NHTSA Computation Methods: Misleading Statistics

By Ed Haas (reprinted here with author's permission)

To further illustrate, there were 16,653 alcohol-related traffic fatalities in 2000, according to the NHTSA. Of these 16,653 alcohol-related fatalities, 12,892 involved at least one driver or non-occupant with a BAC of 0.10 g/dl or greater. 7,326 were the intoxicated drivers themselves, and 1,594 were legally intoxicated pedestrians and pedal-cyclists. The remaining 3,972 fatalities were non-intoxicated drivers, passengers, and non-occupants. So how many actual victims of drunk driving were there in 2000? Excluding the 7,326 legally intoxicated drivers and 1,594 legally intoxicated pedestrians/pedal-cyclists, there remain 3,972 fatalities; but even these deaths cannot be classified as victims because the NHTSA does not indicate which driver was at fault. For example, if a sober driver runs a red light and crashes into a driver who has a BAC of 0.10 or greater, and the sober driver dies, the NHTSA will proclaim that this fatality is alcohol-related, even though alcohol had nothing to do with the crash, and tragically, the intoxicated driver will potentially face vehicular manslaughter charges. Law enforcement will wrongfully conclude that the drunk driver is at fault, and more likely than not, will not execute a thorough traffic scene investigation. Essentially, if alcohol is involved, it is disorderly to blame. Meanwhile, organizations such as MADD are spoon-fed this fatality, and they in turn inappropriately use it to mislead our lawmakers.

Unfortunately, the deception does not end there. According to the NHTSA, on an average, in more than 50 percent of the reported alcohol-related fatalities, alcohol involvement, as determined by actual alcohol testing, is not known. Alcohol test results may not be known for any of several reasons: the test was given but the results were not obtained by the Fatality Analysis Reporting System (FARS); the test was refused; FARS was unable to determine if test were given; or, the test was not given. As a result, the NHTSA imputes alcohol involvement in over 50 percent of the reported alcohol-related traffic fatalities. Imputation, as applied by NHTSA, uses characteristics of the persons involved in the crash to predict alcohol involvement when it is not known. Those characteristics include police-reported drinking, age, sex, restraint-use, type of crash, time of day, and driver of striking or struck vehicle. Sadly, these predicted, unsubstantiated, fatalities are masqueraded as confirmed victims of drunk driving, and have had an unjust role in the shaping of DUI laws across America, along with serving as justification to erode the Fourth Amendment protecting against unreasonable search and seizure without probable cause, by the establishment of sobriety checkpoints.

MADD, in pursuit of nationwide sobriety checkpoints, is partly responsible for the asterisk that was assigned to our Constitutions Fourth Amendment, which protects Americans from unreasonable search and seizure. In the majority ruling, (Michigan Department of State Police v. Sitz, 496 U.S. 444 (1990)) the United States Supreme Court Justice Rehnquist wrote, Drunk drivers cause an annual death toll of over 25,000. Thats simply not accurate data. Assuming that Justice Rehnquist was referring to the statistics of 1989, according to the NHTSA, there were 22,404 alcohol-related traffic fatalities, and as illustrated, not all these fatalities are the result of drunk driving as MADD and the United States Supreme Court contend. The Supreme Court referenced the enormity of the drunk driving problem in our nation to justify the appropriateness of its ruling; that the State has a grave and legitimate interest in curbing drunken driving; yet the statistics they referenced were skewed at the outset. It is curious why the Supreme Court rounded that particular statistic up rather than down, but regardless of their motivation, their claim doesnt match the statistics.

Additionally, the NHTSA utilizes a peculiar computation method to determine the percentage of alcohol-related traffic fatalities. Surely, most people have heard statements such as half of all traffic fatalities are caused by drunk drivers or nearly half of all traffic fatalities are alcohol-related or most recently, approximately 40 percent of all traffic fatalities are alcohol-related. Essentially what the NHTSA is claiming is that out of 100 traffic accidents in which a fatality resulted, 40 of them will be alcohol-related, but as illustrated above, not necessarily the result of drunk driving. For the sake of argument, lets assume that out the 100 traffic accidents in which a fatality resulted, there were indeed 40 drivers who were legally intoxicated. The quick conclusion is that 40 percent of the traffic fatalities are in fact alcohol-related, but that conclusion is incorrect. The best way to demonstrate this fallacy is with a simple analogy. If you have 100 packages of apples, and 40 of the apples are rotten, what is the percentage of rotten apples you have? If you answered 40 percent you are wrong. You cannot begin to solve this problem until you know how many apples are in each package. Lets try this problem again, but this time well provide the necessary information. If you have 100 packages of apples, with 4 apples in each package, and 40 apples are rotten, what is the percentage of rotten apples you have? Now the problem can be solved because we know that we have 400 apples and 40 are rotten, which means 10 percent of the apples are rotten; yet the NHTSA continuously divides drivers by accidents to develop its deceptively high percentages.

To accurately compute percentages there has to be a common denominator. The common denominator is drivers, and when evaluating the percentage of traffic fatalities that are alcohol-related, we have to consider all drivers in the traffic accident. Historically, the NHTSA indicates that approximately 55 percent of all traffic accidents are single-vehicle (SV) and 45 percent are multi-vehicle (MV). It stands to reason that a multi-vehicle (MV) accident involves a minimum of 2 vehicles, if not more. That being said, we will assume that all MV accidents had only 2 vehicles involved (realizing that some MV accidents have three or more drivers involved), bringing the number of drivers to a minimum of 90. Next we need to add the SV accident drivers (55) to the MV accident drivers (90) to arrive at a total of 145 drivers. To determine the percentage of traffic fatalities, we divide 40 drivers with a BAC of 0.01(g/dl) by 145 total drivers to arrive at approximately 27 percent.

The NHTSA persuasively categorizes pedestrian (also pedal cyclists) fatalities within the (SV) statistics, while excluding from its percentage computation, the obvious driver necessary for the pedestrian fatality. If the pedestrian is consider a driver for statistical purposes, then another driver has to also be included in the computation. Historically, approximately 12 percent of alcohol-related fatalities involve pedestrians / pedal cyclists. Consequently, 12 more drivers need to be added to our equation. (12 pedestrians (SV) / 12 drivers who struck and killed the pedestrian)

Considering the aforementioned, the percentage of alcohol-related fatalities is demonstrated as follows: 40 drivers with a BAC of 0.01(g/dl) divided by 157 drivers equals approximately 25 percent of traffic fatalities that are alcohol-related. Remember, 50 percent of these fatalities are predicted due to lack of actual alcohol testing. Additionally, approximately 15 percent of the alcohol-related traffic fatalities involve no driver or pedestrian who is legally intoxicated; that one or more of the participates had a measurable amount of alcohol in their blood, but were below the legal limit within their given states. That being said, realistically the percentage of substantiated alcohol-related traffic fatalities is approximated 12.5 percent, with approximately 65 percent of these fatalities being the driver themselves.

Certainly statistics vary from state to state, but rest assured that the same distortion is occurring wherever you may reside. Also, it is important to remember that approximately 65 percent of all alcohol-related traffic fatalities are in fact the driver themselves, so the actual percentage of innocent victims of drunk driving is minimal, compared to the leading cause of traffic fatalities on our nations highways: Speed! According to the NHTSA, in 2001 there were 12,850 speeding-related traffic fatalities. Speeding overwhelming claims more innocent victims than DUI; so why is there such a disparage between the punishment for speeding versus the punishment for DUI? If saving lives on our nations highways is the primary goal of our legislative and judicial branches of government, then a balanced approach to obtain that goal would result in the revocation of driving privileges for speeding violations along with mandatory prison sentences, hefty fines, safe driving classes, mental evaluations, community service; we could increase the list ad infinitum. Incidentally, the courts have embraced a so-called balancing test to support their rulings on sobriety checkpoints, so it doesnt seem out of the realm of possibilities to challenge the courts and our legislators on their unbalanced approach to highway safety related laws, policies, and punishments directly related to violations of such. Based on the NHTSA statistics, it is reasonable to expect then that a speeding driver is an equal threat, or an even greater immediate threat to public safety than a driver who may have a measurable amount of alcohol in their system, yet the gap between punitive response to each is unnerving. Either the DUI laws are extreme in nature, or the punishment for speeding needs to be completely revamped.

Obviously, if driving privileges were revoked for first offense speeders, civil unrest would erupt across the nation. There seems to be just such civil unrest developing in response to the current DUI laws. Over 1.5 million U.S. citizens are arrested each year for DUI. For over 20 years now, these citizens have fallen victim to the propaganda surrounding their crime, while their respective states inflict unjust punishments upon them. It should be anticipated that eventually the victims of such unjust laws would begin uniting to challenge the very underpinning used to justify the laws in the first place.

One such organization is called Responsibility in DUI Laws. http://www.ridl.us R.I.D.L. has imported the FARS database (same database used by NHTSA to report statistics) into a complex reporting tool called Crystal Reports. Through the use of this tool, R.I.D.L. was able to discover that across the entire nation, the NHTSA nearly doubles the number of instances of drunk drivers. And this is prior to them implementing their "Multiple Imputation" methodology, which adds in more numbers for "unknowns". In their reports, evidence is seen where the NHTSA claims a driver of a fatal accident to be a drunk driver even though that driver was tested and had a blood alcohol content of 0.00, or had a blood alcohol content that was within the legal limit.

To demonstrate the entire scope of this report, consider the statistics obtained from FARS and compiled by R.I.D.L. regarding traffic fatalities for 2001 in the Commonwealth of Pennsylvania. This specific report can be reviewed in its entirety at http://www.ridl.us, along with other state reports. Pennsylvania is a .10 BAC per se commonwealth. The statistics breakdown as follows: NHTSA lists total Pennsylvania fatalities for 2001 = 1,530 NHTSA lists total drivers in fatal accidents = 2,135 NHTSA lists drunk drivers = 575

R.I.D.L. report indicates the following:

Actual drivers with BAC over .10 = 294 Actual fatalities involving drivers with BAC over .10 = 332 Percentage of fatalities involving drunk drivers = 21.70% Number of drunk drivers listed with no proof = 281

Actual percentage of drunk driving in fatal accidents = 13.77%

Recall that earlier in this report, it was indicated that a more realistic percentage for the nation regarding traffic fatalities caused by drunk drivers would be approximately 12.5 percent, bearing in mind that approximately 65 percent of those fatalities were in fact the drivers themselves. Witness how that percentage corresponds closely with the results in the Commonwealth of Pennsylvania.

The significance of this analysis is that Pennsylvania arrests approximately 41,500 citizens each year for DUI. To put that in perspective, 41,500 citizens are criminalized each year in Pennsylvania because approximately 137 persons are killed in accidents involving a legally intoxicated driver, while an undetermined percentage of those fatalities were the fault of the sober driver. Compare that fatality rate to the deaths caused by speeding, and you must find need to pause and question the integrity of our current DUI laws. According to the Pennsylvania DUI Association, in 2001, municipal and state police conducted 302 suspicionless seizures described as sobriety checkpoints. The police contacted 114,959 citizens during these suspicionless seizures, while making only 807 DUI arrests. Alarmingly, the police made 2,227 other arrests while conducted suspicionless seizures under the disguise of sobriety checkpoints.

In conclusion, the states interest in eradicating DUI is no greater and legitimate then the states interest to investigate and prosecute any other ordinary crime. Consequently, suspicionless seizures known as sobriety checkpoints should be immediately recognized as blatant violations of the 4th Amendment of the United States Constitution, and as a violation of the heighten protections provided by individual state constitutions, prohibiting seizures by police without probable cause. A moratorium should be issued immediately prohibiting the employment of suspicionless seizures to enforce current DUI laws.

The courts across this nation have repeatedly compared sobriety checkpoints to checkpoints established by earlier rulings, allowing for such to identify improperly licensed and uninsured drivers along with checkpoints to detect people who may be in this country illegally. There is though, one significant difference between these types of checkpoints. Law enforcement cannot determine whether a driver of a vehicle is an illegal immigrant, or whether they are properly licensed and insured merely by their driving patterns; there are no specific driving habits or patterns that would prompt law enforcement to suspect the aforementioned. That is not the case when investigating a potential DUI. Police have a 20-point alert system that they use to evaluate the driving patterns of a potential DUI. Excessively wide turns, driving excessively right or left within lane, and crossing the centerline are a few of the indicators police currently employ to enforce the current DUI laws. Approximately 99 percent of DUI arrests are made utilizing such techniques.

Finally, although the computation method of imputation maybe useful to develop trends, it should be excluded from shaping laws. Supporters of the current DUI laws and those who advocate even tougher laws, claim that it is their efforts that are responsible for saving lives on our nations highways; yet even this claim cannot be substantiated. In the last 20 years, the leading cause for the reduction of the percentage of traffic fatalities can be attributed to one specific effort; the effort to build safer vehicles. Stronger frames, primary restraining systems (seatbelts) and supplementary restraining systems (airbags) are the leading life saving instruments according to the NHTSA, not necessarily tougher laws.

I invite all criminal defense attorneys and supporters of civil rights to freely distribute the content of this report to your peers, and encourage each to work together to bring about a much needed resurrection of our Fourth Amendment, along with a reformation of our current DUI laws.

Ed Haas is a freelance writer / author and founder of the Muckraker-Report.org
For reprint permission, contact Ed Haas via e-mail at: info@muckraker-report.org