The Police Officer as a Drunk Driving Witness

A study by the National Highway Traffic and Safety Administration [U.S. Department of Transportation Report No. H5-801-230] points out the effect of these individual differences on an officer’s observations and conduct in a DUI investigation:

The officer’s age and experience play a role in his alcohol-related arrest decisions. Younger officers, and those with relatively few years of seniority, tend to have a more positive attitude toward alcohol-related enforcement and make more arrests on that charge than do older officers. This result was found to hold true regardless of the type of department in which the officer serves or the specific type of duty to which he is assigned.

The officer’s personal use of alcohol is inversely related to his level of alcohol-related enforcement. Patrolmen who drink make significantly fewer arrests than those who do not, and those who drink frequently make significantly fewer arrests than those who use alcohol only occasionally.

Lack of knowledge concerning the relationship between alcohol and intoxication is widespread among police officers and imparts a negative influence on alcohol-related enforcement. Most officers underestimate—often by a wide margin—the amount of alcohol a suspect would have to consume in order to achieve the statutory limit of blood-alcohol concentration. This seems to induce a tendency among officers to identify and sympathize with the suspects they encounter.

Specialized training has a strong positive influence on alcohol-related arrests. Patrolmen who have received instruction in the operation of breath testing devices and/or in alcohol-related enforcement—particularly in municipal departments—were found to lack this specialized training.

Specialization in duty assignment can also enhance alcohol-related enforcement. Patrolmen assigned to traffic divisions, in particular, produce higher arrest rates than those charged with general patrol duties.

Near the end of the duty shift, alcohol-related investigations decrease substantially. This is particularly true in departments that have adopted relatively time-consuming procedures for processing alcohol-related arrests.

Weather conditions also affect alcohol-related arrests. There is encouraging evidence that foul weather has a positive influence on the attitude of many officers; they are more appreciative of the risk posed by an alcohol-related suspect when driving conditions are hazardous, and are less likely to avoid the arrest when those conditions prevail.

The suspect’s attitude can have a strong influence on the arrest/no arrest decision. If the suspect proves uncooperative or argumentative, a positive influence for arrest results. Conversely, the likelihood of arrest decreases when the suspect seems cooperative.

The suspect’s race is a key distinguishing characteristic in alcohol-related cases. The officers surveyed—the overwhelming majority of whom were white—reported releasing significantly more nonwhite suspects than they arrested. The data do not suggest that this reflects a greater tendency to exercise discretion when dealing with nonwhite drivers. Rather, the officers seem more willing to initiate an investigation when the suspect is not of their own race.

Suspect’s age is another distinguishing characteristic of these cases, and patrolmen reported releasing significantly more young suspects than they arrested. This appears to stem from two distinct causes. First, young officers exhibit more sympathy for young suspects, i.e., seem less disposed to arrest a driver of their own age group. Second, older officers seem more willing to stop young suspects, i.e., are more likely to conduct an investigation when the driver is young, even if the evidence of alcohol-related violation is not clear.

Suspect’s sex also plays a role in the arrest/no arrest decision. Patrolmen seem more reluctant to arrest a woman for alcohol-related violations, largely because processing of a female arrestee is generally more complex and time consuming.

How effective is the officer as drunk driving witness – and specifically at detecting intoxication in DUI investigations? The general consensus, of course, is that DUI training and experience will significantly improve the ability to detect symptoms of inebriation. A study by scientists at Rutgers University’s Alcohol Behavior Research Laboratory, however, questions whether officers are any better at determining if a driver is under the influence than a random juror might be. [See Langenbrucher & Nathan, Psychology, Public Policy and the Evidence for Alcohol Intoxication, American Psychologist 1070 (Oct 1983). The series of experiments was designed to test the ability of social drinkers, bartenders, and police officers to estimate the sobriety of individuals.

In the first experiment, lay witnesses were tested — 49 individuals who were characterized as “social drinkers.” While these individuals watched, four test subjects — two men and two women — were “prepared” by having them consume varying amounts of alcohol until they reached different levels of blood alcohol; in some cases, no alcohol was consumed. Each subject was then brought into a room where the “witnesses” were seated and interviewed at length to obtain a range of verbal behavior and somatic and cognitive effects. The subject would then leave, his or her walking and balance under constant observation. After later reviewing the opinions of the witnesses, the scientists concluded:

The assumption that social drinkers would prove to be accurate judges of the (blood-alcohol levels) of other persons was not confirmed. On only 4 of 16 occasions did a significant number of subjects correctly classify a target on a three-stage categorical index of intoxication level. If determining whether a man is sober or intoxicated is a matter of common observation, then our subjects apparently lacked this capacity. [See Langenbrucher & Nathan, Psychology, Public Policy and the Evidence for Alcohol Intoxication, American Psychologist 1070 (Oct 1983).]

In the next experiment, the scientists considered a type of witness with more expertise in the area: bartenders. They chose 12 bartenders and then conducted the test in the familiar environment of a large hotel cocktail lounge. The results were, however, familiar:

The bartenders correctly rated a target in only one of four instances. Contrary to expectation, no relationship between years of experience as a bartender and (blood-alcohol level) estimation accuracy was found. These data suggest strongly that these bartenders did not possess and had not acquired special knowledge or skill in identifying intoxicated persons.

Finally, the psychologists turned their attention to the experts: 30 police officers from a number of New Jersey law enforcement agencies with substantial DUI experience. Fifteen of these were tested under conditions similar to those used in the first two experiments; the remaining fifteen were tested under “field conditions”: at night, on roads with a driving subject who is asked to get out of the car and perform a series of field sobriety tests. The results of this experiment should be of considerable interest to any  DUI defense lawyer representing a client charged with driving under the influence of alcohol:

When police observers in the laboratory condition were compared to social drinkers who had experience an identical procedure, no difference in rating accuracy was found. Officers in the arrest analogue condition were somewhat more accurate than their colleagues in the laboratory condition but not significantly so. The scientists concluded:

The results of the three experiments described here are not reassuring. All three of the groups studied—social drinkers, bartenders, and police officers—correctly judged targets’ levels of intoxication only 25 percent of the time.

The California DUI defense firm should consider these studies when cross-examining the “expert” officer in drunk driving cases.