Alcohol and Motors
SINCE the repeal of national prohibition the American people have been faced with a particularly unfortunate pair of circumstances making worse a situation which was bad enough before. The first is that a huge portion of our population is now engaged regularly in the admittedly hazardous activity of driving an automobile. The second is that intoxicating liquor has become cheap and very available. The most flagrant symbol of this dual danger is the roadside bar. Prior to prohibition many cities and states, now dripping wet, were dry and had been so for years. During prohibition a great and increasing number of persons were driving, but the liquor was not so palatable, so cheap, or so readily available.
A large number of us looked forward to repeal as a means of correcting a multitude of obvious abuses, not the least of which was the breakdown of the traditional attitude of respectable people toward drinking and even drunkenness. We feared the danger of increased automobile accidents, but we hoped and rather thought that the improved quality of the liquor, the assured supply, and the opportunity to buy it and consume it calmly in decent places would offset the disadvantages of quantity and ubiquity.
In thus thinking, we failed to consider the body of knowledge developed by workers in the medical and associated sciences on the inescapable physiological and psychological effects which ethyl alcohol has on man. I say we failed to take this information into consideration. Most of us knew nothing about it. We are finding it out now.
We are discovering that it is just as impossible to argue away the effects of alcohol on a human being as it is to legislate away an ancient, popular, and ingrown habit. If you expose yourself to smallpox or measles, you are liable to get smallpox or measles. Nobody disputes the fact. If you drink, you are not the same as you were before. And that cannot be controverted either.
The reader is probably thinking: ‘Of course this is true if you are considering a quantity of liquor such as most prudent people would not think of drinking.’ But unfortunately I am talking about the effect of an ounce and a half of ethyl alcohol or less, such as is contained in no more than three ordinary highballs. It is a cold fact — as I shall attempt to show — that the many drivers who have merely been drinking, as distinguished from the few who are patently drunk, present the real menace to safety on the highways to-day.
I
What is the effect of alcohol on the automobile driver? Not long ago a scientific inquiry into this problem was completed by Poul Bahnsen and K. Vedel-Petersen of the Psychotechnical Institute, Copenhagen. Their work was reported in detail in the Journal ofIndustrial Hygiene for September 1934. Other experimenters have also been busy, some of them in the United States, but for various reasons it would be well to mention the Danish experiments first. Danes are nowise different from Americans, and since ethyl alcohol is the same the world over, this study should throw a great deal of light on our problem over here.
Fourteen experiments were suggested to the two scientists — all contrived to determine the effects of alcohol on functions used in driving an automobile. They selected the four which tested functions having the closest similarity to those which play important parts in driving and which at the same time permitted an accurate measure of performance.
The dose was fixed at 0.6 cubic centimetres of absolute alcohol for each kilogram of body weight of the subject. In other words, a man weighing 150 pounds would receive about an ounce and a half of alcohol. Since whiskey was to be used in the familiar soda mixture, this meant that each subject would drink about three highballs. The dose was given on an empty stomach and the tests were made forty minutes later, because the concentration of alcohol in the blood normally reaches its maximum after this period.
Two groups, each consisting of more than fifty young men, served as subjects. Each group was divided about equally between students and chauffeurs. Many of them were moderate drinkers. The procedure in each experiment was the same. The first group was given preliminary practice and instruction in the experiment, then a first test without alcohol, and then, after the dose, a second test. The second or ‘control’ group followed the same programme except that the alcohol was omitted. The purpose of the second group was, of course, to check the results. To determine the effects of the alcohol, it was necessary to know how normal people would react to the successive tests in each experiment.
The first experiment was designed to measure the effects of alcohol on the speed with which a person makes a selective reaction to a stimulus. A selective reaction is one in which the person must pause to think before acting. For example, if one is driving an automobile and sights danger ahead, one is usually presented with several possible ways to avoid the danger. One may jam on the brakes. One may step on the accelerator. One may choose to swerve the car either to the right or to the left. The danger is the stimulus, and what the driver does is the reaction. It is up to him to select the reaction which will bring him through safely, and ordinarily he has to be quick in making up his mind.
The apparatus used in this test consisted of an oblong piece of cardboard lying on a table, with six circular holes in it of two-inch diameter. The holes were labeled in order from A to F, printed in plain black letters. Behind the cardboard was a square block of metal set on edge. In the face of the block was a slot in which one of the six letters appeared when a trap was sprung. Each time a letter appeared, the person being tested was required to put a stylus in the hole of the cardboard which carried the corresponding letter. The time required to perform this simple operation was measured by an electrical device.
Did the consumption of a few drinks slow up a person’s ability to react in this way? Yes, to the extent of 9.7 per cent, according to these scientists. In other words, if the average man with three drinks under his belt is driving at forty miles an hour when danger pops up ahead of him, he will travel from five to eight feet nearer that danger before deciding what to do about it than he would have traveled if he had not taken the drinks. Any experienced driver knows how very important that margin may be.
The second experiment was contrived to test a muscular reaction. In this there was no element of selection or discrimination. The subject sat before two iron bars suspended vertically. Each bar was about an inch and a half in diameter and a yard in length. A metal plate was struck with a hammer; this released the bars and the subject was required to grasp them, one with each hand, as soon as he could to stop their fall. The height of the fall was recorded on a fixed scale to measure the time of reaction.
The importance of quick muscular reactions is at once apparent to every motorist. If it is necessary to apply the foot brakes, how soon will the driver’s muscles respond? A tire blows out: how soon can he redouble the force of his grasp on the wheel to hold the car in the road? A delay of a fraction of a second might very well mean the difference between a mere flat, tire and a tragic wreck. Does the consumption of alcohol affect this vital ability? The answer again is, yes. In this experiment the muscular reaction of the ‘alcohol’ group was 17.4 per cent slower than that of the ‘control’ group.
The third test was devised to measure the subject’s powder to concentrate his attention. Three strips of white paper about five feet long and four inches wide were suspended vertically and contiguously along a wall. The centre strip was held stationary, while the one at the left moved up and the one at the right moved down. On the strips were painted thick horizontal black lines. As the two outside strips moved, these lines occasionally came abreast of one another to form a straight horizontal line across all three strips. The subject, who was seated back about nine feet, was required to touch a trigger each time this occurred.
Good performance, according to the two scientists, required continuous concentration. Should attention wander for only a moment, the concurrence of a set of lines might be missed. Moreover, the test required the ability to distribute attention, making it cover a large field rather than a single point; for if the eyes were fixed too long on a single place, other combinations would be overlooked. And the test called for the ability to anticipate the meeting of the lines. The subject did not suddenly see a concurrence and then react; in most cases a precalculation took place. Two horizontal lines moving at different speeds — one coming from above and the other from below — were seen approaching one of the lines in the middle stationary strip, and the question arose: Will the three lines cross simultaneously? In deciding this question a more or less conscious estimate of the distance between the three lines was made and the speed at which the two were moving. The picture as it appeared to the subject was ever changing; it consisted of a number of situations in development. Sometimes the subject missed a combination; and again he erred by touching the trigger when he expected a concurrence of lines which did not take place.
The importance of the ability to concentrate one’s attention on the various circumstances which constantly develop at busy street intersections is sufficient to convince any thoughtful person of the great value in determining what effect, if any, a few drinks of whiskey have on this mental attribute. One approaches a corner and sees a car coming from the right and another from the left. Will one have time to cross safely? What if the driver from the left decides to turn to his left? What if the driver from the right accelerates? What if a driver in a car parked at the curb starts up? Can one watch all these factors and judge distances and speed as well with three drinks in one’s stomach as with none? By no means. For in this experiment the subjects who had taken the liquor made 35.3 per cent more errors than those who had not.
The fourth and last experiment of this series was carried out as the subject stood bending over a rotating circular plate about a foot in diameter fitted with metal spikes which stood about an inch high. The spikes were distributed all over the plate. A red line was drawn among the spikes winding circuitously from the centre of the plate to the edge. As the disk rotated, the subject endeavored to move a metal stylus along the red line and within a boundary marked by a flattened wire hoop without touching spikes or hoop.
Here good performance called for the nicest coördination between the eye and the hand, a mastery of the muscles of the hand and arm, concentrated attention, and ’intellectually determined dexterity.’
Again the attributes which the test calls for are so important to the driver that little comment is needed. His ability to do what his eye tells him he needs to do, his control over his hands and arms, the power to concentrate, and the dexterity required to follow through traffic a path sometimes safe and easy, sometimes difficult and dangerous, are fundamental to good driving.
If it were found that an ounce and a half of ethyl alcohol affected adversely these important attributes as much as 50 per cent, would it not be clear that alcohol is deadly stuff in the body of a driver? Well, in this experiment the alcohol group of subjects made a record of 59.7 per cent more errors than the non-alcohol group.
II
Still another experiment, this one by an American, ought to be mentioned. Dr. H. A. Heise, of Columbia Hospital, Milwaukee, developed a test which was made while the subject was actually driving an automobile. The car was rigged up with a gun, so that without warning it would shoot a bullet on to the roadway while the subject was driving. This was the signal for the motorist to press the brake pedal. When he did so the gun would fire another bullet. The distance between the bullet scars was the measure of reaction time. It was found that at the driving rate of thirty miles an hour the five subjects when normal averaged sixteen feet between the marks left by the bullets, but after drinking from one to five ounces of alcohol the distance was twenty-two feet, or an increase of 37 per cent.
All these tests only go to corroborate facts which earlier work on the effects of alcohol had well established. Dr. Walter R. Miles, in his admirable chapter on the ‘Psychological Effects of Alcohol in Man,’ published in the volume Alcohol and Man, reviews the whole subject and presents exceedingly significant data. All of them do not apply specifically to driving an automobile, but they do show conclusively that even tiny quantities of the drug have a measurable effect on the human mechanism.
To demonstrate the effects of a very small quantity of alcohol even after the lapse of considerable time, one experimenter, for two weeks at ten o’clock each morning, threaded needles for twenty minutes. He could do 180 in that time. Then at eleven o’clock at night on the fourteenth day he consumed twenty-five cubic centimetres of alcohol (about an ounce and a half of whiskey) in seventy-five cubic centimetres of liquid. The following morning at ten he threaded needles again, eleven hours after drinking. The drinking and threading continued for ten days, followed by nine more days with no alcohol and finally ten more alcohol days. Despite the small quantity of liquor and the eleven-hour delay, his efficiency in threading needles was 6 per cent less after the doses of alcohol.
Two experimenters desiring to find the immediate effect of increasing quantities devised a ‘dotting machine’ which consisted of a moving tape on which small circular dots were printed. The subject was required to prick the dots with a stylus. Between one and two hours after taking ten cubic centimetres of alcohol, the efficiency of the subject (a very intelligent woman) decreased 21 per cent from normal. Experiments with fifteen cubic centimetres resulted in a decrease of 42 per cent; a dose of twenty cubic centimetres cost 39 per cent in the efficiency of eye and hand coördination; and tests with twenty-five cubic centimetres showed increases in errors averaging 113 per cent.
Exhaustive tests made of marksmanship and typewriting confirmed these findings. Tests of speech, of speed of eye movement, of body sway as the subject stood still with his eyes closed, all revealed decreased motor control after the drinking of very moderate quantities of alcohol. Tests showed that learning was slower after drinking and memory less sure; that ability to distinguish between noises and between colors diminished after drinking. Tests of attention and concentration confirmed the findings of the Danish experimenters. Tests of reasoning and thinking showed measurable losses in ability.
III
Three methods are now being employed by enforcement officers to determine alcoholism in drivers. According to most of the physicians and psychologists who have studied the problem, alcoholism can best be determined and proved in court by an analysis of a body fluid. On the other hand, Dr. Theron W. Kilmer, police surgeon at Hempstead, Long Island, in a recent address before the convention of the International Association of Chiefs of Police, expressed the opinion of many when he declared that such accuracy is not necessary, but that a general examination by a physician is all that is needed. Then there are hundreds of policemen in the country who have been more or less successful in securing convictions with only the testimony of the arresting officers.
A theoretical but interesting objection to the first and most scientific of these procedures is that a driver, regardless of the state of his intoxication, may not be required to give a sample of a body fluid. This is because of the basic law which provides that a suspect need not testify against himself. In actual practice, however, this is not a formidable objection, because drunken people are usually in no state to raise constitutional questions.
Dr. Heise, who became interested in the effects of alcohol on drivers while at Uniontown Hospital, Pennsylvania, reports the results of using a chemical test on drivers in Fayette County. ‘In 1924,’ he says, ‘before a chemical test was used, fifty-five persons were accused of drunken driving. Of these only twenty-one, or 38 per cent, were found guilty or pleaded guilty. In one hundred later consecutive cases in which urinalysis demonstrated that the drivers had been drinking, eighty-seven of the one hundred drivers were found guilty or admitted guilt, while six were pronounced not guilty, but were directed to pay the court costs (meaning, “You are not guilty, but don’t do it again"), and only seven were acquitted.’
In Butler County in the same state, according to both Dr. Heise and E. H. Negley, district attorney, it was almost impossible to get a conviction for drunken driving in 1931 before urinalyses were used. Acquittals on the basis of insufficient evidence were the almost invariable rule. Then a local physician equipped himself with apparatus for urinalysis. During 1932, practically all cases continued to go to court on pleas of not guilty, but in twenty of the first thirty-six cases the test was used and resulted in twenty convictions. Later convictions became so sure that almost all of the accused admitted their guilt. For example, in September of 1934 there were sixteen drunken-driving cases on the docket; no fewer than fourteen of the suspects promptly entered pleas of guilty.
Dr. Kilmer advocates a simpler method. He uses a record card on which he makes detailed notes of the condition of the defendant, and refers to these to refresh his memory in court. When the driver suspected of being ‘under the influence’ is first brought before him the opening question is: ‘Is there anything the matter with you?‘ The common reply, he says, is ‘Not a thing!’ This reply, entered on the card, averts any claim in court that the driver had been in a daze because of illness. The next question: ‘Have you been taking medicine?’ always brings a negative reply, and it then becomes impossible for the defense attorney to claim later that his client was lightheaded because of too much medicine. The doctor’s examination then covers such items as the driver’s temperature, the condition of his pupils, his reaction to light, the tremors of his hands, his ability to touch his forefinger to his nose with his eyes closed, his ability to read and tell time, to walk across the room and turn around and walk back again, to pick up a coin from the floor, his general physical and mental condition.
In defense of the accuracy of Dr. Kilmer’s method, it must be said that when Dr. Heise was making his urinalyses of two hundred persons charged with drunken driving, using both a routine examination not unlike Dr. Kilmer’s and the chemical test, the correlation between the two was so obvious that his colleague attempted to predict the amount of alcohol to be found in the urine from the physical examination alone. The agreement was accurate except in one case, in which the estimate of alcohol content was .30 per cent when the actual amount was only .18 per cent. Later a brain injury was found which accounted for the man’s peculiar reactions.
And now come the policemen, who say their own officers can detect drunkenness and testify successfully in court. The attitude of some of these officers is changing. Lieutenant Franklin M. Kreml of Evanston, Illinois, recently said, ‘I have always considered the drunken-driving question as a sort of “alibi” for the police to hide behind — to distract public attention from inefficient enforcement of traffic rules as they are violated by those who are cold sober; but now I and others are learning that the drunken driver does not present the serious problem in traffic — it is the drinking driver we must detect and discipline. Police who continue to rely on the arresting officer for testimony will not be successful, I am afraid.’
Since it is the drinking driver who presents the real problem, it would seem that the accurate determination of the effective alcohol in a man’s system — that is, the amount in his body fluids — would be the fair and conclusive way to decide whether or not he was unfit to drive. Drunken driving could then be defined as driving when the blood, or urine, or spinal fluid, or even the breath (if apparatus is perfected) contains more than so much alcohol per unit volume. This appears all the more feasible when one considers that there are numerous pathological conditions such as the brain injury mentioned above which create symptoms easily mistaken for intoxication even by physicians. This method also takes account of variations in tolerance to alcohol. It is well known that one man can drink more than another without becoming intoxicated and that at times a man can drink more than he can at other times. Scientists have found that there is a closer correlation between the percentage of alcohol in the body fluids and intoxication than between the amount consumed and intoxication. In other words, once the most capacious drinker absorbs as much alcohol into his blood or urine as the man with the least tolerance to the drug, he shows the same evidences of intoxication as his much-despised brother.
IV
The national prohibition laws were rescinded in December 1933. It is now time to inquire into the effect of repeal on the motor accident rate. Did more people drive after drinking in 1934 than in 1933? And did more accidents involving drinking drivers and pedestrians occur in 1934 than in 1933? Complete statistical answers to these questions are not possible at the time of writing this paper, but the data available are adequate for the practical purpose at hand.
Of twenty-three cities sending special reports to the National Safety Council last fall, twelve indicated that arrests for drunken driving had increased in 1934 as compared to 1933; seven presented figures to show that accidents involving drunken drivers or pedestrians had also increased; two reported that arrests had decreased; and three had no information on the subject.
Arrests for drunken driving during the period from January through October 1934, as compared to the same months of 1933, increased 80 per cent in Duluth, Minnesota; 13 per cent in San Antonio, Texas; 100 per cent in Rochester, New York; 90 per cent in Erie, Pennsylvania; 49 per cent in Richmond, Virginia; 29 per cent in Syracuse, New York; 41 per cent in Elizabeth, New Jersey; 37 per cent in Newark; 39 per cent in Springfield, Massachusetts; 15 per cent in Chicago; and 25 per cent in Milwaukee. An increase of arrests for drunken driving in Cleveland of several hundred per cent may indicate greater vigilance on the part of the police as well as an increase of the dangerous malpractice.
Berkeley, California, for the same period reported that accidents involving drunken drivers increased 14 per cent, and accidents involving drunken pedestrians 125 per cent. Buffalo showed an increase of 55 per cent in the number of fatal automobile accidents in which drivers under the influence of liquor figured. In Chicago the fatal accident experience of drunken drivers showed an increase of 330 per cent, while accidents of drunken pedestrians increased exactly 500 per cent. In the month of September 1933, by the way, 22 drunken drivers were involved in two fatal and 20 non-fatal accidents in Chicago, while in September 1934, 76 drunken drivers were involved in two fatal and 74 non-fatal accidents. In September 1933, 15 drunken pedestrians figured in accidents on Chicago streets; in September 1934, 66 drunken pedestrians were struck by cars.
Detroit experienced a 90 per cent increase in accidents involving drunken drivers and a 52 per cent increase in accidents to pedestrians during the first nine months of 1934 as compared with the same period of 1933. Neither St. Louis nor Pittsburgh presented figures on the increases in accidents involving drunken drivers for the year, but the former reported that while in 1933 intoxicated persons figured in 10.4 per cent of all automobile accidents, in 1934 they figured in 11.8 per cent of the cases. In Pittsburgh, accidents involving drivers who were definitely ’under the influence’ amounted to one per cent of all accidents in 1933, while in 1934 they amounted to 4.5 per cent of the total. A corresponding increase from 2.5 to 6.5 per cent was noted in mishaps involving drunken pedestrians.
Los Angeles was the only city to report a reduction in fatal accidents attributable to intoxicated drivers — a reduction of 17 per cent. But there was an increase in the number of intoxicated pedestrians killed of no less than 600 per cent.
Baltimore reported a decrease in arrests for drunken driving of 39 per cent in 1934 as compared to 1933, and Wilmington presented figures to show that such arrests decreased 12 per cent.
A report from Massachusetts shows that the registrar of motor vehicles suspended 4675 licenses for driving under the influence of liquor during the first ten months of 1934 as compared to 3409 suspensions for the same period of 1933. This represented an increase of 38 per cent.
All this information, fragmentary though it is, would seem valid proof that driving while under the influence of intoxicants has increased since repeal. But how does alcohol compare with other factors as a cause of accidents? We have already seen that in St. Louis and Pittsburgh accidents involving drunken drivers were only 11.8 per cent of the total, or less. R. L. Fomey, statistician of the National Safety Council, reports that fatal accidents involving drivers described as ‘had been drinking’ increased 29 per cent in 1934 as compared with the previous year, and fatal pedestrian accidents 53 per cent. In non-fatal accidents drivers reported under the influence of liquor increased 50 per cent and pedestrians 65 per cent. Even so, however, such accidents amounted to only 7½ per cent of the total for 1934. Mr. Forney, as well as state motor vehicle commissioners, registrars of vital statistics, and city police officials, are convinced that these figures represent, for the most part, only cases of flagrant intoxication, since there is as yet no widely used method for determining whether a driver, though not drunk, has been drinking.
Evidence of woeful under-reporting of alcohol cases comes from many sources. For example, Dr. Heise says that he made an analysis of 119 automobile accidents which resulted in injury or death to 216 persons. These persons, he says, represented a fair distribution of urban, rural, and tourist drivers. The study revealed that 60 per cent of the accidents in which 75 per cent of the injured or killed met their fate involved drivers who had been drinking. It was also found that the ‘alcohol’ accidents were a great deal more severe than the ‘non-alcohol’ accidents; for the latter resulted in the injury or death of about one person per accident, while the ‘alcohol’ accidents accounted for two victims each.
The conclusions to be drawn from this article are inescapable. Drinking is a social custom. It is a style — right now unhappily in the mode. Thus the responsibility of leaders is particularly great. One invites twelve friends in for the evening. They all come in cars. Should cocktails be served? People of influence and strength must set a proper example if drinking driving is to become in the popular mind as reprehensible a practice as it actually is. The problem is one which is more likely to find its solution in law observance than in law enforcement. People must acquire the habit of not drinking if they intend to drive, or of not driving if they have been drinking. Influential citizens can also do much to spur police on to more efficient handling of the incorrigible few who, undeterred by social disapproval or their own conscience, will drink and drive.