Overview of Findings

The importance of unbiased quality and safety information in automotive purchase decisions to many consumers is well known. While governmental agencies such as the National Highway Traffic Safety Administration (NHTSA) have long served as major providers of such data, the Insurance Institute for Highway Safety (IIHS) and its affiliate, the Highway Loss Data Institute (HLDI), have, over the past ten years, become the most publicly visible entities in the sphere of automotive safety research.

Indeed, thanks to its highly successful collaboration with NBC’s Dateline NBC news magazine, it is likely that the IIHS’s thrice-yearly crash test broadcasts (which are viewed by over 10 million people at the time of broadcast and by millions more through follow-up coverage on MSNBC, the NBC Nightly News, and other NBC-affiliated media outlets) are the only vehicle crash tests known to most U.S. vehicle consumers.

The initial data source for the analysis consisted of every high-speed (40 mph), frontal, offset automotive crash tests broadcast on Dateline NBC between April 1995 and January 1, 2002. However, only those vehicles which were produced by U.S. automotive companies, foreign subsidiaries of U.S. companies, or foreign manufacturers with stocks listed on U.S. stock exchanges were included in the study due to the unavailability of foreign stock price data. The final sample consisted of 128 vehicle tests.

Despite employing four different event examination windows (ranging from 3 to 27 trading days in length), there is no evidence that the broadcast of IIHS crash tests resulted in economically or statistically significant changes in manufacturer stock prices. Nor did a detailed examination of vehicle market shares on the month of the test or the following two calendar months reveal changes in vehicle sales levels.

One important hypothesis which would explain the both markets’ apparent disregard for the IIHS crash tests is that such laboratory crash tests have little correlation with real-world accident experience. However, when class-adjusted IIHS crash test scores were compared with class-adjusted actual bodily injury claims experience collected by HLDI, the relationship between the two variables was very high and indicated that same-class vehicles rated highly by the IIHS were likely to experience the lowest injury frequency.

Thus, the lack of consumer and stock market responses, combined with the obvious value of IIHS crash tests to predict vehicle safety experiences, presented something of a puzzle. Accordingly, several additional tests were performed, and these tests constitute the study’s greatest contribution to the literature.

One hypothesis which could explain why IIHS crash tests seemed to be ignored by consumers was that manufacturers “managed” their rebate programs to mitigate the sales effects of unusually good or poor IIHS crash test results. But detailed tests of manufacturer cash rebates around the time of the IIHS broadcasts did not uncover any evidence of such changes in vehicle purchase incentives.

Next, in an effort to ascertain whether consumer estimates of vehicle safety might help explain the apparent inattention of consumers to the information content of the IIHS broadcasts, 80 undergraduate business students were asked for their opinions of the relative safety of every IIHS tested vehicle using the same injury loss scale employed by HLDI. In the preparation of these estimates, the students were given only the vehicle model (e.g., Jeep Grand Cherokee), manufacturer (DaimlerChrysler), and class (e.g., mid-size SUV). No photographs or any kind were provided, nor could the students communicate with each other during the survey.

Correlations between the student estimates of expected vehicular safety and the IIHS crash test scores were negative and statistically significant. However, in results which underscore the importance of the IIHS crash tests for comparisons within a given vehicle class, the student estimates of safety and class-adjusted HDLI loss estimates were not statistically related. Clearly, in same-class purchase decisions, the IIHS crash tests provide much better assessments of expected vehicle safety than the typical consumer estimate.

However, when both the IIHS crash test results and the student estimates of safety were correlated with the raw (non-class-adjusted) HLDI data, a very different and striking picture emerged. Indeed, the correlation between the IIHS crash tests and the non-class-adjusted HLDI injury statistics was very low and statistically insignificant.

This suggests strong support for the IIHS’s disclaimer that “frontal crash test ratings cannot be compared across vehicle type and weight categories.” Unfortunately, the IIHS itself, in its Dateline NBC programming, has mentioned this important disclaimer in only 4 of the last 19 (and only 1 of the most recent 9) crash test broadcasts.

Thus, faced with such an extraordinary incongruity between “tv data” and common sense (e.g., “Surely, that large pickup truck just has to be safer than that tiny little sedan.”) and lacking the explicit knowledge that IIHS crash tests are valid predictors of safety only within the same vehicle class, some consumers may be led to disregard the IIHS crash tests altogether and rely strictly upon their instincts about vehicle safety performance when they purchase a vehicle. But, in so doing, these same consumers will be disregarding an extraordinarily valuable source of safety data valid to help make vehicle purchase decisions within a given market class.

Finally, the results of the study also suggest support for the hypothesis that Dateline NBC/IIHS programs which include the full compliment of all the vehicles in a given tested class (and which include only vehicles from that one class) may provide consumers with the most valuable and informationally unambiguous vehicle safety information.

Significance of the Research

This study provides the results of an important series of tests which attempted to ascertain the influence of the IIHS’s high-speed crash test results on consumer vehicle market shares and manufacturer stock prices. Even more importantly, the study also addresses key questions concerning the overall correlations between IIHS crash test results, NHTSA crash test results, consumer estimates of expected vehicle safety, and actual vehicular loss experience and, in so doing, provide important public policy recommendations concerning the program.