The hedonic impact of "stand-alone" emotional harms - an analysis of survey data.
This paper employs survey data on subjective well-being and a battery of self-assessed health measures Io estimate the hedonic impact of emotional health, as decoupled from its physical counterpart. The disaggregation of global health into physical and emotional components is done with a parochial eye toward tort law, which has historically drawn a distinction between physical and emotional harms, limiting recovery on the latter--particularly "stand-alone" emotional harms--through various common law doctrines. The results of three sets of regression analyses suggest that a range of potentially inactionable emotional conditions, including emotional conditions with no concomitant physical manifestations, exert a significant negative impact on subjective well-being. Further, the emotional health variables included in the models uniformly bore stronger connections to well-being than their similarly-worded and similarly-scaled physical health analogues. In discussing the results, the paper offers a cautious defense of the use of subjective well-being measures--so-called "happiness" data--as a tool for informing the concepts of harm and injury in tort.
Empirical research on subjective well-being ("SWB"), or "happiness research," as it is commonly called, (1) is not a particularly new field of study. (2) However, a recent resurgence of interest in SWB has drawn a fresh batch of commentators and practitioners into its wake--empirically minded legal scholars. Early adopters in the legal academy have offered insights on a range of issues including research on happiness to civil damage awards, (3) the impact of crime on our lives, (4) and the implications of a hedonic version of cost-benefit analysis. (5) Given the rising interest in the data and methodology underlying such studies, the increasing sophistication in which the studies are executed, and the cultural resonance of happiness as an element of human flourishing--as evidenced by the ongoing treatment of the subject in the mainstream media (6)--current scholarly offerings on the subject appear to be the mere tip of the iceberg.
This paper employs survey data on SWB and a battery of self-assessed health measures to estimate the hedonic impact of emotional health as decoupled from its physical counterpart. The disaggregation of global health into physical and emotional components is done with a parochial eye toward tort law, which has historically drawn a distinction between physical and emotional harms, limiting recovery on the latter through various common law doctrines. While the law of torts has made significant strides toward removing the distinction between physical and emotional harms, recovery for emotional harms is still limited in scale and scope, and tied, in various ways, to physical harm. (7) The emotional and physical distinction is often defended on practical grounds, by reference to the nebulous nature of mental injuries, the inability of courts to effectively distinguish genuine emotional harms from fraudulent ones, and the difficulty in valuing such injuries. (8) However, underlying such arguments is an enduring suspicion--often, but not always, implicit--about the unimportance of emotional tranquility to our lives. (9) Thus, the distinction remains despite challenges from the other side of the debate, who view the distinction between emotional and physical injuries as anachronistic, (10) unfair, (11) and discriminatory. (12)
Section I of this article outlines the rules of recovery for mental and emotional harms in tort actions by comparing them to the rules of recovery for physical injuries. The terms "mental" and "emotional" are used interchangeably to refer to a broad category of mental health conditions that are commonly, even if incorrectly, understood to be distinct from physical conditions. This negative view of "mental" and "emotional" health--which covers anxiety, inability to concentrate, depression, anguish, grief, psychosis, humiliation, fright, shock and other negative emotions distinct from physical pain (13)--is a legal concept, not a medical one. (14)
Section II provides background on the burgeoning empirical literature on SWB, including an overview of findings on SWB's connection to health as well as a discussion of the applications of empirical research on SWB in the legal field. Also included in Section II is a discussion of the usefulness of SWB as a proxy for harm in the tort context. Without endorsing the view that recovery in tort should be completely determined by hedonic considerations, this article argues that SWB--with its broad conceptual reach and its mostly intuitive correlations with several important quality-of-life measures--captures something meaningful about what makes us "whole," both as aggrieved litigants and as individuals.
Section III outlines the data and methods employed in the core analyses of the paper. The study comprises three sets of regression analyses, each of which capture emotional health in different ways in an effort to gauge its impact on SWB. Each of the three regression sets contains two distinct models that decouple the hedonic impact of physical and emotional health. In the first model, physical and emotional health are both included among the variables then compared with another variation performed on survey subpopulations that are free of physical health issues. The second model is an import to the law of torts, which approximates the hedonic impact of so-called "stand-alone" or "non-parasitic" emotional harms, which are emotional harms that are not occasioned by or accompanied with physical symptoms. Such claims are treated with particular suspicion in the tort context. (15)
Section IV presents three primary results: (1) emotional harms exert a significant impact on SWB; (2) the impact of emotional health on SWB appears to be largely independent of physical health status; and (3) the emotional health variables uniformly bear stronger connections to SWB than their similarly-worded and similarly-scaled counterparts regarding physical health. To convey a rough sense of the impact of emotional health on SWB, this article includes rough monetary equivalents for various degrees of change in emotional health.
Section V examines the various limitations of the analysis, explaining how such limitations might color the interpretation of the findings and impact their utility vis-a-vis tort law. A question of fundamental importance is addressed at the outset of the section: Are emotional harms and SWB so conceptually similar as to make the analysis question begging? The article argues that, while the concepts of emotional health and SWB are related and, at times, conflated, they are sufficiently distinct to make the analysis meaningful. The issue of external validity is also raised in Section V, since this analysis seeks to bring emotional health data from outside the tort context to bear upon the judicial treatment of emotional injuries within tort law. Finally, the broad framing of the emotional health survey questions, which reveals little about the specific clinical nature of the emotional problems that underlie survey responses, is addressed.
Ultimately, the current inquiry is neither intended to provide a definitive conclusion about the judicial treatment of emotional distress, nor is the analysis aimed at generating precise monetary estimates of the impact of specific types of emotional harm; rather, the current inquiry is an effort to test whether a range of potentially non-actionable emotional harms, including stand-alone harms, impact well-being in a significant way.
I. LEGAL BACKGROUND
The distinction between physical and emotional harm, and the relative devaluation of the latter, has deep historical roots in tort law. (16) In the early twentieth century, there was a categorical denial of recovery for standalone emotional injuries, (17) but, as time passed, courts began to allow recovery for mental and emotional injuries arising from, or concomitant with, physical injuries. (18) The restraint of "parasitic" mental harms, which are those harms tied to physical injuries, was seen as a practical tool for limiting claims:
The fear of imaginary injuries and fictitious suits, the belief in self-responsibility for mental well-being, the difficulty of monetarily valuing emotional harms, the lack of tools and standards for measurement of emotional ills, and the nascent state of the behavioral sciences all combined to preclude recovery for emotional suffering. Yet even at this juncture in history, an unarticulated basis for rejecting claims of emotional distress was probably the relative devaluation of emotional injuries compared to physical injuries. (19)
By the mid-twentieth century, the requirement of a physical injury was relaxed for intentionally inflicted emotional distress ("IIED"), only to be replaceed with a series of different limitations on recovery. Courts initially required, as a precondition for IIED recovery, either a physical manifestation of the emotional disturbance or a "physical impact" associated with the tortious behavior. (20) The physicial impact and physical manifestation requirements then gave way to a different set of limiting factors--a requirement of "extreme and outrageous" conduct on the part of the defendant and a plantiff-side threshold requirement allowing recovery only for "severe" emotional disturbance. (21)
Both requirements were meant to limit emotional distress claims. Liability for extreme and outrageous conduct is "found only where the conduct has been so outrageous in character, and so extreme in degree, as to go beyond all possible bounds of decency, and to be regarded as atrocious, and utterly intolerable in a civilized community," such that "the recitation of the facts to an average member of the community would arouse his resentment against the actor, and lead him to exclaim, Outrageous!" (22) Extreme and outrageous conduct is hard to show. For example, harassment and improper termination by an employer, (23) issuance of a false positive HIV test during a third trimester of pregnancy, (24) disturbance of a relative's grave by exposing the deceased, (25) and failure to clearly warn an employee of "substantial amounts of asbestos" in a confined work site (26) have all been deemed insufficiently outrageous to create liability. The exacting standards for outrageousness is indicated by prior dispositions of IIED claims in New York, where "every [IIED claim considered by this court] has failed because the alleged conduct was not sufficiently outrageous." (27) In addition to the difficult hurdle of proving outrageousness, the emotional injury must also be "so severe that no reasonable person could be expected to endure it." (28) Emotional harms entailing "nightmares, difficulty sleeping, extreme loss of self-esteem and depression, requiring additional psychological treatment and counseling" (29) have failed to rise to the requisite level of severity.
The twin requirements of "extreme and outrageous" conduct and "severe" emotional distress are interpreted so as to effectively nullify what is commonly known as the "eggshell skull" or "thin skull" rule in the context of emotional harm. (30) That is, while the defendant takes the plaintiff as she was found in the case of physical harms, an individual predisposed to emotional harm "may not recover if an ordinary person would not have suffered serious emotional disturbance" from the negligent activity at issue. (31) The requirement of extreme and outrageous conduct and the severe distress threshold remain good law in nearly every state to this day. (32)
The trajectory of the law with respect to negligently inflicted emotional distress (NIED) is similar to, if somewhat slower than, that of IIED. Early twentieth century cases denied recovery for non-parasitic emotional harms arising from negligent behavior. (33) As with IIED, the physical injury rule led to a marginally more permissive rule that allowed recovery for emotional distress where such distress was occasioned by a physical impact or manifested itself physically. (34) A majority of states currently require either a physical impact or manifestation as a precondition for non-bystander NIED claims, (35) in order to "remove an emotional distress claim from the realm of speculation" (36) or, equivalently, "[t]o ensure that the emotional injury is sufficiently serious to be afforded legal protection." (37)
In certain instances, the physical impact requirement has been abandoned altogether. Individuals suffering from serious emotional harm can sometimes recover absent an actual physical impact if the behavior of the tortfeasor placed them at immediate risk of physical harm and a close relative is physically injured by the defendant's negligent behavior. (38) A small number of states follow this so-called "zone of danger" rule or a close variant (including, in some cases, additional requirements such as physical manifestation of emotional distress). (39) A growing majority of states now follow an even more permissive rule with respect to NIED--the Dillon rule, which was articulated by the Supreme Court of California in Dillon v. Legg. (40) Dillon substitutes the relatively rigid zone of danger requirements with a flexible set of considerations, including the nature of the relationship between the emotionally disturbed plaintiff and the physically injured party, the physical proximity of the plaintiff to the underlying accident, and the nexus between the witnessing of the accident and the emotional disturbance at issue. (41) At the time of this writing, more than half of the states follow a rule based on Dillon. (42) In sum, the law has greatly expanded the range of circumstances under which individuals can recover for emotional injuries.
Notwithstanding such strides, however, emotional injuries remain "in large part, tied to either physical impacts, physical manifestations of injury, or other proxies for emotional distress." (43) Moreover, suspicion about the importance of mental helaht to our lives--as opposed to the various practical arguments against greater recovery for mental harms--appears not to have faded. The following characterization of stand-alone mental harms in the Second Restatement is still cited by courts today (44) despite the presence of an updated version:
emotional distress which is not so severe and serious as to have physical consequences is normally in the realm of the trivial, and so falls within the maxim that the law does not concern itself with trifles. It is likely to be so temporary, so evanescent, and so relatively harmless and unimportant, that the task of compensating for it would unduly burden the courts and the defendants. (45)
The Third Restatement of Torts is far less dismissive of emotional harms than its predecessor, yet still favors physical health over emotional tranquility. (46) Recent cases cast a similarly jaundiced eye on stand-alone emotional injuries, which are labeled as "everyday," (47) "usually trivial," (48) "often transient," (49) and "a part of the price of living among people." (50) The Supreme Court of Hawaii echoes such sentiments, cautioning against more expansive definitions of emotional distress, which it warns might "curry.... neurotic patterns in the population." (51) Such sentiments also find support in the legal academy. (52)
The blanket marginalization of mental harms vis-a-vis their physical counterparts is especially stark given the lax requirements concerning the magnitude of physical injury, or "impact," often required to authenticate emotional trauma--"[physical] [c]ontact, however slight, trifling, or trivial" (53) and "bodily injuries, even though trivial or minor in character" (54) are considered sufficient to support a cause of action for emotional distress. In the eyes of the law, therefore, the smallest quantum of physical injury is inherently more real and reliable than any number of emotional harms.
The devaluation of emotional harms extends beyond the courtroom--damages for physical injuries are excludable from taxable income while those arising from emotional distress are not. (55) The differential treatment of physical and emotional harms in the Internal Revenue Code implies that awards for emotional distress are windfall gains rather than restorative transfers. In the words of one comentator, the distinction made by I.R.C. [section] 104(a) "suggests a fundamental distrust on the part of Congress in the reality of emotional distress." (56)
Despite a number of challenges to the distinction between emotional and physical injuries, the relative devaluation of the former endures within tort. The argument that emotional harms are singularly difficult to estimate appears somewhat disingenuous given the wide variation in awards for similar types of physical injuries. (57) The concern over false claims, and the related claim that emotional harms are endogenous to legal rules on recovery, (58) underestimates the effectiveness both of forensic psychologists in identifying malingering injuries asserted by mental health claimants (59) and the complementary role of jurors in the detection of same. Moreover, such arguments elide the fact that physical injuries, too, are susceptible to the problem of malingering. (60) The notion that mental harms are somehow trivial is at odds with wide swaths of research that demonstrate the impact of poor mental health on everyday life (61) and the role of mental states in the etiology of various types of illnesses, (62) many of which manifest long after the initial emotional harm. Finally, a number of commentators have argued persuasively that the various limitations on recovery for emotional harms disproportionately impact women. (63)
II. THE EMPIRICAL STUDY OF WELL-BEING--A BRIEF PRIMER
SWB research generally proceeds through the gathering of survey data on various aspects of people's lives and examining the statistical interplay between self-assessed well-being and its putative correlates. By asking people to assess and report their own levels of well-being, the studies replace the standard "objective" economic measure of well-being, like consumption behavior or revealed preferences, with subjective measures. The purview of SWB research extends beyond the realm of basic economic indicia, exploring the impact of demographic, attitudinal, social and political categories on well-being. Prior work, for example, has sought to identify the relationship between SWB and such diverse phenomena as household income, (64) race, (65) gender, (66) marital status, (67) television viewing, (68) aspects of governance, (69) bereavement, (70) crime, (71) religious participation, (72) educational attainment, (73) unemployment, (74) sexual activity, (75) and even internet access. (76)
Well-being is commonly elicited through simple, single-item survey questions. (77) Other methods of acquiring data on well-being include the "experience sampling method" (ESM) and "day reconstruction method" (DRM), both of which rely on repeated reports of well-being over time. (78) Using pagers to signal to, or beep, study participants at various times throughout the study interval, ESM research allows for real-time assessments of well-being, recorded soon after, or contemporaneously with, various life events. (79) DRM, as the full name suggests, calls for daily reflections on well-being. (80) Moment-by-moment measures of SWB, it should be noted, bear strong correlations with single-item measures of SWB. (81)
Measures of SWB exhibit a number of favorable psychometric properties. The reliability, or stability over time, of SWB measures has been established by test-retest studies, (82) and alternative framings of SWB tend to converge with each other. (83) Moreover, the validity of SWB constructs is supported by their alignment with third-party assessments of well-being (84) and physical markers of well-being such as heart rate, blood pressure, (85) and "Duchenne" smiles. (86) Measures of SWB also respond in expected ways with important life events--positive income increases and marriage, for instance, have been shown to correspond with higher reported levels of well-being. (87)
A. Limitations of SWB Research
Notwithstanding their favorable qualities, survey measures of well-being--particularly single item measures of the type used in the current analysis--are susceptible to various types of bias. Such survey measures can be influenced by seemingly trivial events that precede the administration of surveys, are sensitive to the ordering of questions, and are culturally mediated in a variety of ways, rendering global comparison problematic. (88) Additionally, single-item measures may be subject to social desirability effects and vulnerable to "peak-end" valuation, or a tendency to recall the last and most intense aspects of an event, rather than the entire experience. (89) More fundamentally, SWB measures force respondents to reduce complex and varied feelings about life into a single number. (90) These limitations have led many to question the validity of single-item SWB measures, particularly when they are offered as straightforward measures of happiness or as a perfect proxy for economic conceptions of utility or human welfare. (91)
The limitations of the single-item SWB measures are not, however, fatal. Split-sample surveys aid in the identification of ordering effects. Further, "[t]he idiosyncratic effects of recent, irrelevant events are likely to average out in representative population samples," (92) thus alleviating the concern about their distortionary impact on global life evaluations. Moreover, retrospective evaluations of SWB have been shown to predict subsequent behavior in a number of domains, providing further support for their validity. (93) More importantly for this analysis, the cognitive and evaluative components of single-item SWB responses--a feature that is part and parcel "of judgment and of memory" effects described above--may actually be a boom in cases where an enduring sense of well-being, rather than an unfiltered measure of "hedonic flow," is required. (94) The relationship between single-item SWB measures and the operative concept of harm in the tort context is discussed further in Section II.C.
B. Health and Subjective Well-Being
The relationship between SWB and self-assessed health is well established in the literature. Often touted as the "strongest predictor of subjective well-being" during adulthood, (95) perceived health is a function of both objective health status--physician-rated health, (96) mortality, (97) functional decline, (98) and health care utilization (99) all being robust correlates--and affective orientation. Self-assessed health, in other words, is partially determined by actual health and partially determined by individual tendencies to see things in either a positive or negative light. (100) The two components of self-assessed health correspond to two different frameworks for analyzing SWB--so called "bottom-up" theories, in which life events and circumstances combine to determine SWB at the level of experience, and "top-down" theories, which emphasize the role of genetics and personality in the determination of SWB. (101)
Much of the extant research on the connection between health and SWB, however, employs measures that subsume both emotional and physical health. (102) Among the studies that do account for emotional health, none have sought to specifically analyze subpopulations suffering from no physical harm. (103) In short, there appears to be no analysis of the hedonic impact of "stand-alone" emotional harms. The utility of studies using such "global" health measures is thus limited in two ways. First, to the extent that self-assessed health, alone, is used as a proxy for physical health, the model is likely to overestimate the relationship between health and well-being. (104) Secondly, without a corresponding marker of emotional health, the results are useful only to policy domains in which the distinction between physical and emotional health is of no great importance. Where the distinction between physical and emotional health does matter, as it does in the law of torts, separate measures of physical and emotional health are required.
C. Subjective Well-Being and Civil Damages
Research on SWB offers a unique opportunity to circumvent conventional economic analyses of damages and supplement the ad hoc valuation methods currently used in courts. (105) Accordingly, a number of recent papers bring empirical results from the SWB literature to bear on the issue of non-economic civil damages. Whereas economic analyses seek to construct a hypothetical demand curve around risk avoidance by examining consumption behavior, (106) the hedonic method proceeds by examining the diminution in well-being that typically accompanies various categories of injuries or life circumstances. The underlying question thus becomes: to what extent are diminutions in individual well-being the appropriate metric to identify injuries and estimate non-economic damage awards? In other words, does making an injured party whole mean restoring the injured party to some pre-injury hedonic level or hedonic status quo ante? (107)
There are a several reasons to think that SWB is a useful tool for informing the patchwork of rules used to determine damages in tort, even if one reasonably concludes that hedonic responses to injuries are not the sole criterion for identifying and quantifying damage awards. Two key features of SWB that stand out are its simplicity and its broad conceptual reach. SWB, as measured by single-item survey measures, is a global evaluation of one's life--an aggregation of positive affect, negative affect and goal-fulfillment, filtered and weighted through a subjective lens that reflects one's own values. Indeed, it is difficult to imagine a single measure that better captures human functioning and flourishing, which may be why the concept has enjoyed such resonance through time. Indeed, SWB still resonates today:
[P]eople the world over think SWB is very important. In a survey of college students from 17 countries, Diener found that happiness and life satisfaction were both rated well above neutral on importance (and more important than money) in every country, although there was also variation among cultures. Furthermore, respondents from all samples indicated that they thought about happiness from time to time. Thus, even those from relatively unhappy societies value happiness to some extent. (108)
Moreover, SWB also moves in expected ways with many items on so-called "objective-lists," or philosophical inventories of fundamental human freedoms and entitlements without which "no society can lay claim to basic justice." (109) One such list, offered by prominent philosopher Martha Nussbaum as an alternative to SWB survey measures, includes health, access to education, political and material autonomy, and the ability to enjoy recreation and social activities. (110) That all of the above items are positively correlated with SWB is evidence of its usefulness, if not as a conclusive measure of human functioning, then at least as an efficient tool to capture quality of life in certain judicial contexts, where injuries must often be reduced to a single dollar figure.
Recognizing the potential of SWB as a meaningful, if imperfect, proxy for the quality of life, legal commentators have begun to apply SWB research in the tort context. Sunstein's recent piece on the issue is illustrative. Noting the tendency of individuals to overestimate the adverse impact of many physical problems and the fact that "[t]hose who face such problems experience unexpectedly little in the way of hedonic loss," (111) Sunstein suggests that courts should draw a line between "harms that impose enduring losses, such as chronic pain and mental illness, and harms that do not, such as losses of fingers and toes." (112) While Professor Sunstein is clear that not all injuries are to be defined hedonically--he carves out an exception for loss of capability--his argument ultimately rests on the importance of well-being to our conception of injuries. The enduring hedonic losses associated with unemployment (113)--losses above and beyond those associated with lost wages--have also been invoked by proponents of broader recovery rules in the employment discrimination context. (114)
Other commentators are less sanguine about the use of hedonic measures to identify injuries in tort, particularly where such measures are posited as a replacement for, rather than a complement to, existing notions of harm:
[P]eople care about many things that are not purely hedonic, such as meaning, capabilities, and range of feeling and experience. If this is the case, it would be seriously misguided to ignore the expressed distaste of all three groups for the health conditions in question and to base valuations of noneconomic damages on a notion of well-being that is far narrower than that adopted by individuals themselves. (115)
Similarly, Swedloff and Huang push back against the growing revolution of SWB research in legal academe, echoing the above concerns and noting that many of the findings in the literature, such as adaptation, are not as robust and well-understood as they seem. (116) Further, challenging Sunstein's conclusions, Swedloff and Huang argue that juries might already be compensating for the fading psychological impact of injuries. (117)
Ultimately, much of the resistance to the incorporation of SWB research into tort law appears to stem from a fear that hedonic responses, if accepted as a workable proxy for harm, will supplant other considerations entirely and, perhaps, be used to limit certain types of recovery. Insofar as the instant analysis seeks neither to define damages in strictly hedonic terms nor to limit damages of a certain type, the above concerns may be allayed. Nonetheless, the import of the findings presented below depends on whether hedonic states ultimately matter in the context of tort damages.
III. DATA AND METHODS
The data used in the following analyses are obtained from the General Social Survey (GSS) series, which is representative of the non-institutionalized, English-speaking U.S. population over the age of 18. (118)
The repeated cross-sections of the GSS contain data on a broad range of demographic and attitudinal dimensions, including well-being, health, income, education, family composition, health and political ideology. (119) To date, the GSS survey has been administered to more than 55,000 respondents--approximately 1,500 each year between 1972 and 1993 (with some gaps) and approximately 3,000 every other year from 1994 to the present. (120)
However, only a fraction of the GSS data are usable in the present analyses. While the survey instrument for subjective well-being--the dependent variable in all three regression models--appears in all waves of the GSS, the self-assessed health variables used here are confined to the year 2000 wave and the year 2006 wave. (121) After omitting observations without all of the variables of interest, the number of observations in the regressions below range from 910 to 2,236.
A. Key Variables
Subjective well-being, the dependent variable in all of the analyses, is measured on a one to three scale based on the following GSS survey instrument: "Taken all together, how would you say things are these days would you say that you are very happy, pretty happy, or not too happy?" (122) The independent variables used to capture mental health--"MNTLHLTH," "DIDLESSE," and "CRELESSE"--pose the following questions, respectively, to GSS survey respondents:
MNTLHLTH: "Now thinking about your mental health, which includes stress, depression, and problems with emotions, for how many days during the past 30 days was your mental health not good.?" (123)
DIDLESSE: "During the past 4 weeks, have you had any of the following problems with your work or other daily activities as a result of any emotional problems (such as feeling depressed or anxious)? ... Accomplished less than you would like?" (124)
CRELESSE: "During the past 4 weeks, have you had any of the following problems with your work or other daily activities as a result of any emotional problems (such as feeling depressed or anxious)?.... Didn't do work or other activities as carefully as usual?" (125)
The analogous physical health survey instruments--used in some, but not all, of the regression models--pose the following questions to respondents:
PHYSHLTH: "Now thinking about your physical health, which includes physical illness and injury, for how many days during the past 30 days was your physical health not good?" (126)
DIDLESSP: "During the past four weeks, have you had any of the following problems with your work or other regular daily activities as a result of your physical health? Accomplished less than you would like?" (127)
LIMITEDP: "During the past four weeks, have you had any of the following problems with your work or other regular daily activities as a result of your physical health? Were limited in the kind of work or other activities?" (128)
Each of the three mental health variables appears in two different ordinary least squares (129) regressions: one of which contains the analogous physical health variables and one which does not contain the analogous physical health variables, but is performed on the survey subpopulation that reports no physical problems. The control variables, all of which bear consistent and significant connections with SWB in prior work, include marital status, age, race, and gender. (130)
Income, as one of the most consistent and strong correlates of SWB, (131) is also included in the analysis. The inclusion of log household income is particularly important in light of its likely correlation with both reported physical and mental health. As previously mentioned, though the physical and mental health variables used in the current analysis are intended to shed light on the impact of various injuries that might arise in tort cases, the actual health limitations reflected in the data are not necessarily the result of tortious behavior. In fact, the data reveals little about the etiology or subsequent treatment of the conditions. Insofar as individuals with higher incomes are better able to treat or mitigate the effects of health conditions, the omission of income would bias the coefficients on the mental health variables. Moreover, vis-a-vis low-income individuals or families, those of greater economic means are generally exposed to less risk and are better educated about health conditions. Thus, given the goal of this inquiry--to measure the impact of emotional harms on well-being, given the fact that they occur and given the fact that some individuals may already be treating them--income should be included among the variables. The inclusion of income also allows for the calculation of monetary equivalents for changes in emotional health. For all of the regressions, the survey (SVY) settings in STATA are used to adjust for the complex sampling design of the GSS.
Tables 1-3 present the results of the regressions, and Table 4 translates the hedonic impact of the various health conditions into rough monetary equivalents. (132) The control variables generally bear the expected relationships with SWB. Other things equal, married respondents report higher SWB than their unmarried counterparts, whether they are divorced, separated or never married. Black respondents report lower SWB than white respondents. As in other studies, female respondents report higher SWB than males, other things equal, though the female dummies failed to reach significance in any of the regressions. Taken together, age and its square indicate a U-shaped (curvilinear) relationship with SWB, though the age variables reached significance only in the first set of regressions. (133)
As expected, the log household income variables are significant in all of the regressions, exhibiting a positive, if modest, relationship with SWB. The modest size of the income coefficients, it should be noted, leads to extremely large monetary equivalents for the health conditions. Reassuringly, this finding is in line with other studies using SWB and life-satisfaction measures to value intangible goods. (134) It should be noted, further, that the figures provided in Table 4 do not account for hedonic adaptation, which may limit the amount of time in which well-being is negatively impacted by the relevant health condition. (135)
Table 1 presents the results of the regressions that capture emotional health with the MNTLHLTH variable. MNTLHLTH is significant at the (p<.01) level in both regressions while PHYSHLTH fails to reach significance at the (p<.10) level. The magnitude of the MNTLHLTH coefficient is similar across the standard and stand-alone models, suggesting that the MNTLHLTH variable is not capturing the emotional effects of physical conditions to a great extent, or, alternatively, that the impact of emotional health on SWB is not contingent on the presence of physical conditions. In the standard specification, each day of poor emotional health corresponds to a diminution in SWB of approximately 0.20 units--a decrease commensurate with a $14,733 decrease in family income. Each day of standalone poor emotional health corresponds to a diminution in SWB by 0.017 units--a decrease in SWB commensurate with a $9,800 decrease in family income.
Respondents whose emotional conditions led them to accomplish "less than they would have liked" in the four weeks preceding the survey--as captured by the DIDLESSE variable--experienced between a 0.371 and 0.378 decrease in SWB, depending on the specification. The monetary equivalents for DIDLESSE are $256,363 and $203,441 of annual household income. That is, the diminution in SWB associated with answering "yes" to the DIDLESSE question would be approximately offset by a $256,563 rise in household income for the standard specification and an approximately $203,441 rise in household income for the stand-along specification. Answering "yes" on DIDLESSP--accomplishing less due to physical problems--corresponds to a $84,368 change in family income. Both DIDLESSE and DIDLESSP, where present, were significant at the (p<.01) level.
Being limited in work or other activities due to emotional health (over the four weeks preceding the survey) impacts SWB negatively, to the tune of .222 hedonic units (. 185 for the stand-alone specification). The inability to do work or other activities as carefully as usual in light of physical problems is associated with a 0.140 drop in SWB. The monetary equivalents for CRELESSE are $133,019 for the standard model and $99,555 for the stand-alone model. The monetary equivalent for LIMITEDP is $83,886. CRELESSE and LIMITEDP were significant at the (p<.01) level in the standard specification, and CRELESSE was significant at the (p<.05) level in the stand-alone specification.
The three above analyses suggest that emotional health conditions, including those free of concomitant physical manifestation or symptoms, impact SWB significantly. In fact, in all of the regressions accounting for both physical and mental health, mental health uniformly bore stronger connections to SWB than physical health. The results gain significance when viewed through the lens of tort law, because many of the emotional harms found to have a significant impact on SWB in the instant inquiry would be subject to dismissal in court. The following sub-sections address the limitations of the analysis, both in general and as applied to the tort context.
A. The Distinction Between SWB and the Emotional Health
Given the conceptual similarity between SWB and emotional health, the following question warrants serious attention: are mental harms, as defined in the survey instruments MNTLHLTH, DIDLESSE, and CRELESSE, so closely related to SWB as to make the current inquiry question-begging? A full response to this question (136) merits a closer look at the construct of well-being and the language of the three mental health questions used in the analysis. MNTLHLTH asks survey respondents to tally the number of days in which their mental health, "including stress, depression, and problems with emotions," was "not good." (137) DIDLESSE and CRELESSE both inquire about "emotional problems (such as feeling depressed or anxious)" that limit daily activities. (138) While the precise conceptual bounds of terms such as emotional problems and mental health are not self-defining, the questions appear to confine themselves to experienced emotions or moods rather than higher-order evaluative judgments about individual flourishing.
In short, SWB measures of the type employed herein are inextricably bound up with aspirations and values in a way that emotions, simpliciter, are not. The content of the mental health survey questions used in this analysis can be contrasted with that of SWB. The characterization of SWB by Diener reflects its broad conceptual reach--SWB "include[s] people's emotional reactions to events, their moods, and judgments they form about their life satisfaction, fulfillment, and satisfaction with domains such as marriage and work." (139) SWB "includes the evaluations and declarations that individuals make about the quality of their lives that are based on the review, weighting, and summation of the quality of experiences, accomplishments, relationships, and other culturally relevant and valued ways of functioning in life." (140) In short, SWB measures of the type employed herein are inextricably bound up with aspirations and values in a way that emotions are not. (141) Moreover, even if the mental health questions used herein can be reasonably interpreted as inquiries about negative affect, there is evidence that positive affect and negative affect are not simple opposites of each other; (142) rather, they are independent features of a "circumplex" of affect. (143) Thus, under accepted definitions of SWB, it is quite distinguishable from the types of mental health conditions covered by MNTLHLTH, DIDLESSE, and CRELESSE.
B. Applicability of Results to the Tort Context
1. Broad Survey Questions--What Conclusions Can We Draw?
As mentioned earlier, the emotional health variables used in this analysis reveal little about the specific nature of the emotional conditions that underlie any particular numerical survey response. Rather, the questions are framed broadly, requiring the respondents to impose their own interpretations on the questions in order to arrive at a response. In answering the MNTLHLTH question, for example, each respondent makes her own decision about whether or not a given condition merits inclusion in the tally of days in which her "mental health [was] not good." (144) To varying extents, this problem applies to all three sets of mental health variables used in this analysis. The CRELESSE and DIDLESSE variables, which frame emotional health in functional rather than clinical terms, afford a considerable amount of interpretive freedom to respondents in distinguishing between de minimis stresses and those that hinder regular functioning enough to warrant reporting them. As a result, each numerical survey response reflects a distribution of health issues, both in terms of their clinical characterization and their severity. As mentioned previously, the survey questions also reveal nothing about the length of time that the respondents have suffered from or adapted to the conditions that underlie their responses. The broad framing of the emotional health questions would thus effectively preclude any attempt to generate precise damage awards for specific emotional health issues. (145)
The nature of the self-assessed health measures also renders the comparison between physical health and emotional health problematic. Without any guidance as to the nature of each type of injury, any cross-category comparison is fundamentally an apples-to-oranges comparison. In other words, if there is a systematic difference in the threshold for inclusion across the physical and mental categories, the comparison will be difficult to interpret. For example, if people count de minimis physical pain in the tally, but only count extremely serious mental issues in the count, the comparison between MNTLHLTH and PHYSHLTH will be difficult to interpret. Though the similar wording and scaling of the physical and emotional health survey questions makes direct comparisons tempting, this aspect of the results should be interpreted cautiously.
Ultimately, however, the more modest goal of this inquiry is to test whether emotional harms--including instances of stand-alone emotional distress--impact SWB at all. Recall that claims based on emotional distress of any severity would be barred in a variety of circumstances: (1) in jurisdictions that deny NIED claims altogether, if the harm is the result of negligence; (2) in IIED cases where the offending behavior is not of a sufficiently egregious nature, regardless of the severity of the resulting emotional harm; (3) in jurisdictions that require a physical impact or manifestation, regardless of the severity of the emotional disturbance. Thus, although the GSS survey questions obscure the precise nature of the emotional harms that underlie the survey responses, such aspects are not of paramount importance given the limited goals of this inquiry.
2. External Validity
The use of survey data from outside the tort context to bear upon issues within tort raises issues of external validity. To some extent, this analysis revolves around the notion that harms are harms, whether they are due to the tortious behavior of another or not. However, psychological studies suggest that this is not the case. Causal attribution concerning the etiology of health conditions impacts both the experienced severity of the condition as well as the coping process. (146) In other words, individuals who blame others for their health conditions tend to have worse symptoms and recover more slowly. Thus, to the extent that some of the emotional health conditions experienced by the survey populations used in the instant inquiry were not the result of tortious behavior or otherwise not the fault of others, the figures presented in Tables 1-3 are likely conservative. (147) Again, were the goals of this inquiry to precisely determine monetary equivalents for specific health conditions caused by tortious behavior, the use of survey data from outside the tort context would be inappropriate. In the context of this analysis, however, the conservative nature of the estimates actually bolsters the claims. That is, if emotional harms, many of which are likely not the result of tortious behavior, bear a significant negative impact on SWB, then emotional damages sustained at the hands of a tortfeasor would--per the above findings on causal attribution--presumably bear an even greater impact on SWB.
Of course, the relevance of the results still hinges on one's belief in the importance of hedonic states. Ultimately, those inclined to think of legally cognizable injuries in partially hedonic terms would place more stock in the findings. Moreover, the results presented herein are not directly responsive to the many practical arguments often used to defend the physical and emotional distinction in tort law. To the extent that such practical arguments serve as pretext for a more fundamental suspicion about the importance of emotional health, though, the results suggest that the current treatment of emotional harms in tort are misguided. Further, if malingering can be curbed through vigorous cross-examination and expert testimony, the rules of tort law have little reason to disfavor emotional harms, as a category, by subjecting them to harsher standards. Rather, the jury or judge should be allowed to assess the merits of each case on an equal footing, regardless of the category of injury.
Notwithstanding significant strides towards what might be called mental health parity in tort law, the treatment of physical injuries remains privileged. Using survey data on SWB, this paper shows that a range of emotional harms that might be subject to dismissal in courts--including standalone claims of emotional distress--bear a significant impact on SWB. To the extent that the unequal treatment of physical and emotional harms is based not upon practical concerns but upon the belief that mental health is less important to the quality of life the findings presented here challenge the distinctions currently drawn in tort law.
(1.) Following the convention of some scholars in this area, the terms "happiness" and "subjective well-being" are used interchangeably to refer to the various types of survey-based measures employed in these analyses.
(2.) Philosophical reflections on happiness date back to ancient civilization and contemporary social science journals have been engaged in the study of well-being for around thirty years, the first notable piece being Richard Easterlin's seminal 1974 article examining the relationship between money and well-being. See Richard A. Easterlin, Does Economic Growth Improve the Human Lot? Some Empirical Evidence, in NATIONS AND HOUSEHOLDS IN ECONOMIC GROWTH 89, 89-125 (Paul A. David & Melvin W. Reder eds., 1974).
(3.) See, e.g., Cass R. Sunstein, Illusory Losses, 37 J. LEGAL STUD. S 157 (2008); Peter A. Ubel & George Loewenstein, Pain and Suffering Awards." They Shouldn't Be (Just) About Pain and Suffering, 37 J. LEGAL STUD. S195 (2008); Rick Swedloff & Peter H. Huang, Tort Damages and the New Science of Happiness, 85 IND. L.J. 553 (2010).
(4.) See Mark A. Cohen, The Effect of Crime on Life Satisfaction, 37 J. LEGAL STUD. S325 (2008).
(5.) See John Bronsteen, Christopher Buccafusco & Jonathan S. Masur, Welfare as Happiness, 98 GEO. L.J. 1583 (2010); Eric Posner & Matthew Adler, Happiness Research and Cost-Benefit Analysis, 37 J. LEGAL STUD. S253 (2008).
(6.) See, e.g., David Leonhardt, He's Happier, She's Less So, N.Y. TIMES, September 26, 2007, www.nytimes.com/2007/09/26/business/26leonhardt.html; Jeremy Capalan, Getting Serious About Happiness, TIME, Apr. 3, 2007, www.time.com/time/healtlYarticle/0,8599,1606395,00.html; Claudia Wallis, The New Science of Happiness, TIME, Jan. 9, 2005, www.time.com/time/magazine/article/0,9171,1015902,00.html.
(7.) See RESTATEMENT (THIRD) OF TORTS: LIAB. FOR PHYSICAL & EMOTIONAL HARM ch. 8, scope note (Tentative Draft No. 5, 2007).
(8.) See id.
(9.) See, e.g., Reynolds v. Highland Manor, Inc., 954 P.2d 11, 13 (Kan. Ct. App. 1998) ("[E]motional distress is a common experience of life and is usually trivial." (quoting Freeman v. Kansas State Network, Inc., 719 F. Supp. 995, 1001 (D. Kan. 1989))); Soucek v. Banham, 503 N.W.2d 153, 164 (Minn. Ct. App. 1993) ("[E]motional distress is highly subjective, often transient, and easily alleged." (quoting Garvis v. Emp'rs Mut. Cas. Co., 497 N.W.2d 254, 257 (Minn. 1993))).
(10.) See Betsy J. Grey, Neuroscience and Emotional Harm in Tort Law." Rethinking the American Approach to Free-Standing Emotional Distress Claims, in 13 LAW AND NEUROSCIENCE 203 (Michael Freeman ed., 2011).
(11.) See Nancy Levit, Ethereal Torts, 61 GEO. WASH. L. REV. 136, 163-74 (1992).
(12.) See Martha Chamallas, The Architecture of Bias: Deep Structures in Tort Law, 146 U. PA. L. REV. 463,468 (1998).
(13.) See, e.g., GTE Sw., Inc. v. Bruce, 998 S.W.2d 605, 618 (Tex. 1999) ("Emotional distress includes all highly unpleasant mental reactions such as embarrassment, fright, horror, grief, shame, humiliation, and worry."); Campos v. Ysleta Gen. Hosp., 836 S.W.2d 791, 795 (Tex. App. 1992) ("Severe emotional distress is generally defined as painful emotions such as grief, severe disappointment, indignation, wounded pride, shame, despair or public humiliation.").
(14.) See RESTATEMENT (THIRD) OF TORTS: LIAB. FOR PHYSICAL & EMOTIONAL HARM ch. 8, scope note (Tentative Draft No. 5, 2007) (defining "physical impairment of the human body" as "physical injury, illness, disease, and death" and noting that the "definition of physical harm is meant to preserve the ordinary distinction between physical harm and emotional disturbance.").
(15.) See Levit, supra note 11; John J. Kircher, The Four Faces of Tort Law: Liability for Emotional Harm, 90 MARQ. L. REV. 789 (2007).
(16.) For exhaustive legal history of the distinction between mental and physical injuries, see Levit, supra note 11, at 140-58; see also Kircher, supra note 15 (offering an exhaustive analysis of the treatment of mental and physical harms in tort, including fifty state surveys of the law).
(17.) The case of Lynch v. Knight is frequently cited for the proposition that pure emotional disturbance--mental harm not accompanied by physical injury--is not a legally cognizable harm. See Lynch v. Knight,  11 Eng. Rep. 854 (H.L.) (appeal taken from Ir.). For cases echoing this sentiment, see, e.g., Monteleone v. Coop. Transit Co., 36 S.E.2d 475,475-76 (W. Va. 1945) ("There can be no recovery in tort for an emotional and mental trouble alone without ascertainable physical injuries."), overruled by Heldreth v. Marrs, 425 S.E.2d 157 (W. Va. 1992); Int'l & G.N.R. Co. v. Sammon, 79 S.W. 854, 855-56 (Tex. Civ. App. 1904) ("A distinction must be observed between those cases which establish what is called the 'Texas rule,' which allows damages for mental anguish on breach of contract, and those which deny recovery for mental anguish in cases of tort unaccompanied by physical injury.").
(18.) See Levit, supra note 11, at 140-58.
(19.) Id. at 141-42.
(20.) See, e.g., State Rubbish Collectors Ass'n v. Siliznoff, 240 P.2d 282 (Cal. 1952) (allowing recovery where the defendant intentionally subjected the plaintiff to serious mental distress of which physical injury resulted); Emden v. Vitz, 198 P.2d 696, 699 (Cal. Dist. Ct. App. 1948) ("[I]f the primal cause of this [physical] injury is tortious, it is immaterial whether it is direct, as by a blow, or indirect through some action upon the mind.").
(21.) RESTATEMENT (SECOND) OF TORTS [section] 46(1) (1965).
(22.) Stancuna v. Schaffer, 998 A.2d 1221, 1227 (Conn. App. Ct. 2010) (quoting Little v. Yale Univ., 884 A.2d 427, 431-32 (Conn. App. Ct. 2005)).
(23.) See Palmer v. Arkansas Council on Econ. Educ., 40 S.W.3d 784, 791-92 (Ark. 2001).
(24.) See Johnson v. Methodist Hosp., 226 S.W.3d 525, 529 (Tex. App. 2006).
(25.) See Jaynes v. Strong-Thorne Mortuary, Inc., 954 P.2d 45, 50-51 (N.M. 1997).
(26.) See Thomas v. BSE Indus. Contractors, Inc., 624 So. 2d 1041, 1043 (Aia. 1993).
(27.) Howell v. New York Post Co., 612 N.E.2d 699, 702 (N.Y. 1993).
(28.) E.g., Jones v. Warner, 686 S.E.2d 835, 839 (Ga. Ct. App. 2009) (quoting Abdul-Malik v. AirTran Airways, 678 S.E. 2d 555 (Ga. Ct. App. 2009)).
(29.) Harris v. Kreutzer, 624 S.E.2d 24, 34 (Va. 2006).
(30.) See J. Stanley McQuade, The Eggshell Skull Rule and Related Problems in Recovery for Mental Harm in the Law of Torts, 24 CAMPBELL L. REV. 1 (2001).
(31.) RESTATEMENT (THIRD) OF TORTS: LIAB. FOR PHYSICAL & EMOTIONAL HARM [section] 46 cmt. i (Tentative Draft No. 5, 2007).
(32.) See Kircher, supra note 15.
(33.) See, e.g., McCardle v. George B. Peck Dry Goods Co., 177 S.W. 1095, 1096 (Mo. Ct. App. 1915) ("The rule of law, in this respect, may be stated in this way: If the negligent act causes fright and terror, but not bodily injury, there is no liability, though bodily injury should result from the fright. But, if the negligence causes physical injury and fright, there is liability for both .... ") abrogated by Bass v. Nooney Co., 646 S.W.2d 765 (Mo. 1983).
(34.) See Kircher, supra note 15, at 806-31.
(35.) See id. at 810-31.
(36.) Hawes v. Germantown Mut. Ins. Co., 309 N.W.2d 356, 360 (Wis. Ct. App. 1981).
(37.) O'Donnell v. HCA Health Servs. of New Hampshire, Inc., 883 A.2d 319, 324 (N.H. 2005).
(38.) Dillon v. Legg, 441 P.2d 912, 924-25 (Cal. 1968).
(39.) Id. at 924-25; see generally Kircher, supra note 15.
(40.) J. Mark Appleberry, Negligent Infliction of Emotional Distress. A Focus on Relationships, 21 AM. J.L. & MED. 301,309-10 (1995).
(41.) Dillon, 441 P.2d at 924-25.
(42.) See, e.g., Paugh v. Hanks, 451 N.E.2d 759, 763-65 (Ohio 1983); Graves v. Estabrook, 818 A.2d 1255, 1257-59 (N.H. 2003); Kinard v. Augusta Sash & Door Co., 336 S.E.2d 465,467 (S.C. 1985); see generally Kircher, supra note 15, at 824.
(43.) See Levit, supra note 11, at 146.
(44.) See, e.g., Ware v. ANW Special Educ. Coop. No. 603, 180 P.3d 610, 618 (Kan Ct. App. 2008) (citing RESTATEMENT (SECOND) OF TORT [section] 436A, cmt. b (1965)).
(45.) RESTATEMENT (SECOND)OF TORTS [section] 436A, cmt. b (1965).
(46.) See RESTATEMENT (THIRD) OF TORTS: LIAB. FOR PHYSICAL & EMOTIONAL HARM [section] 4 (2010).
(47.) See Thomas v. Brewer's Food Mart, Inc., No. 100,972, 2009 WL 4639572, *5 (Kan. Ct. App. 2009).
(48.) Reynolds, 954 P.2d at 13.
(49.) Soucek, 503 N.W.2d at 164 (quoting Garvis v. Emp'rs Mut. Cas. Co., 497 N.W.2d 254, 257 (Minn. 1993)).
(50.) Renville v. Fredrickson, 101 P.3d 773, 776 (Mont. 2004) (quoting Sacco v. High Country Indep. Press, Inc., 896 P.2d 411,426 (Mont. 1995)).
(51.) Rodrigues v. State, 472 P.2d 509, 520 (Haw. 1970).
(52.) See, e.g., RICHARD EPSTEIN, TORTS 275 (1999) (Emotional harms are often "so small that the law should take no notice of them, given the extensive costs of cranking up the legal system.").
(53.) Deutsch v. Shein, 597 S.W.2d 141,146 (Ky. 1980).
(54.) Brown v. Philadelphia Coll. of Osteopathic Med., 674 A.2d 1130, 1135 (Pa. Super. Ct. 1996) (quoting Potere v. City of Philadelphia, 112 A.2d 100, 104 (Pa. 1955)).
(55.) See 1.R.C. [section] 104(a) (2006) ("emotional distress shall not be treated as a physical injury or physical sickness").
(56.) J. Martin Burke & Michael K. Friel, Getting Physical: Excluding Personal Injury Awards Under the New Section 104(A)(2), 58 MONT. L. REV. 167, 184 (1997).
(57.) See, e.g., Richard Abel, General Damages are Incoherent, Incalculable, Incommensurable, and Inegalitarian (But Otherwise a Great Idea), 55 DEPAUL L. REV. 253 (2006); Joseph H. King, Jr., Pain and Suffering, Noneconomic Damages, and the Goals of Tort Law, 57 SMU L. REV. 163 (2004); Mark Geistfeld, Placing a Price on Pain and Suffering: A Method for Helping Juries Determine Tort Damages for Nonmonetary Injuries, 83 CALIF. L. REV. 773, 786 (1995); Randall R. Bovbjerg, Frank A. Sloan & James F. Blumstein, Valuing L([e and Limb in Tort: Scheduling "Pain and Sufjering, "' 83 Nw. U. L. REV. 908 (1989).
(58.) See RESTATEMENT (THIRD) OE TORTS: LIAB. FOR PHYSICAL & EMOTIONAL HARM ch.8, scope note (Tentative Draft No. 5, 2007) ("[G]iving legal credence to and permitting recovery for emotional disturbance may increase its severity.").
(59.) See Wiley Mittenberg et al., Identification of Malingered Head Injury on the Wechsler Memory Scale- Revised, 5(1) PSYCHOLOGICAL ASSESSMENT 34 (1993) ("[A] variety of methods are available to detect malingering in cases of head trauma.").
(60.) See, e.g., John E. Meyers & Anh Diep, Assessment of Malingering in Chronic Pain Patients Using Neuropsychological Tests, 7 APPLIED NEUROPSYCHOLOGY 133, 133-35 (2000) (finding evidence of malingering in chronic pain claimants); Wiley Mittenberg et al., Base Rates of Malingering and Symptom Exaggeration, 24 J. CLINICAL & EXPERIMENTAL NEUROPSYCHOLOGY 1094, 1094-99 (2002) (presenting results of a survey on probable malingering indicating lower base rates of malingering for depressive and anxiety disorders than for physical pain and mild head injury).
(61.) See, e.g., Robert L. Dupont et al., Economic Costs of Anxiety Dis'orders, 2 DEPRESSION &ANX1ETY 167 (1996).
(62.) See Linda G. Russek et al., The Harvard Mastery of Stress Study 35-Year Follow-up: Prognostic Significance of Patterns of Psychophysiological Arousal and Adaptation, 52 PSYCHOSOMATIC MED. 271,278 (1990).
(63.) See Robert J. Rhee, A Principled Solution for Negligent Infliction of Emotional Distress Claims, 36 ARIZ. ST. L.J. 805, 842 (2004); Martha Chamallas, Removing Emotional Harm from the Core of Tort Law, 54 VAND. L. REV. 751,752 (2001); Chamallas, supra note 12, at 498 ("[W]omen have taken on the lion's share of the emotional work in our society."); Lucinda M. Finley, Female Trouble. The Implications of Tort Reform for Women, 64 TENN. L. REV. 847, 860 (1997).
(64.) See, e.g., Richard A. Easterlin, Income and Happiness. Towards a Unified Theory, 111 ECON. J. 465 (2001).
(65.) See Russell P. D. Burton et al., Roles, Race and Subjective Well-Being: A Longitudinal Analysis of Elderly Men, 28 Soc. INDICATORS RES. 137 (1993), available at http://www.j stor.org/stable/27522665.
(66.) See Betsey Stevenson & Justin Wolfers, The Paradox of Declining Female Happiness, 1 AM. ECON. J.: ECON. POL'Y 190 (2009).
(67.) See Alois Stutzer & Bruno S. Frey, Does Marriage Make People Happy, or Do Happy People Get Married?, 35 J. SOCIO-ECONS. 326 (2006).
(68.) See Bruno S. Frey et al., Does Watching TV Make Us Happy?, 28 J. ECON. PSYCHOL. 283 (2007).
(69.) See Bruno S. Frey & Alois Stutzer, Happiness, Economy and Institutions, 110 ECON. J. 918 (2000).
(70.) See Andrew J. Oswald & Nattavudh Powdthavee, Death, Happiness, and the Calculation of Compensatory Damages, 37 J. LEGAL STUD. 8217 (2008).
(71.) See Cohen, supra note 4, at S325.
(72.) See Christopher G. Ellison, Religious Involvement and Subjective Well-Being, 32 J. HEALTH & SOC. BEHAV. 80 (1991), available at http:jstor.org/stable/2136801.
(73.) See Alex C. Michalos, Education, Happiness and Wellbeing, 87 Soc. INDICATORS RES. 347 (2008).
(74.) See Rafael Di Tella et al., Preferences over Inflation and Unemployment: Evidence from Surveys of Happiness, 91 AM. ECON. REV. 335 (2001).
(75.) See David G. Blanchflower & Andrew J. Oswald, Money, Sex and Happiness." An Empirical Study, 106 SCANDINAVIAN J. ECONS. 393 (2004).
(76.) See Elizabeth Sparrow, The Information Dividend. Why IT Makes you "Happier," BRIT. COMPUTER SOC'Y (Sept. 2010), http://www.bcs.org/upload/ pdf/info-dividend- full-report.pdf.
(77.) See, e.g., TOM W. SMITH, PETER V. MARSDEN, HOUT MICHAEL, & KIM JIBUM, GENERAL SOCIAL SURVEYS, 1972-2010: CUMULATIVE CODEBOOK 327 (2010), available at http://publicdata.norc.org:41000/gss/documents//BOOK/GSS_Codebook.pdf (measuring well-being with the survey question "Taken all together, how would you say things are these days--would you say that you are very happy, pretty happy, or not too happy?"); see also infra Part III.A.
(78.) See Daniel Kahneman, Alan B. Krueger, David A. Schkade, Norbert Schwarz, & Arthur A. Stone, A Survey Method for Characterizing Daily Life Experience." The Day Reconstruction Method, 3 SCIENCE 1776-1780 (2004); Christie Napa Scollon et al., Experience Sampling: Promises and Pitfalls, Strengths and Weaknesses, 4 J. HAPPINESS STUD. 5 (2003).
(79.) See Napa Scollon et al., supra note 78.
(80.) See Kahneman et al., supra note 78.
(81.) See Michael Eid & Ed Diener, Global Judgments of Subjective Well-Being: Situational Variability and Long-Term Stability, 65 Soc. INDICATORS RES. 245, 262-68 (2004).
(82.) See id.
(83.) See Ed Diener, Assessing Subjective Well-Being: Progress and Opportunities, 31 (2) Soc. INDICATORS RES. 103, 146 (1994).
(84.) See Ed Sandvik et al., Subjective Well-Being: The Convergence and Stability of Self-Report and Non-Self-Report Measures, 61 (3) J. PERSONALITY 317 (1993).
(85.) See Jonathan Shedler et al., The Illusion of Mental Health, 48(11) AM. PSYCHOL. 1117 (1993).
(86.) See Paul Ekman et al., The Duchenne Smile: Emotional Expression and Brain Physiology 11, 58(2) J. PERSONALITY & SOC. PSYCHOL. 342, 342 (1990) (defining a Duchenne Smile as "[s]miling in which the muscle that orbits the eye is active in addition to the muscle that pulls the lip comers up"). But see Eva G. Krumhuber & Antony S.R. Mansted, Can Duchenne Smiles be Feigned? New Evidence on Felt and False Smiles, 9(6) EMOTION 807 (2009) (challenging the validity of Duchenne smiles as indicators of positive emotion).
(87.) See Betsey Stevenson & Justin Wolfers, Economic Growth and Subjective Well-Being." Reassessing the Easterlin Paradox, in BROOKINGS PAPERS ON ECON. ACTIVITY 1, 5 (2008).
(88.) See Norbert Schwarz & Fritz Strack, Evaluating One "s Life: A Judgment Model of Subjective WelLBeing, in SUBJECTIVE WELL BEING: AN INTERDISCIPLINARY PERSPECTIVE 2747 (Fritz Strack, Michael Argyle & Norbert Schwarz eds.,1991).
(89.) Time spent on childcare and work, for example, impact single-item measures of well-being more consistently than they do repeated measures such as DRM and ESM, perhaps because "[r]espondents who answer abstract evaluative questions about activities are likely to be reminded that both work and childcare are desirable aspects of their life." Daniel Kahneman & Alan B. Krueger, Developments in the Measurement of Subjective Well-Being, 20 J. ECON. PERSPECTIVES 3, 13 (2006), available at http://www.krueger.princeton.edu/PDF%20of%20Khneman %20Krueger%20paper.pdf.
(90.) Martha Nussbaum's recent commentary on the empirical study of well-being, for example, characterizes SWB measures as blunt instruments that "bully" respondents into a reductive exercise: "people are simply told that they are to aggregate experiences of many different kinds into a single whole, and the authority of the questioner is put behind that aggregation." Nussbaum goes on to note that SWB fails to account for "bad pleasures" and "good pains." Martha C. Nussbaum, Who is the Happy Warrior? Philosophy Poses Questions to Psychology, 37 J. LEGAL STUD. S2, S86-S99 (2008).
(91.) See id.
(92.) Kahneman & Krueger, supra note 89, at 7.
(93.) See id. at 7 (noting that job satisfaction is a strong predictor of employee turnover).
(94.) Id. at 9-10.
(95.) Morris A. Okun & Linda K George, Physician- and Self-Ratings of Health, Neuroticism and Subjective Well-Being Among Men and Women, 5(5) PERSONALITY & INDIVIDUAL DIFFERENCES 533,538 (1984).
(96.) See Noreen Goldman et al., The Role of Clinical Risk Factors in Understanding Self-Rated Health, 14 ANNALS EPIDEMIOLOGY 49 (2004), available at http://www.princeton.edu/-ngoldman-papers-2004 Clinical_Risk_Factors.pdf.
(97.) See B. BurstrOm & P. Fredlund, Self Rated Health: Is It as Good a Predictor of Subsequent Mortality Among Adults in Lower as Well as in Higher Social Classes?, 55( 11 ) J. EPIDEMIOLOGY COMMUNITY HEALTH 836, 838 (2001).
(98.) See C. Jagger et al., Factors Associated with Decline in Function, Institutionalization and Mortality of Elderly People, 22(3) AGE & AGEING 190 (1993); George A. Kaplan et al., Factors Associated with Change in Physical Functioning in the Elderly: A Six-year Prospective Study, 5 J. AGING & HEALTH 140 (1993).
(99.) See Seppo Miilunpalo et al., Self-Rated Health Status as a Health Measure: The Predictive Value of Self Reported Health Status on the Use of Physician Services and on Mortality in the Working-Age Population, 50(5) J. CLINICAL EPIDEMIOLOGY 517 (1997).
(100.) See generally Okun & George, supra note 95. Further, the stigma associated with mental health issues and treatment could potentially lead individuals to distort their answers to survey questions regarding their emotional health. In the healthcare context, for example, individuals sometimes remain silent about mental health conditions, avoid or discontinue treatment, and engage in harmful "self-prejudice." See, e.g., Patrick Corrigan, How Stigma Interferes with Mental Health Care, 59(7) AM. PSYCHOL. 614 (2004). However, in the context of an anonymous survey, respondents have less reason to distort their answers to avoid stigma. Moreover, the correlation between self-assessed health and various objective measures of health, as discussed above, suggests that they are reliable.
(101.) See, e.g., Bruce Headey et al., Top-Down Versus Bottom-Up Theories of Subjective Well-Being, 24 Soc. INDICATORS RES. 81 (1991), available at http://repub.eur.nl/res/pub/22191/91b-full.pdf.
(102.) See, e.g., Angus Deaton, Income, Health and Well-Being Around the World." Evidence From the Gallup World Poll, 22(2) J. ECON. PERSPS. 53 (2008), available at http://www.gallup.com/se/127634/income-health-wellbeing -around-world-evidence-gallupworld-poll.aspx.
(103.) See, e.g., Nattavudh Powdthavee & Bernard van den Berg, Putting Different Price Tags on the Same Health Condition: Re-Evaluating the Well-Being Valuation Approach, 30(5) J. HEALTH ECONS. 1032 (2011), available at http ://www.powdthavee.co.uldresources/ Health+compensation+version+2+6+11.pdf.
(104.) See Okun & George, supra note 95.
(105.) See, e.g., Abel, supra note 57; King, supra note 57, at 163.
(106.) In cases where injuries can be remedied through the replacement of a fungible (or nearly fungible) good, market value provides a clear indicator of the extent of the injury. In the context of injuries to irreplaceable or non-market goods, economics can provide only limited answers. Contingent value studies estimate the value of non-market goods through the use of surveys, posing various hypothetical scenarios to individuals and asking them to provide an amount of money that they would pay to avoid a hypothetical risk or the amount of money they would accept to assume a hypothetical risk, such as polluted air. The resulting values are called, respectively, willingness-to-pay (WTP) values and willingness-to-accept (WTA) values. Another econometric technique used to estimate the implicit value of life and limb uses consumption behavior in real market settings rather than survey data. One version of this technique employs compensating wage differentials, or differences in wage rates associated with different levels of workplace risk, to evaluate implicit values of the losses associated with such risk. Where the risk in question is the risk of death, the resulting value is called the "value of a statistical life" or VSL. VSL estimates, and those estimated by way of contingent value studies, are used primarily in the regulatory context, as opposed to trials. See Eric Posner & Cass Sunstein, Dollars and Death, 72 U. CHI. L.R. 537, 549 tbl. 2 (2005) (noting that, pursuant to executive orders 12291 and 12866, regulatory agencies have used VSL values to perform cost-benefit analyses of proposed regulations for decades).
(107.) While the adoption of the term "hedonic damages" by a number of courts might have hastened the application of happiness research to this area of study, the application of happiness research to damage issues is a natural extension of the economic analysis of damages. Moreover, the application of happiness research to civil damage awards extends beyond the scope of what are commonly called "hedonic damages." See, e.g., Sunstein, supra note 3, at S157.
(108.) Ed Diener et al., The Evolving Concept of Subjective Well-Being: The Multifaceted Nature of Happiness, in ASSESSING WELL-BEING: THE COLLECTED WORKS OF ED DIENER 69-70 (Ed Diener ed., 2009).
(109.) See Nussbaum, supra note 90, at S105.
(110.) See id. at S110-11.
(111.) Sunstein, supra note 3, at S166.
(112.) Id. at S158.
(113.) See Richard E. Lucas et al., Unemployment Alters the Set-Point for Life Satisfaction,15 PSYCHOL. SCI. 8 (2004), available at http://www.psychologicalscience.org/pdf/lucas.pdf.
(114.) See generally Scott A. Moss & Peter H. Huang, How the New Economics Can Improve Employment Discrimination Law, and How Economics Can Survive the Demise of the "Rational Actor, " 51 WM. & MARY k. REV. 183 (2009).
(115.) Ubel & Loewenstein, supra note 3, at S197.
(116.) See Swedloff & Huang, supra note 3, at 564.
(117.) See id, at 580.
(118.) See SMITH ET AL., supra note 77, at vii.
(119.) See generally id.
(120.) See id. at vii.
(121.) See id. at 1434-35, 1509-10.
(122.) SMITH ET AL., supra note 77, at 327. The SWB variable used in the current analysis is a recoded version of the GSS variable "HAPPY." In the original version of HAPPY, higher levels of SWB correspond to lower response scores; in the recode, higher response values reflect higher levels of SWB.
(123.) Id. at 1510.
(124.) Id. at 1435.
(126.) Id. at 1509.
(127.) Id. at 1434.
(129.) Importantly, ordered profit estimates--the form typically chosen for ordinal-level dependent variables--yield essentially the same results as the OLS regressions described below.
(130.) See Stevenson & Wolfers, supra note 66; Burton et al., supra note 65; Stutzer & Frey, supra note 67; Frey et al., supra note 68; Frey & Stutzer, supra note 69; Oswald & Powdthavee, supra note 70; Cohen, supra note 4; Ellison, supra note 72; Michalos, supra note 73; Tella et al., supra note 74; Blanchflower & Oswald, supra note 75; Sparrow, supra note 76.
(131.) See Stevenson & Wolfers, supra note 66.
(132.) Since the household income variable is log-transformed, each percent change in income impacts SWB, in unit terms, by 1/100th of the income coefficient. The monetary equivalents can be approximated by multiplying the ratio of the coefficients for household income and the relevant health variable by the average sample household income. The product of these two terms is the approximate amount of money that would bring an individual of average household income to the same level of SWB that would be predicted, other things equal, in the absence of the health condition. For example, in the first regression model presented, the variable MNTLHLTH has a coefficient of (0.020) and log of household income has a coefficient of (0.055). Multiplying (.020/.055) by $40,516.70 yields a monetary equivalent of $14,733.
(133.) One plausible explanation for the non-significance of age and age-squared in the latter two regressions is the presence of the health variables, which likely capture some of the SWB variance explained by age.
(134.) See, e.g., Powdthavee & van den Berg, supra note 103.
(135.) See, e.g., Andrew J. Oswald & Nattavudh Powdthavee, Does Happiness Adapt? A Longitudinal Study of Disability with Implications for Economists and Judges, 92 J. PUB. ECON. 1061 (2008).
(136.) The quickest answer to the question is statistical in nature. If the various measures of mental health and the construct of SWB are, in fact, identical, they would bear correlations that approach unity, rendering other aspects of life insignificant in determining SWB. However, the correlations between SWB and emotional health variables used in this analysis bear far more modest correlations, ranging from approximately .3 to .4.
(137.) SMITH ET AL., supra note 77, at 1510.
(138.) Id. at 1435.
(139.) Ed Diener et al., Personality, Culture and Subjective WelLBeing: Emotional and Cognitive Evaluations of Life, 54 ANN. REV. PSYCHOL. 403, 404 (2003), http://education.ucsb.edu/janeconoley/ed 197/documents/DienerPersonalitycultureandsubject ivewell-being.pdf.
(140.) Corey L.M. Keyes, Mental Health in Adolescence: Is America's Youth Flourishing?, 76(3) AM. J. ORTHOPSYCHIATRY 395, 395 (2006).
(141.) To the extent that repeated measures of SWB, such as ESM and DRM, more closely capture moment-by-moment affective experience without also capturing the cognitive/reflective component of SWB, such items would be harder to distinguish from the emotional health measures used herein.
(142.) See Ed Diener & Robert A. Emmons, The Independence of Positive and Negative Affect, 47(5) J. PERSONALITY & SOC. PSYCHOL. 1105 (1985). This conclusion is not entirely free of controversy. See, e.g., Boris Egloff, The Independence of Positive and Negative Affect Depends on the Affect Measure, 25 PERSONALITY & INDIVIDUAL DIFFERENCES 1101 (1998). However, since SWB has a cognitive/evaluative component in addition to the affective components, the distinction between SWB and the emotional health variables used herein can still be maintained.
(143.) See, e.g., Lisa Feldman Barrett & James A. Russell, The Structure of Current Affect: Controversies and Emerging Consensus, 8 CURRENT DIRECTIONS 1N PSYCHOL. SCI. 10 (1999), available at http://www.affective-science.org/ pubs/1999/FBRussel199.pdf.
(144.) SMITH ET AL., supra note 77, at 1510.
(145.) With more detailed information about the underlying emotional conditions, one could compare damage awards issued in court cases with monetary equivalents for similar conditions. Such a comparison would indicate whether certain types of injuries are overvalued or undervalued in court, at least in terms of their impact on SWB. Unfortunately, the range of health conditions covered and the level of generality at which the health questions are framed in the GSS make it difficult to map such survey responses onto tort verdicts in any coherent way.
(146.) See, e.g., Charlotte vanOyen Witvliet, Thomas E. Ludwig & Kelly L. Vander Laan, Granting Forgiveness or Harboring Grudges: Implications for Emotion, Physiology, and Health, 12(2) PSYCHOL. SCI. 117 (2001), http://datapsych.udel.edu/rsimmons/PSYCH467/witvliet %20et%20al.,%202001.pdf.
(147.) The estimates might also be conservative for another reason--because they are not motivated by financial gain, respondents have less of a reason to exaggerate their conditions or limitations.
David DePianto, Visiting Assistant Professor of Law, Sandra Day O'Connor College of Law at Arizona State University. J.D., Georgetown University Law Center; Ph.D. Univ. of California, Berkeley. For their assistance on this project, I am indebted to Robert Cooter, Robert MacCoun, Justin McCrary, Jonathan Masur, Peter Huang, David Kwok and Ashley Rubin.
Table 1--OLS Regression Results Dependent Variable: Subjective Well Being Independent Variables of Interest: "MNTLHLTH" (# of days in past month that mental health was "not good") Variable Standard Regression: "Stand-Alone" Regression: MNTLHLTH -0.020 *** -0.017 *** (0.002) (0.003) PHYSHLTH -0.003 (0.002) Age -0.024 *** -0.027 *** (0.006) (0.007) Age Squared 0.000 *** 0.000 *** (0.000) (0.000) Female 0.036 0.025 (0.029) (0.037) Black -0.064 -0.011 (0.043) (0.046) 2006 -0.024 (0.032) Log HH Income 0.055 *** 0.053 ** (0.019) (0.024) Widowed -0.341 *** -0.346 *** (0.076) (0.093) Divorced -0.274 *** -0.242 *** (0.038) (0.048) Separated -0.413 *** -0.359 *** (0.075) (0.110) Never Married -0.289 *** -0.315 *** (0.039) (0.051) R-squared. .1382 .0998 N 2236 48575 N (subpop) 1474 legend: b/se (* p<.05; ** p<.01; *** p<.001); Standard errors are in parentheses Notes: Stand-Alone regression was performed on respondents who indicated, via the PHYSHLTH variable, that they experienced no bad days of physical health in the past month. "2006" is a dummy variable indicating the year the survey was administered; the reference category is the year 2004. The reference category for the marital status dummies is "married." Table 2--OLS Regression Results Dependent Variable: Subjective Well Being Independent Variables of Interest: "DIDLESSE" (emotional condition impaired work, daily activities) Variable Standard Regressions: "Stand-Alone" Regressions: DIDLESSE -0 371 *** -0.378 *** (0.056) (0.063) DIDLESSP -0 122 *** (0.036) Age -0.001 0.001 (0.006) (0.008) Age Squared 0.000 0.000 (0.000) (0.000) Female 0.046 0.050 (0.036) (0.041) Black -0.176 *** -0.165 ** (0.052) (0.066) Log HH Income 0.046 * 0.057 ** (0.024) (0.026) Widowed -0.167 ** -0.077 (0.068) (0.096) Divorced -0.166 *** -0.173 *** (0.050) (0.056) Separated -0.307 *** -0.275 ** (0.110) (0.131) Never Married -0.205 *** -0.140 ** (0.052) (0.056) R-squared .1500 .1120 N 1196 49250 N (subpop) 910 legend: b/se (* p<.05; ** p<.01; *** p<.001); Standard errors are in parentheses Notes: Stand-Alone regression was performed on respondents who indicated, via the DIDLESSP variable, that they experienced no bad days of physical health in the past 4 weeks. The reference category for the marital status dummies is "married." Table 3--OLS Regression Results Dependent Variable: Subjective Well-Being Independent Variables of Interest: "CRELESSE" (less attention on work and/or less careful work due to emotions) Variable Standard Regressions: "Stand-Alone" Regressions: CRELESSE -0.222 *** -0.185 ** (0.054) (0.079) LIMITEDP -0.140 *** (0.045) Age -0.001 0.005 (0.006) (0.008) Age Squared 0.000 -0.000 (0.000) (0.000) Female 0.038 0.055 (0.037) (0.039) Black -0.156 *** -0.184 *** (0.054) (0.070) Log HH Income 0.053 ** 0.057 ** (0.025) (0.027) Widowed -0.197 *** -0.099 (0.069) (0.108) Divorced -0.186 *** -0.188 *** (0.050) (0.054) Separated -0.338 *** -0.314 *** (0.107) (0.117) Never Married -0.211 *** -0.145 ** (0.054) (0.056) R-squared .1125 .0862 N 1194 49242 N (subpop) 928 legend: b/se (* p<.05; ** p<.01; *** p<.001); Standard errors are in parentheses Notes: Stand-Alone regression was performed on respondents who indicated, via the LIMITEDP variable, that they experienced no bad days of physical health in the past 4 weeks. The reference category for the marital status dummies is "married." Table 4--Monetary Equivalents for Emotional and Physical Conditions Variable Standard Regression: "Stand-Alone" Regression: MNTLHLTH $14,733 $9,800 PHYSHLTH * $2,210 N/A DIDLESSE $256,563 $203,441 DIDLESSP $84,368 N/A CRELESSE $133,020 $99,555 LIMITE DP $83,886 N/A * Not significant at the 90% (p<.10) level.
|Printer friendly Cite/link Email Feedback|
|Publication:||Law and Psychology Review|
|Date:||Jan 1, 2012|
|Previous Article:||Mind matters: mental disability and the history and future of the Americans with Disabilities Act.|
|Next Article:||Social security disability law and the obstacles facing claimants with mental disabilities.|