In Mercado v. Ahmed, 756 F. Supp. 1097 (N.D. Ill. 1991), the district judge found that Stan Smith's proposed testimony, essentially similar to that admitted in the case now before us, was inadmissible as lacking general acceptance of the validity of his methods among economists.Summary of this case from K.M. Leasing, Inc. v. Butler
No. 86 C 6986.
February 8, 1991.
James T. Crotty, Zacarias R. Chacon, Crotty and Hoyne, Chicago, Ill., for plaintiff.
Richard French, Michael Ortyl, French, Kezelis Kominiarek, Chicago, Ill., for defendants.
Is the testimony of an economist on the cash value of the lost pleasure of life admissible?
We ought to begin inquiry into any proposed "scientific, technical, or other specialized knowledge," Fed.R.Evid. 702, by asking two questions. First, if the masters of the knowledge were to examine a set of facts, would all the masters come to the same conclusion? If they would their opinions are reliable. If the opinions can be verified or proven in some way the opinions are valid. If all structural engineers opine that the East Bridge will not stand against any hurricane, their opinions are reliable. If the bridge stands through the hurricane, their opinions are reliable but invalid. If the bridge falls with the first hurricane, their opinions are reliable and valid.
Were complete reliability and validity the only hallmark of science, there would be precious little science. Not only is it difficult to achieve reliability and validity, it is often difficult to decide what shall constitute reliability and validity.
Reliability is the simpler concept: It requires only that all experts would say the same things under the same circumstances. But who are the experts? Are all holders of Ph.D. or M.A. degrees in a given field experts? Or are all with "X" years of experience in the field, or all who pass a certifying examination experts? Who decides who are the experts?
How do we test the proposition that all experts would hold the same opinion? These questions are answered by consensus.
Life in a courtroom teaches there is very little reliability; equally qualified experts testify to propositions in utter contradiction. This teaching is flawed. Much science is reliable. Where nearly all who master some special knowledge are in agreement, their opinions are rarely heard in court. Agreement ends conflict, and the courtroom is an arena of conflict. Reliable opinions quietly enter courtrooms in the form of stipulations of fact, the uncontradicted testimony of the fingerprint expert, or the questioned document examiner. Or, in some cases, courts may belittle themselves by excluding reliable expert evidence such as courts did for too many years in paternity cases.
Validity is a notion bristling with thorns and snares, among them the imponderable question of how we know anything with certainty.
The provisional solution has been an agreement (or agreements) that verification of science is to be found in the accuracy of its predictions. A scientific principle is valid if its use tells us that some defined thing will occur and we perceive that it does occur. The bridge falls in the hurricane. Sometimes prediction is subtler: A theory explains a past occurrence and predicts that we will, if we look in a certain place, find proof of this. The scientist looks for and detects the background radiation that would have resulted if the universe began with a big bang, as one theory proposes. This reliance upon prediction as validation comes from the positivist tradition in philosophy. This is today the accepted method for deciding what is science, what is scientifically or technically known. But it may not be a comprehensive standard for all belief and knowledge.
Quine wrote (in language used by many professional philosophers): "From impacts on our sensory surfaces, we in our collective and cumulative creativity down the generations have projected our systematic theory of the external world. Our system is proving successful in predicting subsequent sensory input. How have we done it?" W.V. Quine, The Pursuit of Truth, at 1 (Harvard 1990).
Hans-Georg Gadamer, a philosopher in some vogue with legal academics, says, in essence, that the highest form of knowing in the human or social realm is not discovery or explanation of fact but understanding; and understanding is not achieved by following a set method, but rather consists in moving to one's place in the continuing tradition. This understanding is like that of legal (or theological) interpretation — applying given rules and texts to new cases to create new texts and rules. Gadamer's high regard for legal interpretation may explain the high regard in which he is held by some legal interpreters. See H.-G. Gadamer, Truth and Method (J. Weinsheimer D. Marshall trans. 2d rev. ed. 1989).
Prediction and its confirmation are often hard to find. Most scientific theories arise from the human thought in response to data perceived by imperfect human senses. Even data measured by machines is (and may always be) inexact. Validation becomes more difficult when science concerns itself more with things that we do not sense, things such as quarks, gravitors, and strange attractors, whose existence emerges at the end of a long chain of inference. A scientific theory is often proposed because it explains existing data. Explanation is not enough. Many discredited, even fanciful, theories explain what we have seen, heard, done, or felt. If explanatory power were enough to make science, then witchcraft, astrology, and paranoid delusions are science. It is possible to provide many alternative explanations for a set of past events. One of the most profound challenges to the admission of psychiatric testimony is that psychiatric theory often has only the power to explain, and psychiatrists have expressed the view that they cannot and ought not to be predictors of human conduct. Indeed, there are several explanatory psychiatric theories.
C. Geertz, Local Knowledge, at 78-79 (Basic Books 1983).
See brief of amicus curiae for the American Psychiatric Association in Estelle v. Smith, 447 U.S. 934, 100 S.Ct. 3036, 65 L.Ed.2d 1129 (1980).
Predictions that a bridge on Florida's southeast coast will fall in a gale is likely to be quickly and easily tested. But many scientific theories (and usually the most important ones) predict a very large number of events. We cannot be really sure of theory validity until all the theory's predictions are tested. Classical physics explained the physical world and predicted innumerable events. Its predictions were consistently verified. At the turn of this century, classical physics explained and predicted everything except black body phenomena and the result of Michelson-Morley experiments about the speed of light. To explain and predict these two phenomena, classical physics was ultimately subordinated to quantum physics and the theories of relativity. The point is that validation never ceases. New data might always explode accepted theory. Indeed we have never decided what minimum of validation is required for provisional acceptance. A single true prediction can destroy a competing scientific theory. A single true prediction will seldom validate a theory. Neither law nor science accepts the maxim veritas in uno, veritas in omnibus.
The tentative nature of all theory has never stopped us from relying on theory because it is pragmatic to do so. Some knowledge is better than none. Classical physics still works quite well for ordinary mechanics. Astronomy went from a theory of spheres (Eudoxus) to a theory of circles around the earth (Ptolemy) to a theory of circles around the sun (Copernicus) to a theory of ellipses around the sun (Kepler). Yet a sailor teaching a child the phases of the moon and navigation by the north star has had no need to alter his essential teaching for over three millennia.
and epicenters and epicycles, etc.
The sailor's lore would be valid and Eudoxus, Ptolemy, Copernicus, and Kepler would have agreed with all of it and with each other. Their disagreements (the unreliability) arises elsewhere. It is true the way in which science develops over time is seriously debated. An influential view is found in T. Kuhn, The Structure of Scientific Revolutions (2nd ed. 1970). Even under Kuhn's view the good scientific theory is simple, consistent, broad, accurate, and a fount of prediction — for Kuhn these may not be the crucial virtues but they remain essential. See Kuhn, "Objectivity, Value Judgment and Theory Choice" (Lecture 1973).
The court, like the sailor, is interested in what is reliable and valid for its need.
Over many years judges have paid too little attention to these twin requirements of reliability and validity. The reasons for this, I suggest, are several.
One is the Frye test which admits as science that which scientists agree is science. Frye v. United States, 293 F. 1013 (D.C. Cir. 1923). By adopting this rule, courts signaled their dependence on the "community" of experts to decide admissibility. The Frye rule does not reject reliability and validity as standards for science. Indeed, in practice the scientific community presumably applied these tests and informed the court of its findings.
Two is the fact that some expert opinion evidence is admissible under longstanding custom. For example, psychiatric opinion of insanity is admissible because anyone's opinion of insanity is admissible given an adequate basis for the opinion. When the evidence was first admitted, it was not examined for reliability or validity.
Three, courts tend to devalue reliability as a guide to admissibility. Judges are accustomed to hearing diametrically opposed statements of purported truth. The absence of reliability in eyewitness accounts of facts, opinions about motives and so forth is part of the human condition — elimination of the unreliable is one goal of science. Where it can be eliminated, we can have science; where it cannot, we have no science.
Four, courts often ask juries and themselves to decide difficult questions of fact. Under the stress of this task, the trier of fact will sometimes entertain evidence in the hope it will ease the task. From a formal perspective there may be little harm in this practice. The law's formal solution to the undecidable fact question is found in burdens of persuasion. Where the fact question is too difficult to decide with any degree of confidence, the party with the burden of persuasion loses. This formal solution is unpopular in practice. It is a last resort in civil cases. Also, it is applied very rarely in cases of credibility conflict. In my experience jurors are far more likely to reach an impasse than to say that they cannot decide who is telling the truth and return a verdict for defendants in a civil case. Judges and juries use their common sense and experience to decide, in light of all the evidence, which scientist is correct. With scientists as well as lay witnesses, judges and juries feel obligated to come down on one side or another to reach a verdict by deciding whom to believe rather than reaching a verdict by believing no one.
Courts and scientists seek truth. This they have in common. Courts do more than seek truth; they resolve disputes in society. Indeed they seek truth only to resolve the dispute. Resolving disputes is the court's primary purpose, and the certain discovery of truth is not indispensable to ending disputes. Courts are satisfied if a claim is shown to be somewhat more likely to be true than not. Courts are not to be deterred from making a decision because the matter is beset with doubt. The pure scientist seeks truth for its own sake and admits as truth only that which is relatively free from doubt.
The dividing line between what courts do and what scientists do may not be absolutely clear at all times, in all contexts. Yet judges and lawyers rarely regard their work as science and, in this respect, nearly all scientists agree with them.
The practice of physicians is a useful example. As scientist, the physician explores anatomy, biochemistry and so forth. Reliability and validity are often achieved. As healer, the physician tries to resolve problems presented by disease and defect. These resolutions are sometimes scientific, sometimes not. It is not uncommon for physicians' opinions about diagnosis and treatment to be less than reliable and valid. This is why physicians often say (with less truth today than in the past) that medicine is an art, not a science.
We do not practice science in court and we do not insist that our expert witnesses do either. Rule 702 allows us to hear not only scientific knowledge but "scientific, technical, or other specialized knowledge [which] will assist the trier of fact to understand the evidence or to determine a fact in issue." Fed.R.Evid. 702.
The twin hallmarks of reliability and validity ought to serve as a standard against which experts' evidence is to be measured. It may be too much to ask that the courtroom expert be adept in a discipline with the reliability and validity of parts of inorganic chemistry. But it is reasonable to inquire how close the expert's discipline is to achievement of reliability and validity and to take this into account in deciding whether to admit the expert's opinion. Absent some degree of reliability and validity, we may have words that sound "scientific, technical, or . . . specialized" but we will not have knowledge.
It can be argued that courts ought simply to admit opinions of experts without any inquiry whether the area of expertise is a field of knowledge. If the expertise is flawed it can be met with evidence that proves these flaws. For example, the prosecutor in a criminal case might counter a defense psychiatrist with another psychiatrist to testify that any psychiatrist, including the defense psychiatrist, does not know enough to testify with any degree of certainty. Courts will sometimes defend the admission of expert testimony because the other side can defend itself by calling its own expert. Sherrod v. Berry, 629 F. Supp. 159 (N.D.Ill. 1985). As a principle that justifies admission of evidence, it goes too far because it would permit an astrologer to take the stand and say, "Smith, an Aries, could not have had (or must have had) an intent to defraud on November 18." Surely another astrologer could be found to disagree or to say that astrology is not knowledge.
In theory, astrology could be a science if reliability and validity were established, so could graphology (discovering mental states from handwriting), or phrenology (discovering character from conformation of the skull). The door to science's house is open to anything reasonably reliable and valid. Much of modern science rests on premises that would have been (and, in some cases, were) laughed at within our century. A good deal of medicine has become science only in the latest decades by using statistical techniques, and systematic studies and by improving methods of observing the operation of the living, uninvaded human body.
I conclude then that before I admit the evidence of an expert in a discipline whose acceptance by courts has not been established by binding precedent, I should assess the degree to which the discipline is characterized by reliability among its practitioners and by validation of its theories. The higher the degree of reliability and validity, the more likely the evidence will be admitted. Neither courts nor scientists agree what precise degree is required for entry into the canon of science or admissibility. I would be satisfied if we excluded what we sometimes admit, that is, expertise with virtually no reliability and no validity.
The Seventh Circuit's decision that affirmed the admissibility of this evidence, 827 F.2d 195 (7th Cir. 1987), was vacated, at 835 F.2d 1222 (7th Cir. 1988). The Seventh Circuit en banc then reversed the District Court on other grounds and stated: "We leave [the district court's other evidentiary rulings] for the district court on remand to decide in light of this court's prior discussions of those matters, specifically those found in our earlier vacated opinion." 856 F.2d 802, 808. Those discussions are not binding precedent.
There is a price to be paid for this and, as the required degree of reliability and validity increases, so does the price. The price is the exclusion of evidence shown by later studies (perhaps years later) to have been scientific. This price ought to be paid and courts have so said. In fact, it is inherent in the application of Frye. Courts will lag behind scientists as far as science is concerned. The risk to justice from pseudoscience is substantial, and we avoid this risk by requiring some showing of reliability and validity either by direct proof or by proof of acceptance by the appropriate scientific community.
I do not suggest that the inquiry into reliability and validity is by itself sufficient. The witness must be "qualified as an expert by knowledge, skill, experience, training or education." Fed.R.Evid. 702. The knowledge must also "assist the trier of fact to understand the evidence or to determine a fact in issue." Id. The ability of the opposing party to call an expert with a contrary opinion may protect against bad science — the incompetent use of a real science — but it does not offer adequate protection against false or unproved science.
Reliability and validity are good tests for any expertise, e.g., the police officer explaining drug or gambling business argot and procedure or the mechanic explaining the operation of a thresher.
The question today is whether I ought to interpret the law to permit an economist to testify as an expert on the monetary value of the pleasure of Brian Mercado's life. Brian Mercado is about eleven years old. He has significant mental and emotional deficits. He will likely never be able to hold a job or live alone. Psychiatric treatment is likely to be a consistent necessity. Brian Mercado was in kindergarten when he went with his family to visit the Museum of Science and Industry. In the parking lot he was struck by a taxicab and was injured. The jury found the taxicab driver was negligent. Assuming that the taxicab accident caused Brian Mercado an injury which turned him from a normal child into a severely disabled one, there is no question that Brian Mercado has lost a considerable degree of the pleasure of life. Assuming that Illinois law allows recovery for this lost pleasure of life the jury would have to quantify the dollar value of this loss. Brian Mercado offers an economist whose testimony I heard outside the jury's presence and who is prepared to opine that this value in Brian Mercado's case lies between $1,500,000 and $2,500,000. The taxicab driver and the taxicab company say that this is not admissible.
The jury did not find this to be so. There was medical testimony and testimony by Brian Mercado's kindergarten teacher that Brian Mercado's woes (and they are quite real) existed before the taxicab struck him. The jury accepted this evidence against contrary testimony of other physicians and members of Brian Mercado's family.
Sometimes called hedonic damages.
The testimony here does not depend on regression analysis which is often the basis of the correlations made by social scientists. Regression analysis is an accepted method of scientific investigation to establish correlations by controlling all factors which cause some particular state of facts. In practice the use of the method is flawed by poor basic data, unstated assumptions and unskilled application. Some think that regression analysis can never really work. A. Wildavsky, "What Would One Have to Believe to Believe in Multiple Regression", 27 Contemp. Psychology 903 (1982).
This kind of evidence is well described in T. Miller, Willingness to Pay Comes of Age: Will the System Survive?, 83 Nw.U.L.Rev. 876 (1989). In brief, Miller notes that economists are researching the "ways to measure the value that individuals place upon reducing the risk of dying" by examining the markets. Id. at 878-79. They examine "what people actually pay — in dollars, time discomfort, and inconvenience — for small reductions in health and safety risks." Id. at 879. Of particular significance, economists have estimated the values people place on risk reduction based on the following factors: 1) the extra wages employers pay to induce people to take risky jobs; 2) the demand and price for products — such as safer cars, smoke detectors, houses in polluted areas, and life insurance — that enhance health and safety; 3) the tradeoffs people make among time, money, comfort, and safety — in studies involving pedestrian tunnel use, safety belt use, speed choice, and drivers' travel time; and 4) surveys that ask people about their willingness to invest money to enhance their health or safety. Id. at 880-81.
However, there is no basic agreement among economists as to what elements ought to go into the life valuation. There is no unanimity on which studies ought to be considered. There is a lack of reliability. In fact, Smith was prepared to testify based on seventy or eighty studies; Miller relies on twenty-nine; in Sherrod v. Berry, 629 F. Supp. 159, 163 (N.D.Ill. 1985), Smith testified on the basis of fifteen studies. Smith acknowledged that more studies could be done on the willingness-to-pay issue. In particular Smith noted that further studies will focus on a set of consumers to uncover when these consumers make or do not make choices for safety, and these results may help establish validity. The fact that the bottom lines of most studies (between less than $100,000 to more than $12,000,000) arguably do not wind up very far apart (by some definitions of "very far") may be coincidence and not the result of the application of a scientific method.
T. Miller, "Willingness to Pay", at 883.
Survey of attitudes and views of others as a basis for concluding something is true is not necessarily wrong. Some science as it comes into court is the result of consensus by practitioners of some area of expertise that a certain law of nature is correct. What is wrong here is not that the evidence is founded on consensus or agreement, it is that the consensus is that of persons who are no more expert than are the jurors on the value of the lost pleasure of life. Even if reliable and valid, the evidence may fail to "assist the trier of fact to understand the evidence or determine a fact in issue" in a way more meaningful than would occur if the jury asked a group of wise courtroom bystanders for their opinions.
As I noted in open court, I do not regard this proferred testimony and this theory as either pro-defense or pro-plaintiff. Nor do I question the possible utility of this evidence for a legislature wishing to adopt, by statute, a table of damages.
For the reasons stated here and in open court, I grant the defendants' motion to bar testimony of plaintiff's expert, Stan V. Smith, on the issue of hedonic damages.