Concept
Testing for Habit
Habit and routine practice
Testing for Habit
Rule 406 admits evidence of a person's habit or an organization's routine practice to prove conforming conduct on a particular occasion, but its text gives no yardstick for telling when a pattern of behavior has become a habit.1 The definition does the first sorting: a regular response to a repeated specific situation is habit, and a general disposition is character. Applying that definition to a witness's actual testimony is harder, and courts do it by examining three features of the pattern offered. Specificity asks how narrowly the triggering situation and the response are drawn. Regularity asks how often the response followed the situation and how many instances support the claim. The third feature asks how deliberate the conduct is, because conduct chosen anew on each occasion looks like character, while conduct that has become semi-automatic looks like habit.2
All three features measure how closely the pattern approaches a reflex set off by a recurring trigger. A pattern that scores high on each can be laid beside the disputed event and support an inference about that event without any detour through the person's general character.3
The three features are guides, and no formula combines them. Precise standards for how many instances or how much uniformity will suffice cannot be formulated.4 The trial judge therefore has considerable discretion in deciding whether conduct is narrow enough to be habit and whether the evidence of repetition is enough to establish it.5 The subsections below take the features one at a time, beginning with the one that fixes what the court is testing.
Specificity
Specificity is the first question a court asks of any offered pattern, because regularity and automaticity can be judged only after the pattern has been stated precisely enough to test. The factor has two dimensions that rise and fall together. One is the situation, meaning how narrowly the triggering occasion is drawn. The other is the response, meaning how particular the conduct is that follows. The Advisory Committee's examples of habit each pair one kind of situation with one type of conduct: going down a particular stairway two stairs at a time, giving the hand signal for a left turn, alighting from railway cars while they are moving.6 A court sorts a proffer by asking whether the pattern, as described, resembles those examples or resembles a disposition spread across a person's whole life.
Narrowness matters because habit evidence earns its weight through the match between the pattern and the disputed occasion. When the situation is narrow, the jury can lay the pattern beside the facts and ask whether the day in question was one the pattern covers. When the response is particular, a witness can describe it concretely, and little room remains for ambiguity about what the person did. Always locking the car door on leaving the car is a pattern of that kind. A trait such as calmness or honesty is far more ambiguous, and even when it is proven, its connection to conduct on any single occasion is weak.7
Testimony that Marcus is a careful driver describes the same man at the same wheel but names no situation and no act. The trait covers every intersection, every lane change, and every parking space, so the jury could use it only by reasoning that a careful person probably crossed carefully that day. That is the character inference the propensity prohibition forbids. The same underlying fact can be offered either way: proof that a driver always stops at stop signs and looks both ways would usually be excluded if offered to show the driver is cautious and law-abiding, yet it is relevant to how the driver acted on one occasion when it is offered as habit.8
Much evidence is excluded simply because it fails to reach the status of habit. Evidence of intemperate “habits,” for example, is generally excluded when offered to prove drunkenness in accident cases, and evidence of other assaults is inadmissible to prove the one at issue in a civil assault action.9 The label the proponent or the witness attaches therefore does not matter, because the court measures the breadth of the conduct actually described. Suppose the prosecutor offered specific examples of Stan's lies instead. Specific instances are a recognized way to prove a habit, and a later section takes them up, but a method of proof cannot narrow a description that was broad when offered. Instances of lying on unrelated subjects to unrelated listeners would still support only the trait.
A proponent whose first offer is overbroad can often repair it by redefining the situation and the response narrowly, provided a witness actually observed the narrower pattern.
Redescribing the evidence does not manufacture facts. The narrower account is admissible only because a witness saw it, and it commits the proponent to a precise claim that the opponent can test on cross-examination. Even a narrowly drawn pattern must still clear the other factors, since a practice observed twice, or one that depends on a fresh choice each time, may not amount to habit.
Regularity and Sampling
The regularity factor asks whether a pattern has recurred often enough and consistently enough to deserve the name habit. Habit and routine-practice evidence draws its probative force from extreme regularity, so the court asks whether the proponent has shown one regular response to a repeated specific situation, and it weighs the adequacy of the sample and the uniformity of the response.11 The Advisory Committee's Note names the same two considerations as key factors and adds that precise standards for measuring their sufficiency cannot be formulated.12 A pattern that falls short leaves the factfinder with only a general tendency to act in a certain way, and that is propensity proof.
Courts give regularity a measurable form. In Wilson v. Volkswagen of America, Inc., the court held that examples offered to establish habit must be “numerous enough to base an inference of systematic conduct” and must establish “one's regular response to a repeated specific situation.” It named “adequacy of sampling and uniformity of response” as the key criteria, or, in the words of an article the Advisory Committee's note cites with approval, the “ratio of reactions to situations.”13 Put practically, the court counts the occasions on which the person responded in the claimed way and sets that count against the occasions on which the same situation arose. The ratio matters because of the inference the jury is asked to draw. The disputed occasion is one more instance of the situation, and the jury reasons that it probably went the way the earlier instances went. When the response followed the situation nearly every time, that reasoning is strong. When it followed only a little more than half the time, the pattern shows only that either outcome was possible.
The ratio can be trusted only if the sample behind it is adequate and the response is uniform, and those are separate demands. The first demand is for enough instances: a handful of observations cannot distinguish a standing practice from coincidence, so the examples must be numerous enough to support an inference of systematic conduct.14 The second is for enough uniformity that the pattern is rarely broken. Foundational evidence is weakened by variable conduct, and also by observations that span only a short period or a small number of instances, although a deviation now and then does not defeat a habit.15 A large sample with many exceptions fails on uniformity, and a perfectly uniform sample of two instances fails on adequacy.
The Fifth Circuit's decision in Reyes v. Missouri Pacific Railway Co. applies these considerations to a real record. The court reversed the district court and held that four prior convictions for intoxication were inadmissible under Rule 404. They were proffered to prove that Reyes acted in conformity with his character on the night of the accident, and they failed to constitute habit evidence. The court explained that although no precise formula can be proposed for when behavior becomes consistent enough to rise to habit, “adequacy of sampling and uniformity of response” are the controlling considerations.16 Four convictions say nothing about how Reyes responded to any defined situation, so the jury could infer only that he was the kind of person who drinks too much. That is the inference the Advisory Committee's note identifies when it observes that evidence of intemperate “habits” is generally excluded as proof of drunkenness in accident cases.17
A raw count does not decide the question, and two other decisions show why. The district court in Loughan v. Firestone Tire & Rubber Co. admitted testimony about Loughan's drinking as habit and distinguished Reyes on the ground that Reyes rested on four particular incidents. Firestone, by contrast, offered evidence of drinking over an extended period together with evidence of Loughan's regular practice of carrying a cooler of beer at or about the time of the accident, and the court relied on the cumulative effect of that evidence in measuring its sufficiency.18 The Advisory Committee's note, for its part, treats the rule as consistent with Whittemore v. Lockheed Aircraft Corp., which upheld admission of evidence that the plaintiff's intestate had on four other occasions flown planes from the defendant's factory for delivery to his employer airline. That evidence was offered to prove that he was piloting, rather than riding as a guest, when a plane crashed en route for delivery.19 Four instances supported the inference in Whittemore and fell short in Reyes because a sample is measured against the situation it must represent. Four deliveries from one factory to one employer repeat a single narrow situation, while four convictions for intoxication sample an open-ended one, any occasion on which the person might drink.
At the bottom of the scale, a single prior act cannot show habit, because one instance supplies no ratio and no way to tell a standing practice from a one-time event.
The 95% and 55% figures in this section mark the poles of the scale and do not state thresholds. Because no precise standard of sufficiency can be written, the trial judge has considerable discretion in deciding whether the foundational evidence of repetitious activity establishes a habit.20 A strong sample also cannot rescue a description drawn too broadly, because regularity is measured against a situation that the specificity factor has already required the proponent to state narrowly.
Semi-Automatic Response
The third factor asks how much conscious deliberation a pattern involves, and it takes its name from the Advisory Committee's description of habit. That description gives the hand signal for a left turn, the particular stairway taken two steps at a time, and alighting from moving railway cars as examples, and it observes that the doing of habitual acts may become semi-automatic.22 Each example is something a person does when a recurring trigger appears, without stopping to decide what to do. Conduct at the other end of the range is chosen anew each time, and it looks more like a preference or a trait.
The factor matters because of variation. A response that follows its trigger without a fresh decision gives the actor no moment at which to weigh alternatives, so it has little room to differ from one occasion to the next. A response that depends on a decision each time is exposed to mood, convenience, and competing demands, and any of those can change the outcome on the one day that matters in the lawsuit. The prospect of change weakens the inference that the disputed occasion matched the earlier ones.
The contrast is sharpest in Levin v. United States. The accused offered testimony about his religious practices to show that he was at home observing the Sabbath, and so was not out obtaining money through larceny by trick. The D.C. Circuit held the testimony properly excluded: “It seems apparent to us that an individual's religious practices would not be the type of activities which would lend themselves to the characterization of ‘invariable regularity.’ Certainly the very volitional basis of the activity raises serious questions as to its invariable nature, and hence its probative value.”23 A person who observes the Sabbath chooses to do so week after week, and because the choice expresses belief, the pattern cannot be treated as a response that simply follows its trigger. The driver who signals every left turn stands differently. The intention to turn prompts the signal, the signal takes little thought, and a witness can describe the act precisely. Both patterns may be regular in the everyday sense, but only the signal is performed without deliberation, and that difference is why a court is more likely to receive it.
The factor is a guide, not a gate. A rule excluding every pattern that involves some thought would shut out proof that is regular and specific, and that proof would lose none of its power to show what happened on the day in question.
That outcome fits one court's statement of the elements of habit as regularity, specificity, and an involuntary or semiautomatic response. The same opinion adds that the last element does not require a reflexive or nonvolitional response, only a uniform one.24 On that reading, the semi-automatic quality is best understood as evidence of uniformity. Volition remains relevant because it is a reason to doubt uniformity, as Levin shows for an activity people choose to vary, but a specific practice that is in fact followed without exception does not lose its force because the actor thinks while performing it.
The Advisory Committee's description blends two theories of habit, and the difference between them explains why this factor is treated flexibly. Under the psychological theory, the response must be semi-automatic, which means it uniformly follows a stimulus, though uniformity is only one aspect of that theory. Under the probability theory, the question is whether the behavior is specific and sufficiently regular. Regularity is a requirement of both theories, because a response that does not uniformly follow its stimulus cannot be semi-automatic.25
The drafters' primary reason for advancing two theories was to promote trial court discretion. A judge presented with proof that a litigant regularly reads mystery novels before bed, offered to show the litigant was reading on the night in question, could exclude it under the psychological theory because reading is volitional, or admit it under the probability theory because the behavior is specific and sufficiently regular.26 A proponent whose pattern involves choice will therefore press the probability theory and point to uniformity, while an opponent will press the psychological theory and point to the volitional character of the act. An automatic act observed only twice still fails the regularity factor, and a deliberate practice observed hundreds of times without exception can satisfy it.