers to differentiate between vehicles with different scores that 
received the same number of stars.

Today, the Australia NCAP makes vehicle rating information available on 
its Web site, through press releases, and through a safety brochure. 
Figure 45 shows how the program communicates its overall and pedestrian 
ratings to consumers on the Internet.

Figure 45: Example of Australian NCAP Rating of a 2004 Mid-size 
Passenger Car:

[See PDF for image]

[End of figure]

According to Australia NCAP officials, the Australia NCAP also 
publishes the Crash Test Update, a brochure that provides new crash 
test results about twice a year. In addition to an overall star rating 
for each type of tested vehicle, the brochure presents star ratings 
with comparative bar graphs showing how well vehicles scored within the 
star levels. Figure 46 shows the brochure Australia NCAP officials 
provide for consumers.

Figure 46: Australian NCAP July 2004 Crash Test Update Brochure:

[See PDF for image]

[End of figure]

[End of section]

Appendix VI: European New Car Assessment Programme:

The European New Car Assessment Programme (Euro NCAP) provides 
information for consumers with an assessment of the safety performance 
of some new vehicles sold in Europe. The program was established and 
began rating vehicles in 1997. Its main purposes are to make 
comparative safety rating information available to consumers for 
vehicles in the same class and to provide incentives for manufacturers 
to improve the safety of their vehicles. The program is operated and 
funded by a consortium of six European governments--Catalonia, France, 
Germany, the Netherlands, Sweden, and the United Kingdom--and of 
various motoring and consumer organizations throughout Europe, 
including the General German Automobile Association (Allgemeiner 
Deutscher Automobil-Club e V); German Federal Ministry for Traffic, 
Building and Housing (Bundesministerium für Verkehr, Bau-und 
Wohnungswesen); United Kingdom Department for Transport; Dutch Ministry 
of Transport--Public Works and Water Management; FIA (Fédération 
Internationale de l'Automobile) Foundation for the Automobile and 
Society; Catalonia Department of Employment and Industry (Departament 
de Treball i Indústria); International Consumer Research and Testing; 
French Ministry of Equipment (Ministère de l'Equipement); Swedish Road 
Administration; and Thatcham. The Euro NCAP crash testing program was 
modeled from the U.S. NCAP (1979) and the Australia NCAP (1992).

The decision process for Europe involves the use of technical working 
groups and subgroups to examine vehicle safety issues and make 
recommendations for change. Such groups are investigating the 
feasibility of incorporating such safety features as braking and 
handling, visibility and lighting, ergonomics, driver information, and 
whiplash into Euro NCAP. The automobile industry and public safety 
organizations may be involved in providing research or opinions, but 
the committees are free to make decisions they believe appropriate. 
Generally, decisions are made through two working groups, one for 
primary safety systems and one for secondary safety systems, that 
perform research and analysis. The Euro NCAP allows industry 
representatives to participate in the discussions of the subgroups of 
its two technical working groups. Also, the technical working groups 
and automobile manufacturers engage in direct dialogue in industry 
liaison meetings to address issues such as whiplash.

Each member of the Euro NCAP is required to sponsor at least one 
vehicle for crash testing each year. The vehicles are normally acquired 
by the Euro NCAP Secretariat by various methods, including purchasing 
directly from dealers and selecting from manufacturers' production 
lines. The Euro NCAP tests vehicles in categories--superminis, family 
cars, executive cars, roadsters, off-roaders, and multipurpose 
vehicles. The following further describes (1) the testing conducted, 
(2) the methods used for developing the vehicle crash ratings, and (3) 
the approaches taken to share the safety results with the public.

Testing Conducted:

The Euro NCAP performs three vehicle crash tests, a pedestrian 
protection test, and a child restraint test. The three crash tests are 
the 40 percent offset frontal test, the perpendicular side-impact test, 
and the side-impact pole test. The frontal test involves a moving test 
vehicle traveling at 40 mph (64 km/h) crashing into an offset 
deformable aluminum barrier where 40 percent of the vehicle's width 
engages the barrier on the driver's side. The deformable barrier used 
is a crushable aluminum honeycomb face attached to a solid barrier. The 
deformable structure is designed to replicate the essential 
characteristics of the front end of another car. Two instrumented 50TH 
percentile Hybrid III dummies (each weighing about 194 pounds) are used 
to collect data during the crash and are placed in the front driver's 
and front passenger seats.

In the side-impact test, a moving trolley with a deformable barrier is 
towed at about 31 mph (50 km/h) into a stationary test vehicle at a 90 
degree angle centered on the driver seating position. This test 
simulates a side-impact collision. The moving deformable barrier has a 
mass of 2,095 pounds (950 kg) compared with 3,015 pounds (1,367 kg) for 
the U.S. barrier. The European barrier's face is smaller and much 
softer than the face of the barrier used in the U.S. NCAP. However, 
Euro NCAP officials said that because the barrier strikes a vehicle at 
a 90 degree angle, their side-impact test is more aggressive than 
NHTSA's side-impact test. One instrumented 50TH percentile EuroSID-II 
dummy (weighing about 176 pounds) is used to collect data during the 
crash and is placed in the front driver seat.

The pole side-impact test consists of a vehicle placed on a platform 
and propelled at 18 mph (29 km/h) into a cylindrical pole. The pole has 
a diameter of 10 inches (254 mm), and its vertical axis is aligned with 
the front seat dummy's head. One instrumented 50TH percentile EuroSID-
II dummy is used to collect data during the crash and is placed in the 
front driver's seat.

The pedestrian protection test evaluates the impact of dummy parts 
against the bumper, hood, and windshield areas of a vehicle. Adult and 
child-size dummy parts are propelled at specified areas of the hood and 
front bumper of a vehicle to simulate a 25 mph (40 km/h) car-to-
pedestrian collision. The test simulates the impact of a lower leg 
against a bumper, a thigh against the lower edge of the hood, and adult 
and child heads against the upper portion of the hood.

The child protection test evaluates a vehicle's ability to protect 
children by assessing the performance of the vehicle's child restraint 
system in front and side-impact tests. During these tests, two child-
size dummies are placed in the manufacturer's recommended child 
restraints in the rear seat of a vehicle. In the frontal test, a dummy 
with the weight and size of an 18-month-old child (about 24 pounds) is 
placed behind the passenger, and a dummy with the weight and size of a 
3-year-old child (about 33 pounds) is placed behind the driver. In the 
side-impact test, the positions of the two dummies are reversed.

Scoring Tests:

The Euro NCAP bases its assessment of crashworthiness on three types of 
observations made during or after a crash test: (1) dummy measurements 
of forces to the body, used to assess injuries; (2) five measurements 
of vehicle deformation, used to assess the vehicle's structural 
performance; and (3) post-crash inspection data for six areas, which 
are termed "modifiers" because problems in any one of them may result 
in a penalty that modifies the vehicle's assessment score.

In the offset frontal crash test, two instrumented Hybrid III dummies 
are positioned in the front driver's seat and front passenger seat to 
measure injuries to four regions of the body: (1) head and neck; (2) 
chest; (3) knee, femur, and pelvis; and (4) legs and feet. The five 
structural measurements provide vehicle deformation data, indicating 
the amount and pattern of intrusion into the occupant compartment. The 
post-crash inspection provides information about air bag stability, 
steering column movement, A-pillar movement, structural integrity, 
hazardous structures in the knee impact area, and brake pedal movement. 
The dummy measurements and the vehicle deformation data are combined to 
generate a score--up to four points--for each body region. This score 
may be modified by findings from the post-crash inspection.

In the side-impact and pole tests, injury measurements are recorded on 
one EuroSID-II dummy positioned in the front driver's seat. These 
measurements provide data for assessing injuries to four body regions: 
the head, ribs, abdomen (chest or thorax), and pelvis. No structural or 
post-crash inspection data are gathered during these tests. Thus, the 
score for each body region is based on the dummy measurements alone.

In the pedestrian test, readings taken from the adult and child-size 
dummy parts (head and lower limbs) are used to assess the risk of 
injury. The two different size dummy heads are tested at six different 
areas of the hood, and the lower limbs are tested at three areas, for a 
total of 18 impacts tested for each vehicle. Depending on the injury 
measurements recorded from the dummy parts, each impact can receive up 
to 2 points, and the maximum number of points that can be received is 
36 points. See table 6.

Table 6: Euro NCAP Pedestrian Test Assessment Criteria:

Type of test: Child head;
Impact locations: 6;
Maximum: points: 12.

Type of test: Adult head;
Impact locations: 6;
Maximum: points: 12.

Type of test: Upper leg;
Impact locations: 3;
Maximum: points: 6.

Type of test: Lower leg;
Impact locations: 3;
Maximum: points: 6.

Type of test: Total;
Impact locations: 18;
Maximum: points: 36.

Source: Euro NCAP.

[End of table]

The child protection test consists of three assessments that are based 
on (1) dummy measurements and dynamic evaluations, (2) marking 
requirements for child restraint systems, and (3) a vehicle-based 
assessment. Points reflect the results of the three assessments. The 
first assessment uses dummy measurements taken from the two child 
dummies in the frontal and side tests, together with dynamic 
evaluations of ejection from the child restraint system and head 
contact within the vehicle. Another assessment evaluates whether the 
markings on the child restraint fully comply with the test 
requirements. The final assessment evaluates how easily the child 
restraint system can be used inside the vehicle.

A combined star rating is used to show the protection level achieved in 
the offset frontal and side impact tests together. The score for this 
rating is the sum of the scores achieved in these two tests--up to 16 
points for the frontal test and up to 18 points for the side test, for 
a maximum of 34 points. For both tests, each of four body regions can 
receive up to 4 points, for a cumulative score of 16 points per test, 
and for the side test, 2 additional points can come from an optional 
pole test, which assesses protection for only one body region--the 
head. The pole side-impact test is an option for the manufacturer of a 
vehicle that has head-protecting side air bags. Finally, up to 3 more 
points can be earned for having a safety belt reminder system. The 
points are translated into stars, as shown in table 7.

Table 7: Euro NCAP Front and Side-Impact Star Rating System:

Star rating: 1;
Points: 1-8.

Star rating: 2;
Points: 9-16.

Star rating: 3;
Points: 17-24.

Star rating: 4;
Points: 25-32.

Star rating: 5;
Points: 33-40.

Source: Euro NCAP.

[End of table]

If the crash tests demonstrate a high risk of a life-threatening 
injury, indicated by an injury score of 0 for the head, chest, abdomen, 
or pelvis, then a warning note is added to the overall rating. Euro 
NCAP uses a "struck star" to convey this warning. When the star is 
struck through, it highlights concern that there is a serious risk of 
injury in at least one vulnerable body region. These concerns are based 
on data from the offset frontal test for the head or chest and from the 
side-impact test for the head, chest, abdomen, or pelvis. A star cannot 
be struck because of findings from post-crash inspections showing the 
effects of modifiers.

Euro NCAP provides a separate rating of one to four stars to show the 
level of pedestrian protection. The score for this rating sums the 
results of the 18 impact tests of dummy parts propelled into the 
specified areas of the bumper and hood. A vehicle can earn up to 2 
points for each test, for a maximum of 36 points. The points are 
translated into stars, as shown in table 8.

Table 8: Scoring Basis for Euro NCAP Pedestrian Ratings:

Star rating: 0;
Points: 0.

Star rating: 1;
Points: 1-9.

Star rating: 2;
Points: 10-18.

Star rating: 3;
Points: 19-27.

Star rating: 4;
Points: 28-36.

Source: Euro NCAP.

[End of table]

Euro NCAP also provides a separate rating of one to five stars to show 
the level of child protection. Currently, the tests on which this 
rating is based can produce a maximum of 49 points, but the rating 
scale allows further points to be awarded for future developments in 
child protection. Table 9 shows how the points are translated into 
stars.

Table 9: Scoring Basis for Euro NCAP Child Protection Star Ratings:

Star rating: 0;
Points: 0.

Star rating: 1;
Points: 1-12.

Star rating: 2;
Points: 13-24.

Star rating: 3;
Points: 25-36.

Star rating: 4;
Points: 37-48.

Star rating: 5;
Points: 49-60.

Source: Euro NCAP.

[End of table]

Sharing Results with the Public:

Vehicle rating information is available on the Euro NCAP Web site, 
through press releases, and through popular consumer magazines. Figure 
47 shows the ratings that the program makes available to consumers on 
the Internet--a front and side-impact rating, a pedestrian protection 
rating, and a child restraint protection rating. The pedestrian 
protection rating is intended to encourage manufacturers to start 
designing for pedestrian protection. The child restraint protection 
rating is based on a vehicle's performance using the child seats 
recommended by that vehicle's manufacturer. Specifically, the rating 
depends on the fitting instructions for the child seats, the car's 
ability to accommodate the seats safely, and the seats' performance in 
front and side impact tests.

Figure 47: European NCAP Rating of a 2004 Mid-size Passenger Car:

[See PDF for image]

[End of figure]

In addition to star ratings, the Euro NCAP uses color-coded dummy 
injury diagrams to show how specific body regions performed in the 
frontal, side, and pole crash tests. The color-codes are: Good (green), 
Adequate (yellow), Marginal (orange), Weak (red), and Poor (brown). The 
colored injury diagrams display the risk of injury to the various body 
regions, as shown in figure 48.

Figure 48: Dummy Injury Diagrams of Driver and Passenger in Frontal 
Test, and Driver in Side Test:

[See PDF for image]

[End of figure]

The Euro NCAP divides its testing into two phases and releases the 
results twice a year, in November and June. The results are posted on 
the program's Web site, issued in press releases, and published by What 
Car? (a British car magazine), Which? Car (a magazine owned and 
produced by British consumer associations), and the General German 
Automobile Association (ADAC) magazine. Other consumer magazines in 
Europe provide additional crash test information.

[End of section]

Appendix VII: Japan New Car Assessment Program:

The National Agency for Automotive Safety and Victims' Aid (NASVA) 
conducts the Japan NCAP and is funded by the government through the 
Ministry of Land, Infrastructure, and Transportation. According to 
NASVA officials, the Automobile Assessment Committee, made up of 12 
members appointed by the ministry, oversees the program. The committee 
includes four working groups, each focusing on specific areas: crash 
tests, tests of active safety systems such as brakes, pedestrian tests, 
and tests of child restraint systems. NASVA officials conduct research 
in these areas and propose changes to the program that must be approved 
by the committee. NASVA officials said that the Japan NCAP is funded 
through appropriations from the Compulsory Automobile Liability 
Insurance that every car owner must pay.

The Japan NCAP began testing vehicles in 1995, starting with a full 
frontal collision test. The program added the side-impact test in 1999 
and the offset frontal test in 2000. Vehicles are selected for testing 
on the basis of sales. By 2004, the program had evaluated 79 vehicles 
representing over 80 percent of those that were on the market at that 
time. Ratings for 60 of these vehicles were carried over from previous 
years' testing, and ratings for 19 vehicles were based on tests 
performed in 2003. Testing is conducted at the Japan Auto Research 
Institute under the control and supervision of NCAP officials. The 
institute crash tests cars, minivans, and SUVs and performs other NCAP 
tests, such as the brake and pedestrian tests. The research laboratory 
has one track for conducting frontal and side-impact tests. In these 
tests, either the vehicle is towed to strike the barrier, or, in side-
impact tests, the barrier is moved to strike the vehicle. In 2005, the 
institute plans to open a new test facility with multiple tracks that 
will enable researchers to conduct vehicle-to-vehicle crash tests at 
various angles.

Testing Conducted:

The Japan NCAP performs a variety of safety tests and rates vehicles 
according to the results. It conducts three types of crash tests--a 
full frontal test, an offset frontal test, and a perpendicular side-
impact test. In addition, it performs a braking test, which measures 
the performance of an active safety system that enables a driver to 
avoid a crash. The program further assesses how easily doors are opened 
and occupants are removed after a crash and how well vehicles perform 
if they strike pedestrians. The program also evaluates how well child 
safety seats perform.

The Japan NCAP is the only program that conducts both the full frontal 
and the offset frontal crash tests. Together, the two tests assess both 
the potential for injuries caused by intrusion and the effectiveness of 
the vehicle's restraint system. The full frontal test is performed by 
towing a vehicle to collide with a rigid barrier at 55 km/h (about 34 
mph). This test simulates a head-on collision between two vehicles of 
the same size traveling at the same speed. The offset frontal test 
involves towing a vehicle into a deformable barrier that represents the 
front end of another vehicle and simulates a head-on collision of two 
vehicles traveling at 40 mph. In this test, only a portion of the front 
end (40 percent) engages the barrier, and the impact on the vehicle 
body is greater than the full frontal test because much of the crash 
energy is distributed to one side of the vehicle. Thus, there is the 
possibility of substantial vehicle deformation, which makes this test 
suitable for evaluating injuries caused to occupants by intrusion into 
the occupant compartment. The program uses the Hybrid III dummy that 
represents a man of about 5 feet 10 inches tall and weighing about 185 
pounds.

The side-impact test propels a moveable deformable barrier weighing 
about 2,090 pounds into the driver's and passenger's side of the 
vehicle, simulating a perpendicular collision at 55 km/h (about 34 
mph). The barrier is shaped like the front end of a car, and because it 
is not rigid, its performance is intended to simulate a vehicle's 
response in an actual collision. A EuroSID-I dummy is placed in the 
driver's seat. This dummy is the same height as the Hybrid III dummy 
but weighs about 20 pounds less. The EuroSID-I dummy was designed to 
measure the risk of injury to the head, chest, abdomen, and pelvis.

The Japan NCAP conducts a braking performance test that measures how 
far a vehicle travels before it stops and how stable it is at the time 
of braking when it is stopped abruptly while traveling at about 62 mph. 
The braking test is a test of an active safety system because it 
enables the driver to avoid a crash. The test is performed under wet 
and dry road conditions for a vehicle with a driver and a weight on the 
front passenger seat. To ensure consistent testing, Japan NCAP 
officials said, the dry road surface temperature must be 95.0 degrees 
plus or minus 18.0 degrees Fahrenheit and the wet road surface 
temperature must be 80.6 degrees plus or minus 9.0 degrees Fahrenheit 
because the temperature of the road surface affects the distance it 
takes to stop the vehicle. Japan NCAP officials also said that all 
braking tests must be performed at the same location because road 
surfaces vary and surface differences could affect test results. 
Professional drivers conduct the tests, and the speed of the vehicle 
and force with which the drivers depress the brake pedal are monitored 
electronically to ensure consistency. Three braking tests are conducted 
to be sure that the result is not due to a flaw in the testing process. 
Figure 49 illustrates the braking test.

Figure 49: Test Vehicle Undergoing Brake Tests Under Dry and Wet 
Conditions:

[See PDF for image]

[End of figure]

In addition, the Japan NCAP assesses and scores the ease with which 
doors can be opened and the dummies removed after a crash test. The 
purpose of the accessibility assessment is to rate how easily emergency 
responders can assist injured persons. The rating is based on whether 
the doors can be opened with one hand, two hands, or whether tools are 
needed to open the doors.

The pedestrian test measures the effect of a pedestrian being hit by a 
vehicle traveling at about 22 mph if the pedestrian's head strikes part 
of the hood or windshield. This test was initiated because pedestrian 
fatalities represent a high percentage of total vehicle fatalities in 
Japan. Dummies modeling the head of an adult or a child (head impactor) 
are projected toward the car hood from a testing machine. The force 
received by the head impactor is measured and then evaluated using a 
head injury criterion. The test is conducted on multiple points on each 
car, and the impact angles differ according to the shape of the front 
part of three types of vehicles--sedan, SUV, and van. Figure 50 
illustrates how the test is performed.

Figure 50: Pedestrian Head Impact Test and Target Area:

[See PDF for image]

Note: The distance between the ground and where the pedestrian's head 
hits the car in crashes is called the wrap-around distance, which is 
measured according to the length of that area. The location of the 
impact area for adult's and children's heads is based on the data from 
actual crash data.

[End of figure]

The pedestrian test is conducted on vehicles with three different body 
types, as shown in table 10.

Table 10: Japan NCAP Vehicle Types Used for Pedestrian Test:

Vehicle type: Sedan;
Definition: With the hood leading edge height less than 835 mm.

Vehicle type: SUV;
Definition: With the hood leading edge height more than 835 mm.

Vehicle type: Van;
Definition: With the hood angle more than 30 degrees.

Source: Japan NCAP, National Agency for Automotive Safety and Victims' 
Aid.

[End of table]

The Japan NCAP also assesses the safety performance of child seats in a 
car crash and the ease of using the seats. Child seats are installed in 
the rear passenger seats of a test vehicle stripped down to its body 
frame. The test uses dummies to represent a 9-month-old child and a 3-
year-old child. The test vehicle is placed on a sled and subjected to a 
shock identical to the test speed used in the full frontal crash test.

Scoring Tests:

The Japan NCAP measures injuries to the head, neck, chest, and upper 
(femur) and lower (tibia) legs for both the full frontal and offset 
frontal crash tests. Points vary by body region, from 2 points for 
upper and lower leg injuries to 4 points for head, neck, or chest 
injuries, according to the extent of injuries as measured by crash test 
dummies. Vehicle deformation is measured after the crash test, and if 
certain limits are exceeded, a point is deducted from the score for one 
body area, according to where the deformation occurred. In addition, 
weighting factors are assigned according to the frequency of injuries 
to these body areas in vehicle crashes. The weighted points for each 
body area are then combined to arrive at separate total point scores 
for the driver and the passenger in full frontal and offset frontal 
crash tests. The maximum score that a vehicle can achieve is 12 points 
because of the way the injuries are weighted.

For the side-impact crash test, the Japan NCAP measures injuries to the 
driver's head, chest, abdomen, and pelvis.[Footnote 71] Four points are 
assigned for each body area and then weighted according to the 
incidence of injuries in this type of accident, with lesser weights 
assigned to the abdomen and pelvis than to the head and chest. Again, 
the maximum score that a vehicle can achieve is 12 points, because of 
the way injuries to the driver are weighted.

The Japan NCAP is the only program that adjusts its test results by 
weighting the injury scores according to historical crash data. NCAP 
officials said they can do this because the police are well trained to 
investigate every accident and provide thorough reports to the 
government.

For the pedestrian test, a series of head injury scores is used to 
assign injury probability levels from 5 (the best) to 1 (the worst). 
The results are then combined to arrive at an overall score. According 
to NCAP officials, vehicles with hoods that are more flexible and 
compress upon impact can receive better scores than those that are 
rigid and leave no room between the hood and the engine for the impact 
to be absorbed.

Child seats are evaluated according to their performance in a collision 
and their ease of use. For the collision test, overall ratings of 
Excellent, Good, Normal, and Not Recommended are assigned. The ratings 
are primarily based on the head and chest injury scores taken from the 
dummies used in the test. Five child seat specialists assessed the ease 
of installation, the ease of understanding the instructions, the 
product warning labels and markings to aid in installation, the 
structural design, and the ease of securing the child in the seat. For 
each area, the specialists assigned points, from 5 (the best) to 1 (the 
worst). The scores given by the specialists were averaged and reported 
separately for each area.

Sharing Results with the Public:

Initially, the Japan NCAP used a four-letter system to rate vehicles' 
crashworthiness, in which "A" reflected the highest scores for 
performance and "D" reflected the lowest scores. As vehicles' 
performance improved, more and more vehicles achieved an "A" rating. To 
help consumers better differentiate vehicles' performance, NCAP 
officials expanded the range of ratings to include AA and AAA. This 
same scale was later converted to six stars. Many vehicles have 
achieved a five-star rating, and some have received a six-star rating 
for occupant protection. In addition to the star ratings, the Japan 
NCAP reports the percentage of possible points that each vehicle 
received and provides a bar chart indicating how well the vehicles 
performed in these tests.

Figure 51 shows how the Japan NCAP communicates its ratings to 
consumers as two overall ratings--one for the driver's seat and one for 
the passenger's seat. The overall safety rating for the driver's seat 
combines the results of the two frontal crash tests (full and offset) 
and the side-impact test. The overall safety rating for the passenger's 
seat includes the results of the full frontal and offset frontal tests. 
The Japan NCAP also provides consumers with star ratings by type of 
test for the driver's and passenger's seats and makes the detailed test 
information available to consumers for each crash test, as shown in 
figure 52. Consumers are also provided with ratings on how difficult it 
was to open the door after the test (openability) and how difficult it 
was to retrieve the dummy from the vehicle after the crash test 
(rescueability), as shown in figure 53 and 54 respectively. Although 
not shown as part of the crashworthiness rating, the ratings for the 
pedestrian tests are provided, as well as the ratings for the child 
restraint seats (Excellent, Good, Normal, and Not Recommended). 
Furthermore, the Japan NCAP has provided consumers with comparative 
information on vehicles' braking capability on wet and dry pavements.

Figure 51: Japan NCAP Rating of a 2003 Mid-sized Passenger Car:

[See PDF for image]

[End of figure]

Figure 52: Example of Japan NCAP Detailed Full Frontal Data Available 
for a 2003 Test:

[See PDF for image]

[End of figure]

Figure 53: Key to Japan NCAP's Door Openability Ratings:

[See PDF for image]

[A] At times, a vehicle may turn sideways (roll over) after a side 
impact. This test is performed to measure how well passengers are 
protected when a vehicle is hit while stopped; it is not intended to 
evaluate the stability of the vehicle while it is in motion.

[End of figure]

Figure 54: Key to Japan NCAP's Rescueability Ratings:

[See PDF for image]

[End of figure]

The Japan NCAP publicizes ratings through the NASVA Web site and 
publishes a brochure that contains summary information on vehicle 
ratings. In addition to the summary ratings, detailed information for 
each vehicle tested is posted on the Web site and published in a book 
by NASVA every year. The Japan Automobile Federation also publishes 
NCAP vehicle ratings in its magazine and posts the information on its 
Web site. According to Japan NCAP officials, vehicle ratings and new 
tests under the program are further publicized in television specials. 
According to NASVA officials, the public response was very positive 
after these specials because the number of hits to NASVA Web site 
increased by 15 to 20 percent in the few days after each broadcast.

[End of section]

Appendix VIII: Different Types of Tests Used by the Programs:

Table 11: Description of Different Types of Tests Used by the Programs:

Test type: Frontal tests.
Full Frontal tests;
Description of test: The test vehicle's entire front end is crashed 
into a fixed rigid barrier at 35 miles per hour (mph), which is 
equivalent to a head-on collision between two similar size vehicles, 
each moving at 35 mph. 

Only the U.S. NCAP and Japan currently conduct this test.[A]; 
Assessment: Frontal tests: The test assesses the ability of vehicle 
restraint systems, such as safety belts and air bags, to restrain 
occupants. 

The test maximizes the energy absorbed by the front of the vehicle so 
that the occupant compartment is more likely to remain intact; 
Considerations: Frontal tests: Test results cannot be compared between 
vehicles of different weight classes.

Test type: Frontal tests:
Off-set Tests;
Description of test: The test vehicle is the equivalent of two vehicles 
of the same weight crashing into each other. The vehicle is crashed 
into a deformable barrier at 40 mph where approximately 40 percent of 
the vehicle's width makes contact with the barrier. 

Australia, Europe, Japan, and the Insurance Institute currently conduct 
this test; Assessment: Frontal tests: The test assesses a vehicle's 
structural integrity and its ability to manage the energy generated in 
the crash entirely on the driver's side of the vehicle. 

This type of crash forces a smaller area of the vehicle's structure to 
absorb the energy from the crash; 
Considerations: Frontal tests: Test results cannot be compared between 
vehicles of different weight classes.

Test type: Side tests;
Angled Side; 
Description of test: The test vehicle is struck by a moving deformable 
barrier at 38.5 mph at an angle. Both the barrier and the driver's side 
of the vehicle are parallel, so that the entire face of the barrier 
impacts the side of the vehicle. Only the U.S. NCAP currently conducts 
this test; 
Assessment: Frontal tests: The test represents an intersection-type 
collision. The barrier striking the test vehicle is shaped like a 
passenger car and is about the size of a medium passenger car, weighing 
3,015 pounds; 
Considerations: Frontal tests: Test results can be compared across 
vehicle weight classes because each vehicle is struck by a barrier of 
the same size and weight. The test does not assess the risks of head 
injury from side impacts from vehicles like SUVs and pickups.

Test type: Side tests;
Perpendicular side with medium barrier; 
Description of test: Frontal tests: The test vehicle is struck by a 
moving deformable barrier traveling at about 31 mph (50 km/h) at a 90 
degree angle centered on the driver side seating position. Australia, 
Europe, and Japan currently conduct this test.[B]; 
Assessment: Frontal tests: The test represents a 90 degree intersection-
type collision. The barrier that strikes the test vehicle weighs 2,095 
pounds (950 kg) and has a front end shaped to simulate the typical 
front end of a passenger car; 
Considerations: Frontal tests: Test results can be compared across 
vehicle weight classes because each vehicle is struck by a barrier of 
the same size and weight.

Test type: Side tests;
Perpendicular side with large barrier;
Description of test: Frontal tests: The test vehicle is struck by a 
moving deformable barrier traveling at 31 mph (50 km/h) perpendicular 
impact into the driver side of a passenger vehicle. The Insurance 
Institute currently conducts this test; 
Assessment: Frontal tests: The test represents a 90 degree intersection-
type collision. The moving deformable barrier that strikes the test 
vehicle weighs about 3,300 pounds (1,500 kg), is higher than other 
barriers with a front end shaped to simulate the typical front end of a 
pickup or SUV; 
Considerations: Frontal tests: Test results can be compared across 
vehicle weight classes because each vehicle is struck by a barrier of 
the same size and weight. The test was designed to encourage automakers 
to provide side-impact head protection by installing air curtain bags 
that extend low enough to protect smaller occupants in front and rear 
seats.

Test type: Side tests;
Pole side test;
Description of test: Frontal tests: The test vehicle is placed on a 
platform and propelled at 29 km/h (about 18 miles per hour) into a 
stationary cylindrical pole producing a side-impact crash on the 
driver's door. Australia and Europe currently conduct this test; 
Assessment: Frontal tests: The test represents a side-impact collision 
involving a narrow object such as a tree or pole. The diameter of the 
pole is 10 inches (254 mm) and aligned with the front occupant's head; 
Considerations: Frontal tests: The side-impact test is demanding on 
vehicles' structures, and heavy vehicles have to cope with higher loads 
than lighter vehicles.

Test type: Other tests;
Rollover tets;
Description of test: Frontal tests: The test vehicle is subjected to a 
series of left/right tests and right/left tests at different speeds and 
steering wheel angles.The first run of tests is conducted at 35 mph, 
with each subsequent run conducted at about 40 mph, 45 mph, 47.5 mph, 
and 50 mph, until the vehicle "tips up" (fails) or attains a speed of 
50 mph on the last run of each test without tipping up (pass). Only the 
U.S. NCAP currently conducts this test; 
Assessment: Frontal tests: The dynamic test represents an untripped 
rollover by simulating a high-speed collision-avoidance maneuver. The 
Static Stability Factor is a measurement used to determine the top-
heaviness of a vehicle, accounting for situations of potential tripped 
rollovers such as when a vehicle strikes a curb; 
Considerations: Frontal tests: Rollovers where the vehicle is not 
tripped account for less than 5 percent of all rollover crashes.

Test type: Other tests;
Pedestrian protection test;
Description of test: Adult and child-size dummy parts (such as heads) 
are projected at specified areas of a test vehicle front end to 
replicate a car colliding with a pedestrian. Australia, Europe, and 
Japan currently conduct this test; 
Assessment: Frontal tests: The test is used to assess the risk to 
pedestrians if struck by the front of a car; 
Considerations: Frontal tests: This test evaluates the front end design 
of vehicles and the effect on pedestrian injuries.

Test type: Other tests;
Child restraint test;
Description of test: Frontal tests: Euro NCAP tests child restraints by 
placing child-size dummies in child seats during frontal and side crash 
tests. The Japan NCAP test evaluates child seats by placing child-size 
dummies in child seats using a test sled instead of actual collisions; 
Assessment: Frontal tests: The Euro NCAP test is used to evaluate child 
protection focusing on the interaction of a vehicle's child restraint 
system and a child seat. The Japan NCAP test evaluates the ease of 
correctly using the child seats; 
Considerations: Frontal tests: The Euro NCAP test evaluates the 
performance of a car seat during a front or side collision. Japan NCAP 
tests child restraint systems independent of vehicle performance in a 
simulated frontal crash and evaluates the ease of use.

[A] The Japan NCAP full frontal test is performed at 55 km/h (about 34 
mph).

[B] The Japan NCAP perpendicular side test is performed at 55 km/h 
(about 34 mph).

[End of table]

[End of section]

Appendix IX: GAO Contacts and Staff Acknowledgments:

GAO Contacts:

JayEtta Hecker, (202) 512-2834:
Glen Trochelman, (312) 220-7729:

Acknowledgments:

In addition to those named above, Vashun Cole, Michelle Dresben, Colin 
Fallon, Kathleen Gilhooly, Doug Manor, Terry Richardson, Beverly Ross, 
Brian Sells, Jena Sinkfield, Stacey Thompson, and Frank Taliaferro made 
key contributions to this report.

(544085):

FOOTNOTES

[1] Economic costs include productivity losses, property damage, 
medical costs, rehabilitation costs, travel delay, legal and court 
costs, emergency services, insurance administration costs, and the 
costs to employers. For more information, see U.S. Department of 
Transportataion National Highway Traffic Safety Administration, The 
Economic Impact of Motor Vehicle Crashes, 2000 (Washington, D.C.: May 
2002).

[2] The Insurance Institute for Highway Safety is a nonprofit research 
and communications organization in the United States funded by the auto 
insurance industry.

[3] The rollover test is nondestructive.

[4] Driver behavior involves the actions taken by or the condition of 
the driver of the motor vehicle, including speeding and violating 
traffic laws, as well as the effects of alcohol or drugs, inattention, 
decision errors, and age. The driving environment associated with 
crashes includes roadway design, roadside hazards, and road conditions. 
Vehicle factors that contribute to crashes include vehicle-related 
failures or design of the vehicle. For additional information on the 
causes of motor vehicle crashes, see GAO, Highway Safety: Research 
Continues on a Variety of Factors That Contribute to Motor Vehicle 
Crashes, GAO-03-436 (Washington, D.C.: Mar. 31, 2003).

[5] Crashworthiness is defined in 49 USC § 32301 as "the protection a 
passenger motor vehicle gives its passengers against personal injury or 
death from a motor vehicle accident." Section 49 USC § 32302, 
subsection (a), requires the Secretary of Transportation to "maintain a 
program for developing . . . information on passenger motor vehicles['] 
. . . crashworthiness." Subsection (b) requires the Secretary to 
"provide to the public information developed under subsection (a) . . . 
in a simple and understandable form." 

[6] NHTSA also rates the ease of use for child safety seats under NCAP, 
which is not included in the scope of this report.

[7] An additional 16 vehicles were also rated for rollover based on a 
calculation of their top-heaviness. 

[8] The Federal Motor Vehicle Safety Standards (CFR Title 49: chapter 
V, part 571) were created under section 103 of the National Traffic and 
Motor Vehicle Safety Act of 1966. The standards cover a broad range of 
safety concerns, from windshield wipers and brakes to crashworthiness.

[9] Federal Motor Vehicle Safety Standards No. 208 and No. 214, for the 
frontal and side tests, respectively.

[10] Vehicle manufacturers face severe civil penalties, potentially 
expensive vehicle safety recalls, and potential legal liability if 
NHTSA's testing indicates that a vehicle does not comply with Federal 
Motor Vehicle Safety Standards.

[11] The dynamic rollover test is a specific driving maneuver. The 
Static Stability Factor is a laboratory measurement.

[12] According to NHTSA officials, they are conducting pilot studies to 
determine the feasibility of conducting more tests under NCAP, 
including a child seat crashworthiness sled test, rating vehicles on 
how well they protect children, braking, and lighting.

[13] See appendix VIII for a diagram of the full frontal crash test 
configuration.

[14] The higher speed creates more crash energy or power and inflicts 
potentially more damage on the vehicle and its occupants. The increased 
speed also allows differences in crashworthiness to be more readily 
observed. The difference in speed is only 5 mph faster, but the total 
kinetic energy released in a 35 mph crash is 36 percent greater than 
that released in a 30 mph crash.

[15] See appendix VIII for a diagram of the angled side crash test 
configuration.

[16] NHTSA performs the angled side test because accident data suggests 
that intersection or perpendicular impacts occur with two moving 
vehicles. To simulate a moving car to moving car crash, the angled or 
"crabbed" side test is used. The test simulates the striking vehicle 
traveling at approximately 34 mph and the test vehicle traveling at 
roughly half that speed--approximately 17 mph--due to the 63/27 degree 
configuration.

[17] The safety standard, No. 214, applies to both sides of the vehicle 
and specifies performance requirements for vehicle occupant protection 
in side crashes. NCAP side tests only apply to the driver's side of the 
vehicle.

[18] Appendix III provides additional discussion of the basis for 
NHTSA's rollover rating.

[19] The head and chest measurements are known as the Head Injury 
Criterion (HIC) and chest deceleration value, measured in Gs.

[20] The forces on the dummy's lower spine and the greater of the 
accelerations of the lower and upper ribs are used in the side NCAP 
star ratings. These accelerations are averaged to produce a measurement 
known as the Thoracic Trauma Index.

[21] The crash test dummies used in the frontal and side NCAP test 
represent the 50TH percentile male or the average male in terms of 
height and weight. The dummy used is approximately 5 feet 9 inches tall 
and weighs approximately 170 pounds.

[22] If these force measurements are sufficiently high to cause serious 
or life-threatening injury, they are reported separately as a safety 
concern but are not included in the star rating calculation. NHTSA does 
not report tibia and feet measurements as safety concerns because there 
are no agency-approved metrics for these body regions. In side impact 
tests, forces to the head are also measured, though they are not 
included in the star rating. They are also reported separately as a 
safety concern if considered high.

[23] See appendix II for more information about how dummy measures and 
information about human injury are combined to develop a probability of 
injury. 

[24] NHTSA began rating a vehicle's risk of rollover in 2001. Rollover 
ratings between 2001 and 2003 were based on the measure of a vehicle's 
top-heaviness because NHTSA had determined that this was a good 
indicator of rollover likelihood. In 2004, NHTSA began to conduct a 
dynamic rollover test as mandated by the TREAD Act. 

[25] A vehicle's Static Stability Factor ranges (SSF) from 
approximately 1.00 to 1.50, with SUVs lying in the lower half and 
passenger cars lying in the upper half of that range. The lower the 
SSF, the more likely a rollover will occur. For rollover ratings, 
vehicles are rated using a statistical risk model that incorporates 
both the vehicle's SSF and its tip or no-tip result in the dynamic 
test. NHTSA does not always subject passenger cars to the dynamic test. 
For some passenger cars, NHTSA imputes or assigns a "no tip-up" if 
other passenger cars that are more top-heavy did not tip up during the 
dynamic test. NHTSA periodically puts vehicles with imputed test 
results through the dynamic test to verify the no tip-up assignation.

[26] See appendix III for a more detailed description of the 
development of the NCAP rollover rating.

[27] NCAP rating information for vehicles prior to model year 1990 is 
archived and not available on the Web site. NCAP information for these 
vehicles has been converted into star ratings and is available to 
consumers upon request.

[28] Visits to the Consumer Reports Web site include those seeking 
rating information on an array of consumer products, not just vehicle 
safety.

[29] The Car Book was originally a government publication. It is now 
produced commercially.

[30] The Insurance Institute for Highway Safety is a nonprofit research 
and communications organization in the United States funded by the auto 
insurance industry.

[31] As discussed earlier, the angled side test involves crashing a 
3,015-pound moving deformable barrier traveling at 38.5 mph into a 
standing test vehicle.

[32] The 50TH percentile adult male dummy represents an average-sized 
male, and the 5TH percentile adult female dummy represents a small 
female or 12-year-old child.

[33] NHTSA has proposed a regulatory revision to its Federal Motor 
Vehicle Safety Standard 214, in which vehicles would have to meet 
additional performance criteria of a pole side test involving a vehicle 
traveling at 20 mph into a rigid pole at a 75 degree angle.

[34] As discussed earlier, NHTSA's rollover test evaluates vehicles' 
rollover resistance using a Static Stability Factor calculation and a 
dynamic test simulating a high-speed collision avoidance maneuver.

[35] Appendixes IV through VII provide details on each country's 
approach to vehicle ratings.

[36] For the side crashes, dummy readings of the chest and lower spine 
are used to compute the Thoracic Trauma Index, which is used to 
determine the star rating.

[37] In the proposed change to the Federal Motor Vehicle Safety 
Standards side impact test, the new pole test would evaluate protection 
to front seat occupants against head, thoracic, and pelvic injuries. 
This would be the first time that head injury criteria would need to be 
met under the side standards.

[38] For the side impact tests, the U.S. NCAP uses the SID-H3 dummy, 
Australia and Euro NCAPs use the EuroSID-II dummy, and Japan's NCAP 
uses the EuroSID-1.

[39] Shopping for Safety: Providing Consumer Automotive Safety 
Information, Special Report 248, Transportation Research Board, 
National Research Council.

[40] Primary safety includes vehicle safety features designed to help 
vehicles avoid crashes. Secondary safety includes vehicle safety 
features designed to help minimize the risk of injury for vehicle 
occupants involved in crashes.

[41] This includes carryover vehicles and corporate twins. Carryover 
vehicles are those that have been tested under NCAP in previous years 
and whose design and safety rating has not changed. Corporate twins are 
two vehicles that are built on the same chassis, share most of their 
underhood and interior components, and have the same air bag and safety 
belt interaction but are sold under different brand names.

[42] NHTSA made these changes in recognition of the need to upgrade the 
level of safety in frontal crashes in the regulations and to reflect 
specific fatality risks in side crashes in the regulations.

[43] The higher test speed will be phased in for increasing percentages 
of vehicles over 3 years.

[44] This is called self protection and partner protection. While the 
occupants of a large vehicle might be protected in a crash, occupants 
of a smaller vehicle, the partner in a vehicle-to-vehicle crash, would 
have less protection.

[45] Initiatives to Address Vehicle Compatibility, NHTSA, June 2003.

[46] According to a safety expert, automobile manufacturers have 
generally opted to increase the strength and stiffness of the front of 
their vehicles within the original vehicle design rather than adding 
additional length to the front of vehicles which could be used to 
reduce the impact on the vehicle struck.

[47] According to NHTSA officials, in such tests, larger vehicles could 
experience a lower change in velocity than smaller vehicles, 
potentially leading manufacturers to incorporate softer structures that 
would not absorb as much energy when these vehicles are in single 
vehicle crashes.

[48] NHTSA officials said they are addressing this issue by proposing a 
change to the Federal Motor Vehicle Safety Standards that would require 
a 20 mph oblique side pole test.

[49] By contrast, 22 percent of those who died in passenger cars in 
2002 were killed in rollover crashes.

[50] Electronic stability control keeps the vehicle under control by 
applying brakes to individual wheels when the system detects loss of 
control or instability.

[51] Preliminary Results Analyzing the Effectiveness of Electronic 
Stability Control (ESC) Systems, NHTSA, September 2004.

[52] Effects of Electronic Stability Control on Automotive Crash Risk, 
C.M. Farmer, Insurance Institute for Highway Safety, 2004.

[53] The U.S. NCAP uses head and chest crash dummy readings in frontal 
crashes and chest and lower spine readings for side crashes because 
according to NHTSA these are the most important indicators of serious 
or fatal injury. The Insurance Institute uses four body regions for its 
frontal crash analysis and three for its side crash analysis. Japan 
uses five body regions for its frontal crash analysis and four for its 
side crash analysis, while Australia and Europe use four for frontal 
and two for side. The other organizations use additional measures to 
capture what can be serious, although not necessarily life-threatening 
injuries such as those to the victim's legs. 

[54] These Safety Concerns were taken from the several thousand vehicle 
ratings available on NHTSA's Web site, including carryover vehicles and 
corporate twins.

[55] Data collected during NCAP tests but not used to calculate star 
ratings are available on the NHTSA Web site. Other data, such as the 
results of force-of-load testing, are available on the NHTSA research 
Web site. 

[56] Officials from vehicle manufacturers said they are developing new 
restraint systems that they believe will be able to determine the 
weight of the occupant and will vary the safety belt restraint and air 
bag deployment to maximize the protection for different-sized occupants 
during a crash.

[57] Automakers also encouraged NHTSA to work towards consistency with 
other countries to lessen the burden on their testing programs. For 
example, they noted that an advanced side impact dummy called "World 
SID" has been developed and that it should be considered for use in 
side crash tests by all side testing programs.

[58] The U.S. NCAP provides four separate star ratings for crash test 
results. There are separate ratings for drivers and passengers (front 
seat) in frontal crashes and separate ratings for drivers and rear 
(driver side) passengers in side-impact crashes.

[59] "Shopping for Safety," Special Report 248, National Academy Press 
(Washington, D.C.: 1996).

[60] Automotive News. Includes new model year passenger vehicles sales 
from September 2003 though February 2004.

[61] The Euro NCAP selects the vehicles from the manufacturers randomly 
by vehicle identification number.

[62] NHTSA used linear regression to determine the relationship between 
a vehicle's probability of rollover per single vehicle crash and its 
Static Stability Factor controlling for road use and state dummy 
variables. It is important to emphasize that this relationship may only 
be imputed to vehicles involved in single-vehicle crashes and not to 
the vehicle fleet at large. 

[63] The National Highway Traffic Safety Administration's Rating System 
for Rollover Resistance: An Assessment, TRB Special Report 265 
(Washington, D.C.: National Academy of Sciences, 2002).

[64] NHTSA's new model uses logistic instead of linear regression. In 
addition, NHTSA performs a log transformation of the Static Stability 
Factor to increase the accuracy of the model for vehicles with low 
Static Stability Factors. 

[65] The Insurance Institute began evaluating head restraint geometry 
in 1995, and dynamic seat/head restraint tests were added in 2004.

[66] Compression refers to the extent body regions are squeezed during 
the impact and is used as an indicator of injury to internal organs.

[67] Passive safety is the term used to refer to automobile designs and 
technologies that help mitigate the injury potential in vehicle crashes 
(sometimes called "crashworthiness"). Passive safety comes from an 
optimized vehicle structure and vehicle restraint technologies.

[68] The Australia New Car Assessment Program is supported by a 
consortium of 15 members of the Australian Automobile Association. The 
membership includes the National Roads and Motorists Association; the 
Royal Automobile Clubs of Victoria, Queensland, South Australia, 
Western Australia, and Tasmania; the Automobile Association of the 
Northern Territory; the FIA Foundation for the Automobile and Society; 
the state road and transport authorities in New South Wales, Victoria, 
Queensland, South Australia, and Western Australia; the New Zealand 
Automobile Association; and the New Zealand Land Transport Safety 
Authority.

[69] Dummy kinematics (movement) are evaluated by how well the safety 
belt and air bag perform and interact with the steering column and 
other vehicle parts to control movement.

[70] The A-pillar is the side support for the roof that is in the front 
of the passenger compartment and the B-pillar provides side roof 
support in the center of the passenger compartment just behind the 
door. 

[71] The side-impact crash test can also be performed on the passenger 
side of the vehicle, but the rating applies to only one occupant 
position.

GAO's Mission:

The Government Accountability Office, the investigative arm of 
Congress, exists to support Congress in meeting its constitutional 
responsibilities and to help improve the performance and accountability 
of the federal government for the American people. GAO examines the use 
of public funds; evaluates federal programs and policies; and provides 
analyses, recommendations, and other assistance to help Congress make 
informed oversight, policy, and funding decisions. GAO's commitment to 
good government is reflected in its core values of accountability, 
integrity, and reliability.

Obtaining Copies of GAO Reports and Testimony:

The fastest and easiest way to obtain copies of GAO documents at no 
cost is through the Internet. GAO's Web site ( www.gao.gov ) contains 
abstracts and full-text files of current reports and testimony and an 
expanding archive of older products. The Web site features a search 
engine to help you locate documents using key words and phrases. You 
can print these documents in their entirety, including charts and other 
graphics.

Each day, GAO issues a list of newly released reports, testimony, and 
correspondence. GAO posts this list, known as "Today's Reports," on its 
Web site daily. The list contains links to the full-text document 
files. To have GAO e-mail this list to you every afternoon, go to 
www.gao.gov and select "Subscribe to e-mail alerts" under the "Order 
GAO Products" heading.

Order by Mail or Phone:

The first copy of each printed report is free. Additional copies are $2 
each. A check or money order should be made out to the Superintendent 
of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or 
more copies mailed to a single address are discounted 25 percent. 
Orders should be sent to:

U.S. Government Accountability Office

441 G Street NW, Room LM

Washington, D.C. 20548:

To order by Phone:

Voice: (202) 512-6000:

TDD: (202) 512-2537:

Fax: (202) 512-6061:

To Report Fraud, Waste, and Abuse in Federal Programs:

Contact:

Web site: www.gao.gov/fraudnet/fraudnet.htm

E-mail: fraudnet@gao.gov

Automated answering system: (800) 424-5454 or (202) 512-7470:

Public Affairs:

Jeff Nelligan, managing director,

NelliganJ@gao.gov

(202) 512-4800

U.S. Government Accountability Office,

441 G Street NW, Room 7149

Washington, D.C. 20548: