Lawrence S. Krieger* with Kennon M. Sheldon, Ph.D.**
George Washington Law Review
This is the first theory-guided empirical research seeking to identify the correlates and contributors to the well-being and life satisfaction of lawyers. Data from several thousand lawyers in four states provide insights about diverse factors from law school and one’s legal career and personal life. Striking patterns appear repeatedly in the data and raise serious questions about the common priorities on law school campuses and among lawyers. External factors, which are often given the most attention and concern among law students and lawyers (factors oriented towards money and status—such as earnings, partnership in a law firm, law school debt, class rank, law review membership, and U.S. News & World Report’s law school rankings), showed nil to small associations with lawyer well-being. Conversely, the kinds of internal and psychological factors shown in previous research to erode in law school appear in these data to be the most important contributors to lawyers’ happiness and satisfaction. These factors constitute the first two of five tiers of well-being factors identified in the data, followed by choices regarding family and personal life. The external money and status factors constitute the fourth tier, and demographic differences were least important. Data on lawyers in different practice types and settings demonstrate the applied importance of the contrasting internal and external factors. Attorneys in large firms and other prestigious positions were not as happy as public service attorneys, despite the far better grades and pay of the former group; and junior partners in law firms were no happier than senior associates, despite the greatly enhanced pay and status of the partners. Overall, the data also demonstrate that lawyers are very much like other people, notwithstanding their specialized cognitive training and the common perception that lawyers are different from others in fundamental ways.
* Clinical Professor of Law, Florida State University College of Law. ** Professor, Department of Psychological Sciences, University of Missouri (Columbia). We particularly appreciate the dedication and focused efforts of the Lawyer Assistance Program directors and bar administrators who made this study possible. Special appreciation also goes to David Shearon, who generously provided his thrivinglawyers.org website for management of continuing legal education records related to this study. We thank Sarah Spacht for research assistance, Hunter Whaley for research assistance and editing suggestions to complete the draft, Mike Prentice and Mark White for technical assistance with data compilation and expression, and Jerry Organ and Daisy Floyd for thoughtful comments on an earlier draft. Deficiencies remain the responsibility of the authors.
February 2015 Vol. 83 No. 2
2015] WHAT MAKES LAWYERS HAPPY? 555
Additional measures raised concerns. Subjects did not broadly agree that the behavior of judges and lawyers is professional, or that the legal process reaches fair outcomes; and subjects reported quite unrealistic earnings expectations for their careers when they entered law school. Implications for improving lawyer performance and professionalism, and recommendations for law teachers and legal employers, are drawn from the data.
TABLE OF CONTENTS
INTRODUCTION …………………………………………. 557
I. BACKGROUND AND PURPOSES FOR THE CURRENT STUDY ………………………………………….. 559
II. THEORY UNDERLYING THE STUDIES……………….. 562 A. Subjective Well-Being as a Measure of Happiness … 562 B. Self-Determination Theory …………………….. 564
III. FOUNDATIONAL STUDIES OF LAW STUDENTS ……….. 565
IV. THE CURRENT STUDY ……………………………. 569 A. Measures…………………………………….. 569 B. The Bar Member Sample ……………………… 570
V. HYPOTHESES …………………………………….. 573 VI. PRIMARY FINDINGS ………………………………. 576 A. Grades, Law Review, and Money Issues ………… 576 1. Law School Grades ……………………….. 576 2. Law Journal Membership………………….. 577 3. Law School Debt and Income After Graduation ………………………………. 578 B. Psychological Need Satisfaction ………………… 579 C. Motivation …………………………………… 579 D. Values ………………………………………. 580 1. Values and Professionalism ………………… 581 E. Autonomy Support……………………………. 582 1. Replicating the Path Model for Autonomy Support, Motivation, and Well-Being ……….. 583 F. Brief Discussion of Primary Findings …………… 584 VII. SECONDARY FINDINGS ……………………………. 585 A. Alcohol Consumption…………………………. 586 B. Attorney Well-Being in Contrasting Work Settings and Practice Types: Testing the Internal-External Factors Dichotomy……………………………. 587 1. Findings………………………………….. 590
THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554
2. Do Attorney Preferences and Work Settings Affect the Factors That Promote Their WellBeing? …………………………………… 593
C. Other Work Variables…………………………. 594 1. Hours Worked, Firm Size, and Billable Hours.. 594 a. Total Hours and Billable Hours ………… 595 b. Size of Law Firm ……………………… 596 2. Position Within Law Firm …………………. 597
3. Litigation Practice and Private and Public Attorneys ………………………………… 599
4. Pro Bono and Community Service Work ……. 599 D. Personal Demographics ……………………….. 600 1. Age……………………………………… 600 2. Gender ………………………………….. 601 3. Race and Ethnicity ……………………….. 602 4. Marriage and Social Support ………………. 602 5. Children …………………………………. 603
E. Law School Ranking………………………….. 604
F. Personal Life and Balance Choices …………….. 607 1. Physical Activities: Exercise, Sports and Martial Arts, and Yoga and Tai Chi…………………607 2. Vacations ………………………………… 608 3. Religious and Spiritual Practice…………….. 609 G. Smaller City Life and Practice …………………. 611 H. Perceptions of Professionalism and Faith in the Justice System………………………………… 611 I. Expected Earnings Compared with Actual Earnings. 613 J. Brief Discussion of Secondary Findings …………. 614 VIII.
A. What Makes Lawyers Happy? …………………. 617
B. Lawyers Are Not Different from Other People with Regard to Their Happiness and Satisfaction ……… 621
C. Improved Well-Being Implies Improved Productivity, Ethics, and Professionalism ……………………. 622
D. What the Findings Mean for Lawyers and Their Teachers and Employers ………………………. 623
IX. LIMITATIONS AND FUTURE DIRECTIONS ……………. 625
CONCLUSION …………………………………………… 626
2015] WHAT MAKES LAWYERS HAPPY? 557
INTRODUCTION “It’s pretty hard to tell what does bring happiness. Poverty an’ wealth have both failed.”1 Legal educators, attorneys, and bar leaders have expressed concern for emotional distress,2 dissatisfaction,3 and unethical or unprofessional behavior among practicing lawyers.4 There is ample literature to raise questions about the mental health of lawyers and law students5; the legal profession, as compared to other occupations,
1 KIN HUBBARD, ABE MARTIN’S BROADCAST 191 (1930). 2 See, e.g., AM. BAR ASS’N, THE REPORT OF AT THE BREAKING POINT: A NATIONAL CONFERENCE ON THE EMERGING CRISIS IN THE QUALITY OF LAWYERS’ HEALTH AND LIVES— ITS IMPACT ON LAW FIRMS AND CLIENT SERVICES (1991); SUSAN SWAIM DAICOFF, LAWYER, KNOW THYSELF: A PSYCHOLOGICAL ANALYSIS OF PERSONALITY STRENGTHS AND WEAKNESSES 3 (2004); Connie J.A. Beck et al., Lawyer Distress: Alcohol-Related Problems and Other Psychological Concerns Among a Sample of Practicing Lawyers, 10 J.L. & HEALTH 1 (1995–96); G. Andrew H. Benjamin et al., The Prevalence of Depression, Alcohol Abuse, and Cocaine Abuse Among United States Lawyers, 13 INT’L J.L. & PSYCHIATRY 233 (1990); Peter H. Huang & Rick Swedloff, Authentic Happiness & Meaning at Law Firms, 58 SYRACUSE L. REV. 335 (2008); Rebecca M. Nerison, Is Law Hazardous to Your Health? The Depressing Nature of the Law, B. LEADER, Mar.–Apr. 1998, at 14; Patrick J. Schiltz, On Being a Happy, Healthy, and Ethical Member of an Unhappy, Unhealthy, and Unethical Profession, 52 VAND. L. REV. 871, 874 (1999). The evidence, although not encouraging, is somewhat mixed; for a thoughtful overview, see NANCY LEVIT & DOUGLAS O. LINDER, THE HAPPY LAWYER: MAKING A GOOD LIFE IN THE LAW 3–7 (2010). 3 For an overview of the many surveys on lawyers’ satisfaction with their legal careers, see generally Jerome M. Organ, What Do We Know About the Satisfaction/Dissatisfaction of Lawyers? A Meta-Analysis of Research on Lawyer Satisfaction and Well-Being, 8 U. ST. THOMAS L.J. 225 (2011).
Results of lawyer job satisfaction surveys are not consistent, likely at least in part because they employ different sampling techniques and different measures to gauge satisfaction. See, e.g., John P. Heinz et al., Lawyers and Their Discontents: Findings from a Survey of the Chicago Bar, 74 IND. L.J. 735, 735–36 (1999); John Monahan & Jeffrey Swanson, Lawyers at Mid-Career: A 20-Year Longitudinal Study of Job and Life Satisfaction, 6 J. EMPIRICAL LEGAL STUD. 451, 452–55, 470 (2009) (reporting positive findings of lawyer career satisfaction, and contrasting them with other reports of high lawyer discontent). It is important to note that satisfaction specifically with career is not a focus of the current study. Rather, we sought to determine overall life satisfaction (which includes satisfaction with career) and positive or negative mood— related but more relevant issues for this study that also employ validated measures to provide reliable findings. See infra Part V. 4 Susan Daicoff discusses a “tripartite crisis,” including low professionalism, low public opinion, and high emotional distress emerging in the legal profession. DAICOFF, supra note 2, at 3; see also Schiltz, supra note 2. 5 See G. Andrew H. Benjamin et al., The Role of Legal Education in Producing Psychological Distress Among Law Students and Lawyers, 1986 AM. B. FOUND. RES. J. 225; Todd David Peterson & Elizabeth Waters Peterson, Stemming the Tide of Law Student Depression: What Law Schools Need to Learn from the Science of Positive Psychology, 9 YALE J. HEALTH POL’Y L. & ETHICS 357, 358 (2009); Kennon M. Sheldon & Lawrence S. Krieger, Does Legal Education Have Undermining Effects on Law Students? Evaluating Changes in Motivation, Values, and Well-Being, 22 BEHAV. SCI. & L. 261 (2004); see also Matthew Dammeyer & Narina Nunez, Anxiety and Depression Among Law Students: Current Knowledge and Future Directions, 23
558 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554
may well harbor a disproportionate number of unhappy people.6 While articles often include anecdotes, observations, and discussion regarding negative (and positive) aspects of law practice, the literature broadly lacks empirical data bearing on the causes or correlates of the problems noted or their possible solutions. More specifically, there has been no theory-driven empirical study investigating the experiences, attitudes, and motivations of practicing lawyers, or how those factors relate to attorney emotional health or well-being.7 The current study was conceived to address this void. Rather than addressing whether lawyers are happy, this study presents data pointing to which lawyers are more, and less, happy in the profession—and specifically why that appears to be true. This Article, then, is intended to provide practical guidance to lawyers, law students, and law teachers seeking to improve their own well-being or that of others—regardless of the level of well-being or ill-being in the profession as a whole. We also discuss important implications of these data for improved performance, productivity, and professionalism.
LAW & HUM. BEHAV. 55, 61 (1999); B.A. Glesner, Fear and Loathing in the Law Schools, 23 CONN. L. REV. 627 (1991); Gerald F. Hess, Heads and Hearts: The Teaching and Learning Environment in Law School, 52 J. LEGAL EDUC. 75 (2002); Lawrence S. Krieger, Human Nature as a New Guiding Philosophy for Legal Education and the Profession, 47 WASHBURN L.J. 247 (2008) [hereinafter Krieger, Human Nature]; Lawrence S. Krieger, Institutional Denial About the Dark Side of Law School, and Fresh Empirical Guidance for Constructively Breaking the Silence, 52 J. LEGAL EDUC. 112 (2002). 6 One of the most concerning studies includes the stark finding that attorneys had the highest rate of depression of any occupational group in the United States. William W. Eaton et al., Occupations and the Prevalence of Major Depressive Disorder, 32 J. OCCUPATIONAL MED. 1079, 1085 tbl.3 (1990). Although this study is somewhat dated, there is nothing in the literature, anecdotally or otherwise, to suggest general improvement in the legal profession. Cf. Rosa Flores & Rose Marie Arce, Why Are Lawyers Killing Themselves?, CNN (Jan. 20, 2014, 2:42 PM), http://www.cnn.com/2014/01/19/us/lawyer-suicides/ (detailing recent suicides among lawyers). If anything, given the negative economic climate and accelerating law school debt in recent years, the well-being of lawyers and law students is likely stagnant or may be eroding further. 7 However, a study with partially related goals but fundamental differences from the current study is ongoing. RONIT DINOVITZER ET AL., AFTER THE JD: FIRST RESULTS OF A NATIONAL STUDY OF LEGAL CAREERS (2004) [hereinafter AJD1]; RONIT DINOVITZER ET AL., AFTER THE JD II: SECOND RESULTS FROM A NATIONAL STUDY OF LEGAL CAREERS (2009) [hereinafter AJD2]. The After the JD study seeks to follow a large segment of U.S. lawyers admitted to practice in the year 2000. AJD1, supra, at 13. It includes a longitudinal design, but a markedly narrower focus than the current study. See id. at 89. The After the JD data include one year of bar admissions and focus specifically on satisfaction with career and job choices. Id. The current study, by contrast, surveys lawyers spanning several decades of practice, and measures depression and global well-being. The current study also employs validated measures for wellbeing, motivation, values, and supervisory support, extending the same measures from previous law student studies to provide a confident empirical context for current attorney data. Thus, for the limited number of topics addressed by both studies, the partially shared goals and very different methodologies suggest they should be viewed together for increased understanding.
2015] WHAT MAKES LAWYERS HAPPY? 559
I. BACKGROUND AND PURPOSES FOR THE CURRENT STUDY We began empirically investigating likely causes8 for the reported well-being issues of lawyers by studying the mental health of law students as they progressed through law school.9 We analyzed the emotional adjustment, life satisfaction, motivations, values, needs, and level of faculty support experienced by students at two contrasting law schools. We then began the current study, extending the same inquiries to practicing lawyers and judges in the United States. We intended this study, when considered in conjunction with the law student studies, to provide a comprehensive picture of the psychodynamics of lawyers, particularly the causes or correlates of their well-being, and to encompass initial law training and varied careers in the law. We report here data on numerous subjective and objective factors related to work and personal life that bear on lawyer well-being. Factors include, for example, the work setting, area of practice, earnings, family and social status, law school achievements, motivations, values, psychological needs, and level of supervisory support of thousands of lawyers. Importantly, the report includes the relative importance (correlation strength) of each such factor for lawyer happiness and satisfaction. The data did, as hoped, fit well with the earlier law student data to generate a coherent picture of the relevant personality dynamics of
8 The cross-sectional design of this large study focuses on correlations, and thus does not permit firm conclusions about cause and effect. This limitation is common, because the design is a virtual necessity for this type of research. See generally BRUNO S. FREY & ALOIS STUTZER, HAPPINESS AND ECONOMICS: HOW THE ECONOMY AND INSTITUTIONS AFFECT HUMAN WELLBEING 13 (2002); Sonja Lyubomirsky et al., The Benefits of Frequent Positive Affect: Does Happiness Lead to Success?, 131 PSYCHOL. BULL. 803, 804 (2005) [hereinafter Lyubomirsky et al., Positive Affect]; Sonja Lyubomirsky, Why Are Some People Happier than Others? The Role of Cognitive and Motivational Processes in Well-Being, 56 AM. PSYCHOLOGIST 239, 240 (2001) [hereinafter Lyubomirsky, Happier than Others]. Consequently, findings are reported in terms of correlations, predictive power, or apparent effects of one factor on or with another. Findings demonstrate the extent to which one variable or occurrence makes it probable that another (typically happiness or unhappiness in this study) will occur, although the precise mechanism by which the two variables may interact may be unclear. Notwithstanding the limitation of a correlational study such as this, the consistency of the many findings and the patterns they present provide substantial confidence in apparent causal relationships suggested by the data. This is particularly true because of the large sample sizes and the consistency of our findings with similar findings in previous related studies that were conducted with longitudinal designs and that reached more firm causal conclusions. We did not deem a longitudinal design practical for the current study, nor was it required to achieve the purposes of the study. 9 Sheldon & Krieger, supra note 5; Kennon M. Sheldon & Lawrence S. Krieger, Understanding the Negative Effects of Legal Education on Law Students: A Longitudinal Test of SelfDetermination Theory, 33 PERSONALITY & SOC. PSYCHOL. BULL. 883 (2007) [hereinafter Sheldon & Krieger, Understanding Negative Effects].
560 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554
law students and lawyers. Although the purposes of the study did not include determination of the overall well-being of lawyers, the current data are consistent with many previous law student findings and add support to concerns for the future well-being of lawyers expressed in those reports10 and in the literature more generally.11 Most particularly, in the context of the previous law school studies, the current data show that the psychological factors seen to erode during law school are the very factors most important for the well-being of lawyers. Conversely, the data reported here also indicate that the factors most emphasized in law schools—grades, honors, and potential career income, have nil to modest bearing on lawyer well-being. These conclusions are explained throughout the findings sections of this Article and are then addressed with brief recommendations for legal educators and employers. As a second purpose of this study, we sought to investigate a question of interest to us and likely many other people: are lawyers fundamentally different from other people regarding the sources of their happiness?12 In the common culture of the United States, lawyers appear to be viewed as different from other people in the most basic ways—particularly lawyers’ levels of honesty and integrity, the way they think, and their ability to relate to or care about others.13 The focus of this survey would provide insight into any differences between lawyers and the general population regarding their sources of happiness.14 A third primary purpose for this study, as alluded to above, was to investigate the actual importance of the principal sources of stress on law school campuses—grades, honors (exemplified by law review positions),15 law school debt, and future earnings—for life after law
10 For a summary of the findings, see infra Parts VI–VII. 11 See supra notes 4–6. 12 The definition and components of well-being and “happiness” as measured in this study are explained infra Part II. 13 “Lawyer” jokes, for instance, commonly address one or more of these negative stereotypes. See, e.g., Thomas W. Overton, Lawyers, Light Bulbs, and Dead Snakes: The Lawyer Joke as Societal Text, 42 UCLA L. REV. 1069, 1082–85 (1995). 14 For a broader consideration of differences between lawyers and other people, see DAICOFF, supra note 2, at 25. Daicoff postulates that a typical “lawyer personality” is distinguished by an ethic of justice rather than an ethic of care, introversion, the Myers-Briggs preference for thinking rather than feeling, and many other traits. Id. at 25–42. If such differences exist, they may be engendered at least in part by basic law school training. For a linguistic analysis of the depersonalization of the law student personality, see generally ELIZABETH MERTZ, THE LANGUAGE OF LAW SCHOOL: LEARNING TO “THINK LIKE A LAWYER” (2007). 15 See, e.g., Benjamin et al., supra note 5, at 247, 249; Peterson & Peterson, supra note 5, at 380, 415; Sheldon & Krieger, supra note 5, at 276 n.3.
2015] WHAT MAKES LAWYERS HAPPY? 561
school.16 The question of interest here was: are these external “grades and money” factors, which commonly define “success” among law students and lawyers, sufficiently related to happiness after graduation to merit the intensity of competition and concern invested in them?17 We sought to measure the persisting association of such factors with later attorney satisfaction and well-being and then compare those associations with the effect sizes18 for well-being of other factors over which students could exert more control—intrinsic psychological factors and choices in work and personal life. We expected that the external stressors dominating the law school experience would prove to be weak predictors of lawyer happiness. If this were true and were communicated to students, it could serve to diminish the level of anxiety and stress on campuses. The study could have implications for two other highly important considerations that relate to well-being: performance and professionalism. Performance is, of course, a primary concern for educators, employers, and lawyers themselves and has been empirically linked to well-being.19 The substantial concerns for unprofessional or unethical behavior among lawyers20 might also be addressed by clarifying the
16 See Krieger, Human Nature, supra note 5, at 306–07; see also LAWRENCE S. KRIEGER, THE HIDDEN SOURCES OF LAW SCHOOL STRESS 4 (2006) [hereinafter KRIEGER, HIDDEN SOURCES] (emphasizing that the competition for grades and high income will not determine student or lawyer well-being). These issues garner substantial attention: administrators and teachers at more than half the law schools in the United States, Canada, and Australia purchased approximately 80,000 copies of this booklet for their students from 2006 to 2014. 17 Although it is commonly believed, but not empirically proven, that such factors are major stressors for students, there is little doubt about the heightened level of distress in many law schools. One study, for example, found the levels of depression on law school campuses to be akin to those in psychiatric populations. Dammeyer & Nunez, supra note 5, at 64; see also Stephen B. Shanfield & G. Andrew H. Benjamin, Psychiatric Distress in Law Students, 35 J. LEGAL EDUC. 65, 72 (1985). 18 “Effect size” connotes the correlation strength of two variables, but does not presume a cause-effect relationship. See, e.g., BARBARA G. TABACHNICK & LINDA S. FIDELL, USING MULTIVARIATE STATISTICS 54 (6th ed. 2013). 19 DAVID G. MYERS, THE PURSUIT OF HAPPINESS 130, 134 (1992); Huang and Swedloff, supra note 2, at 337; Lyubomirsky et al., Positive Affect, supra note 8, at 846; Sheldon & Krieger, Understanding Negative Effects, supra note 9, at 893; see also infra notes 202–09 and accompanying text. 20 A particularly notable article discussing lawyer distress and dissatisfaction is Patrick Schiltz’s stark warning to law students about the “unhappy, unhealthy, and unethical profession” they are seeking to join. Schiltz, supra note 2, at 920. Other than Susan Daicoff’s consideration of lawyer personality and professional behavior, DAICOFF, supra note 2, at 102–06, it is one of the few articles that addresses in a coherent way these two seemingly distinct areas of concern about lawyers—emotional distress and lack of ethical or professional behavior. It is also likely the most frequently cited law review article on these subjects to date, see Fred R. Shapiro & Michelle Pearse, The Most-Cited Law Review Articles of All Time, 110 MICH. L. REV. 1483, 1495
562 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554
sources of lawyer well-being, because known sources of well-being in general populations appear to be identical or closely related to important sources of positive professional behavior.21 All of these considerations are discussed in the context of the data reported below.
II. THEORY UNDERLYING THE STUDIES A. Subjective Well-Being as a Measure of Happiness The term “happiness” is subject to many shades of meaning22 and might seem out of place when applied to serious professionals doing serious work. Nonetheless, most people would agree that happiness is the prime human motivator,23 and certainly lawyers go to work and students go to law school in order to further some goal related to experiencing happiness. We employed the concept of “subjective wellbeing” (“SWB”) to measure happiness in this study, as in our law student studies and in much other research based on Self-Determination Theory (“SDT”).24 We quantified SWB as the sum of life satisfaction and positive affect, or mood (after subtracting negative affect), utilizing established instruments for each factor.25 These affect and satis
(2012) (finding that this article was the fourth most-cited law review article published in 1999), and has been incorporated into numerous law school courses, Telephone Interview with Patrick J. Schiltz (2000) (informing the author that he had received approximately 300 requests from law teachers to use this article in law courses). However, as with the literature generally, this article lacks systematic empirical data to support its recommendations, a concern we seek to address with the current study. 21 Professor Krieger has argued that the sources of both attorney well-being and professional and ethical behavior are found within personality and are essentially the same psychological factors measured in this and our previous law student studies. See Lawrence S. Krieger, The Inseparability of Professionalism and Personal Satisfaction: Perspectives on Values, Integrity and Happiness, 11 CLINICAL L. REV. 425, 427–28 (2005) [hereinafter Krieger, Inseparability]; Lawrence S. Krieger, The Most Ethical of People, the Least Ethical of People: Proposing Self-Determination Theory to Measure Professional Character Formation, 8 U. ST. THOMAS L.J. 168, 169–70 (2011) [hereinafter Krieger, Most Ethical People]. For another discussion of the connections in personality between well-being and professionalism, see DAICOFF, supra note 2, at 99–112. The applicability of all such conclusions would depend on whether attorneys are similar to other people with regard to the sources of their well-being, a principal focus of the current study. 22 For summaries of different approaches to understanding happiness, see generally FREY & STUTZER, supra note 8, at 11–12; LEVIT & LINDER, supra note 2, at 18–48; MYERS, supra note 19, at 23–30; Lyubomirsky, Happier than Others, supra note 8, at 241–42. Cf. Huang & Swedloff, supra note 2, at 339. 23 See, e.g., Lyubomirsky et al., Positive Affect, supra note 8, at 846 (noting happiness as a “prevalent” desire in Western culture); Lyubomirsky, Happier than Others, supra note 8, at 239 (observing that happiness is the primary goal of human existence). 24 See infra Part II.B. 25 For an explanation of the Positive Affect/Negative Affect Scale, see David Watson et al., Development and Validation of Brief Measures of Positive and Negative Affect: The PANAS
2015] WHAT MAKES LAWYERS HAPPY? 563
faction factors provide data on complementary aspects of personal experience. Although moods are experienced as transient, they have been found to persist over time in stable ways.26 Positive and negative affect are purely subjective, straightforward experiences of “feeling good” or “feeling bad” that many people would interpret as happiness or its opposite.27 Life satisfaction, on the other hand, includes a personal (subjective) evaluation of objective circumstances—such as one’s work, home, relationships, possessions, income, and leisure opportunities. The measure of life satisfaction employed in this study is validated by its use in previous social science research and is broader than the concept of career or job satisfaction often discussed regarding lawyers’ attitudes towards their work.28 These complementary components of SWB can diverge for an individual—a person could often feel sad or “down” but also recognize her many positive life circumstances (job, family, finances, etc.); another whose life circumstances are impoverished could feel quite good much of the time. Thus, life satisfaction and affect measure somewhat different aspects of well-being.29 Combining the two variables in one SWB measure has proven an effective way to measure the global idea of a happy life in SDT research.30 Because SWB includes a combina
Scales, 54 J. PERSONALITY & SOC. PSYCHOL. 1063, 1064–65 (1988). For an explanation of the Satisfaction with Life Scale, see Ed Diener et al., The Satisfaction with Life Scale, 49 J. PERSONALITY ASSESSMENT 71, 72 (1985). The wording of the primary measures in the survey instrument may be viewed at: Lawrence S. Krieger & Kennon Sheldon, Attorney Survey, FLA. ST. U.C.L., http://www.law.fsu.edu/faculty/profiles/krieger/attorneysurvey.docx (last visited Mar. 1, 2015) [hereinafter Attorney Survey]. 26 Lyubomirsky, Happier than Others, supra note 8, at 239. Subjective evaluations of happiness also tend to be stable, despite changing experiences. MYERS, supra note 19, at 23. 27 E.g., Lyubomirsky et al., Positive Affect, supra note 8, at 816, 840, 842 (considering short-term positive mood to be the hallmark of happiness and observing happiness to involve more than the absence of negative mood or depression). 28 See, e.g., Organ, supra note 3; see also Ronit Dinovitzer & Bryant G. Garth, Lawyer Satisfaction in the Process of Structuring Legal Careers, 41 LAW & SOC’Y REV. 1 (2007). Authors addressing the question of career satisfaction do not appear to use the same, nor an established, measure, which introduces potential confusion. Monahan and Swanson measured satisfaction with both life and career in a study of University of Virginia law graduates, finding very high satisfaction in both domains. Monahan & Swanson, supra note 3, at 452, 474–75. 29 Though different in some ways, the two aspects of SWB are highly correlated. For our working sample of 6,226 bar members, the relationship of net affect with life satisfaction was .63. A perfect correlation on this scale is 1.0; a strong one is approximately .40 or greater. 30 See Edward L. Deci & Richard M. Ryan, The ‘‘What’’ and ‘‘Why’’ of Goal Pursuits: Human Needs and the Self-Determination of Behavior, 11 PSYCHOL. INQUIRY 227, 243–44 (2000); see also FREY & STUTZER, supra note 8, at 11–12; Ed Diener, Assessing Subjective Well-Being: Progress and Opportunities, 31 SOC. INDICATORS RES. 103, 146–48 (1994) (suggesting multiple scores capturing multiple aspects of SWB, including life satisfaction among others, likely to lead to more sophisticated theories and understanding).
564 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554
tion of these critical but somewhat different aspects of personal experience, we use these and other terms, depending on context, when referring to the concept of happiness.31
B. Self-Determination Theory Both this study and our previous law student research were guided by Self-Determination Theory, a comprehensive theory of human motivation that has been prominent in the psychological literature for more than forty years.32 Tenets of SDT include that all human beings have certain basic psychological needs—to feel competent/effective, autonomous/authentic, and related/connected with others.33 These experiences are considered needs because they produce wellbeing or a sense of thriving34 in subjects, and because a lack of these experiences generates angst, low mood, or low vitality.35 SDT also broadly considers the well-being impacts of different values, goals, and motivations at the basis of behavior. Values or goals such as personal growth, love, helping others, and building community are considered “intrinsic,” while “extrinsic” values include affluence, beauty, status, and power.36 Similarly, motivation for behavior is distinguished based on the locus of its source, either “internal” (the behavior is inherently interesting and enjoyable, or it is meaningful because it furthers one’s own values) or “external” (behavior is compelled by
31 For example, “well-being” and “subjective well-being” are largely interchangeable, but the latter specifically refers to the term of art defined here. “Well-being” and “happiness” are also generally interchangeable. Lyubomirsky, Happier than Others, supra note 8, at 239 n.1. These and other terms, including “satisfaction,” are used in this Article separately or in combination to indicate shades of meaning appropriate to the specific discussion context. 32 See generally Sheldon & Krieger, supra note 5, at 263–64; see also Richard M. Ryan & Edward L. Deci, Self-Determination Theory and the Role of Basic Psychological Needs in Personality and the Organization of Behavior, in HANDBOOK OF PERSONALITY: THEORY & RESEARCH 654, 655–56 (Oliver P. Johns et al. eds., 3d ed. 2008). 33 Kennon M. Sheldon et al., What Is Satisfying About Satisfying Events? Testing 10 Candidate Psychological Needs, 80 J. PERSONALITY & SOC. PSYCHOL. 325, 326 (2001). Although selfesteem was also found to be an important predictor of well-being, we did not include it in this study. The instrument was exceptionally long and our previous studies indicated a subordinate role for self-esteem, because it did not also impact performance as did the other three needs. See Sheldon & Krieger, Understanding Negative Effects, supra note 9, at 884; see also Harry T. Reis et al., Daily Well-Being: The Role of Autonomy, Competence, and Relatedness, 26 PERSONALITY & SOC. PSYCHOL. BULL. 419 (2000); Ryan & Deci, supra note 32, at 654–78. 34 “Thriving” in this Article refers to a combination of well-being and positive performance. 35 See Sheldon et al., supra note 33, at 327. 36 See, e.g., Tim Kasser & Richard M. Ryan, A Dark Side of the American Dream: Correlates of Financial Success as a Central Life Aspiration, 65 J. PERSONALITY & SOC. PSYCHOL. 410, 420–21 (1993); Ryan & Deci, supra note 32, at 660.
2015] WHAT MAKES LAWYERS HAPPY? 565
guilt, fear, or pressure, or chosen to please or impress others).37 Research has established that intrinsic values and internal motivations are more predictive of well-being than their extrinsic and external counterparts.38 Another important construct of SDT is the effect of supportive (versus controlling) supervisors, teachers, or mentors. Research has shown that providing autonomy support39 to subordinates enhances their ability to perform maximally, fulfill their psychological needs, and experience well-being.40 The current study employs measures of all of these well-validated constructs.41
III. FOUNDATIONAL STUDIES OF LAW STUDENTS We initiated our investigation of the developing psychodynamics of lawyers with two published studies of law students.42 Both studies employed longitudinal designs to reliably investigate hypothesized changes during law school in student motivations, values, need satisfaction, and emotional health. If detrimental changes in adjustment were occurring during this foundational phase of professional formation, those changes could predispose graduates to emotional and behavioral problems in later law practice. Further, if data demonstrated likely causes for any negative changes, ongoing problems could be directly addressed and perhaps prevented by law teachers and deans. We studied two very diverse law schools in two different regions of the United States. The specific findings and the patterns within the
37 See Deci & Ryan, supra note 30, at 239–43; Sheldon & Krieger, supra note 5, at 263–64. 38 See Deci & Ryan, supra note 30; Sheldon & Krieger, supra note 5, at 265, 267–70; Kennon M. Sheldon et al., The Independent Effects of Goal Contents and Motives on Well-Being: It’s Both What You Pursue and Why You Pursue It, 30 PERSONALITY & SOC. PSYCHOL. BULL. 475 (2004); Sheldon & Krieger, Understanding Negative Effects, supra note 9, at 888. 39 Autonomy support is generally experienced when a supervisor or teacher conveys respect rather than control to a subordinate or student, by expressing understanding of the preferences of the other and providing her with choices. See infra Part VI.E. 40 Deci et al., Self-Determination in a Work Organization, 74 J. APPLIED PSYCHOL. 580, 589 (1989); Deci & Ryan, supra note 30, at 233–35; see also Sheldon & Krieger, Understanding Negative Effects, supra note 9, at 883–86. 41 Attorney Survey, supra note 25; accord infra notes 75–78. 42 Sheldon & Krieger, supra note 5; Sheldon & Krieger, Understanding Negative Effects, supra note 9. There were, of course, earlier studies documenting more straightforward negative changes in students, particularly anxiety and depression. See, e.g., Dammeyer & Nunez, supra note 5, at 56; Shanfield & Benjamin, supra note 17, at 66. There is also a recent prominent study that supports and further elucidates reasons for the precise negative changes in law students found in our studies. MERTZ, supra note 14. The Mertz study employed an entirely different design and methodology from our studies, and thus adds substantial confidence to our findings and conclusions. Id.; see also Krieger, Human Nature, supra note 5, at 267–70, 296–308 (discussing the impact of the Mertz findings in the context of the Sheldon/Krieger findings and offering strategies to mitigate the negative phenomena revealed by these studies).
566 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554
data are important and foundational for the current study. Those findings confirmed earlier reports of increasing anxiety and depression among students while in law school.43 More importantly, they pointed to reasons for the negative well-being shifts, and thus suggested educational strategies to prevent ongoing problems among students both before and after graduation. They also predicted many of the findings of the current attorney study, providing confidence in the results reported here. The first law school study44 demonstrated the following changes occurring in students after they began law school: marked increases in depression, negative mood, and physical symptoms, with corresponding decreases in positive affect and life satisfaction;45 shifts from helping and community-oriented values to extrinsic, rewards-based values in the first year;46 similar shifts in motivation for becoming lawyers, from salutary internal purposes (for interest, enjoyment, and meaning) to more superficial and external reasons (such as for financial rewards, recognition, or to impress or please others);47 and decreases in values of all kinds after the first year, suggesting generalized demoralization or loss of personal purpose.48 As discussed above, each of these shifts would predict decreased well-being, and that result was apparent in the data.49 As expected, the data also showed that students beginning law school with the most internal motivations and intrinsic values earned higher grades,50 but we also found that those students then shifted to more external (money-oriented) job preferences.51 Thus, the concerning findings extended beyond confirming decreasing student wellness; it also appeared that success in law school (measured by grades) could exacerbate the longer-term negative ef
43 For a summary of earlier findings of anxiety and depression in law student populations, see Dammeyer & Nunez, supra note 5. 44 Sheldon & Krieger, supra note 5. 45 Id. at 270–71 & tbl.1. 46 Id. at 272 tbl.3. 47 Id. 48 Id. at 273. This specific pattern of changes has been reported among students at Harvard Law School. See Note, Making Docile Lawyers: An Essay on the Pacification of Law Students, 111 HARV. L. REV. 2027, 2027 (1998). Neither of our subject schools were Ivy League/ elite schools, so this pattern of apparent demoralization may generalize to many law schools. 49 The study design did not permit firm conclusions about causation, but the consistency of the data certainly suggested this conclusion. See Sheldon & Krieger, supra note 5, at 273. As discussed immediately below, our second study employed additional measures and methods and did more confidently establish causation between related psychosocial factors, well-being, and student performance. 50 Id. at 274–75. 51 Id.
2015] WHAT MAKES LAWYERS HAPPY? 567
fects of the law school experience. More successful students changed career goals to prefer more extrinsically oriented jobs than when they began law school, and thus would be predicted to experience diminished satisfaction and well-being.52 The second study53 further investigated the mechanisms by which the law school experience generated these negative effects on students in these contrasting schools—one with a traditional scholarly focus and the other more focused on quality teaching and practical skills for students.54 We included additional methods and measures to address more subtle and potentially more telling variables—the level of autonomy support that students experienced from their faculties and the level of satisfaction of the students’ needs for autonomy, competence, and relatedness to others.55 This study again confirmed broad negative effects occurring during the three years of law school, including increasing student distress and decreasing internal motivation for legal work.56 The negative effects were most pronounced at the more traditional school.57 In addition, the added measures did reveal important new insights. First, the data demonstrated that all negative outcomes resulted from decreases in satisfaction of the fundamental needs for autonomy, competence, and relatedness to others after students entered law school.58 Of greater practical value for educators, the single factor of autonomy support that students received from their faculties accounted for all of the differences between the two schools in student need satisfaction, and hence
52 This specific pattern has also been described in earlier articles, although not supported with empirical data as here. See Robert Granfield & Thomas Koenig, Learning Collective Eminence: Harvard Law School and the Social Production of Elite Lawyers, 33 SOC. Q. 503, 517–18 (1992); see also Note, supra note 48, at 2040–42 (describing Harvard Law students’ tendency to shift from public interest to corporate law preferences after beginning law school). These findings and predictions were also supported in the current study, finding that lawyers with higher law school grades had chosen more affluent, externally motivated career tracks and were less happy than lawyers with lower grades and income. See infra Part VI.A. 53 Sheldon & Krieger, Understanding Negative Effects, supra note 9. 54 Id. at 886. 55 See supra notes 33–41 and accompanying text (regarding these measures); see also infra note 58 and accompanying text (regarding the significance of employing Structural Equation Modeling). 56 Sheldon & Krieger, Understanding Negative Effects, supra note 9, at 889. 57 Id. at 890. 58 Id. at 893–94. Importantly, this longitudinal study employed Structural Equation Modeling, and the data supported confident conclusions about causation. See id. at 891–93. The consistent symmetry of findings in the current cross-sectional study with those in this previous longitudinal study provides an additional source of confidence in the conclusions we draw from the current attorney data.
568 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554
in all of the other measured outcomes—well-being, career motivation, and academic outcomes (grades and bar exam performance).59 In other words, because of the more autonomy-supportive educational environment at the less traditional law school,60 students there fared broadly better, experiencing greater well-being, more internal motivation, and higher performance than the students at the other school. Notably, this institution had a far lower standing than the other in the hierarchy of law schools (as ranked by U.S. News & World Report),61 suggesting that law school reputation or standing may not relate, or may even relate inversely, to a variety of important student outcomes.62 The American Bar Foundation sponsored a third recent study of the law school experience, which is also important as context for the current attorney research. Professor Elizabeth Mertz63 conducted a linguistic analysis of the initial classroom training of new law students at eight diverse law schools.64 Her findings include a number of effects on law students that represent a fundamental undermining of basic personality structures, much as we found using entirely different methodology.65 Mertz observed, for example, that basic law school training changes student values;66 “unmoor[s] . . . the self”;67 marginalizes fairness, justice, morality, emotional life, and caring for others;68 and exclusively emphasizes competitive processes to the extent that they become the only goal.69 The net result is erosion of the very ability to make an ethical decision.70 Given the similar (and concerning) findings coming from this study and our previous research—studies
59 Id. at 890. 60 We could not empirically determine the factors responsible for the difference in autonomy support, but we reasoned that students might well perceive greater support from the emphases on law practice training (“skills” and clinics) and on faculty teaching expertise at this school, compared to the greater legal theory and research orientation of the second law school studied. Id. at 894–95. 61 U.S. NEWS & WORLD REPORT: AMERICA’S BEST GRADUATE SCHOOLS 60 (2006 ed.). 62 Consistent with these results, the data from the current attorney sample indicated only a negligible association of law school rank with well-being, despite moderately predicting increased income. See infra Part VII.E. 63 Professor Mertz is the John and Rylla Bosshard Professor of Law at the University of Wisconsin Law School. 64 MERTZ, supra note 14. 65 Id. at 10–11. 66 Id. at 1 (quoting SHIRLEY BRICE HEATH, WAYS WITH WORDS 367–68 (1983)). 67 Id. at 137. This study lends support to the generalized personal alienation reported at Harvard Law School. See Note, supra note 48, at 2034, 2038–40, 2044. 68 MERTZ, supra note 14, at 1, 6, 10, 95, 100–01, 120. 69 Id. at 77, 82–83, 95, 100–01, 126–27. 70 Id. at 132.
2015] WHAT MAKES LAWYERS HAPPY? 569
performed with entirely different empirical methodologies at different sets of diverse law schools—the findings provide mutual support and increase confidence that such results generalize to many, and perhaps most, law schools across the country.
IV. THE CURRENT STUDY
A. Measures As previously stated, our primary measure of happiness was SWB, the aggregate result of the mood and life satisfaction components. We supplemented the SWB measures with the depression scale from the Brief Symptom Inventory.71 That scale provided a second, inverse view of well-being, and a direct measure of depression—a matter of concern in the legal profession as previously discussed.72 This scale has been previously published in studies of law students and lawyers.73 In addition, given reports of substance abuse among lawyers, we inquired about the frequency and quantity of alcohol use. We expected this measure to provide another inverse indicator of wellbeing, and, as reported below, this was true with some limitations.74 We assessed likely predictors of well-being, including need satisfaction,75 values,76 motivations,77 and perceived autonomy support at work,78 using the same validated instruments used in our law student studies and previous SDT research.79 We also asked subjects about
71 Leonard R. Derogatis & Nick Melisaratos, The Brief Symptom Inventory: An Introductory Report, 13 PSYCHOL. MED. 595, 603 tbl.6 (1983); see also Sheldon & Krieger, Understanding Negative Effects, supra note 9, at 888 (applying the same measures to law students). 72 Eaton et al., supra note 6. 73 E.g., Beck et al., supra note 2, at 13 (citing LEONARD R. DEROGATIS & PHILLIP M. SPENCER, THE BRIEF SYMPTOM INVENTORY (BSI), ADMINISTRATION, SCORING & PROCEDURES MANUAL § 1 (1982)); Sheldon & Krieger, Understanding Negative Effects, supra note 9, at 888. 74 See infra Part VII.A. 75 See Sheldon et al., supra note 33, at 335–36; see also Sheldon & Krieger, Understanding Negative Effects, supra note 9, at 888. 76 See Tim Kasser & Richard M. Ryan, Further Examining the American Dream: Differential Correlates of Intrinsic and Extrinsic Goals, 22 PERSONALITY & SOC. PSYCH. BULL. 281, 284–86 (1996) (Aspirations Index); see also Sheldon & Krieger, supra note 5, at 265, 267–70. 77 See Kennon M. Sheldon & Andrew J. Elliot, Goal Striving, Need Satisfaction, and Longitudinal Well-Being: The Self-Concordance Model, 76 J. PERSONALITY & SOC. PSYCHOL. 482, 492–93 (1999); see also Sheldon & Krieger, supra note 5, at 265, 267–70. 78 We modified the Learning Climate Questionnaire for the work environment. See Aaron E. Black & Edward L. Deci, The Effects of Instructors’ Autonomy Support and Students’ Autonomous Motivation on Learning Organic Chemistry: A Self-Determination Theory Perspective, 84 SCI. EDUC. 740, 751–55 (2000); see also Sheldon & Krieger, Understanding Negative Effects, supra note 9, at 888. 79 When necessary for clarity and applicability to practicing lawyers and judges, we altered the wording from our law student instruments slightly. For example, a typical item in the auton
570 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554
previous law school experiences (name of school attended, class rank, law journal membership, and amount of debt upon graduation), current working circumstances (office setting, subject area of law practice, hours worked, billable hours required, position if in a private firm, and earnings), personal life choices likely to impact well-being (relationship status, children, exercise, vacations, religious or spiritual practices), and typical demographic information (gender, race and ethnicity, age, and number of years out of law school). Data were analyzed to determine which factors predicted wellbeing and the extent of their apparent impacts.80 We particularly wanted to compare the predictive power of the different categories of subjective and objective factors included in the study, as such information could assist law students and lawyers in making personal life and career decisions. Since the instrument included questions with different response metrics (i.e., dollars for income and debt, percentile for class rank, and level of agreement on Likert scales for psychological measures), we calculated results in terms of standardized Pearson correlation coefficients.81 This standardization permits meaningful comparison of factors expressed in different metrics. Thus, each variable measured was analyzed to determine if it related significantly and substantially82 to well-being, and we report standardized correlations (“r” factors) to indicate how strongly each variable predicts increased or decreased attorney well-being.
B. The Bar Member Sample With essential assistance of bar leaders and Lawyer Assistance Program directors, we were able to sample members of four state bar associations in the United States. The states represent four geographically diverse regions of the country, excluding the Pacific and Mountain West regions. One state is predominantly rural but includes a few large cities, one state is very populous with many major urban centers,
omy support measure for students read: “The faculty and administration listen to how I would like to do things.” The analogous item in the current survey read: “The supervisors listen to how I would like to do things.” Attorney Survey, supra note 25. 80 See supra note 8. 81 See, e.g., TABACHNICK & FIDELL, supra note 18, at 54. 82 Statistical significance is further discussed infra note 100. With such large sample sizes and statistical power, very small results can attain statistical significance but be essentially meaningless. See TABACHNICK & FIDELL, supra note 18, at 54. “Statistical significance is not the same as practical significance.” DAVID S. MOORE & GEORGE P. MCCABE, INTRODUCTION TO THE PRACTICE OF STATISTICS 425 (5th ed. 2006) (emphasis omitted). We therefore focus on strength of correlations, or “effect sizes,” throughout the Article, rather than relying primarily on statistical significance.
2015] WHAT MAKES LAWYERS HAPPY? 571
and two states include a mix of urban and rural areas. The states are also very diverse economically, politically, ethnically, racially, and in their predominant religions. We therefore expected these states to provide a relatively representative view of attorneys and judges in this country. The number of bar members invited to participate in each state ranged from 11,000 to 20,000. Two bar associations from less populous states invited all of their members. The other two states generated random lists of 11,000 and 20,000 invitees respectively. Sampling differences resulted from preferences within the governing bodies of the four bar associations. We expected valid results despite the different approaches, because partial invitee lists were randomly generated and all resulting sample sizes were very large. Invitees were sent an email introducing the project, assuring confidentiality, and providing a link to the online survey. They were told that the survey would remain open for about fourteen days, and a reminder e-mail was sent towards the end of the open period. Of necessity, the instrument was lengthy, because we sought to investigate and compare many dimensions of attorney experiences. As an incentive to participate, all subjects were offered a continuing legal education (“CLE”) program at no cost. The content of the programs in the four states was similar; bar personnel in two states created programs, while the other two states used a video program created by Professor Krieger. The CLE programs were intended to assist participating lawyers by educating them about simple choices that could improve their level of adjustment and well-being. Subjects could access their program via a link that was provided only after completion of the survey, so that the CLE content could not bias responses to the survey. Subjects were not made aware of the purpose and focus of the CLE programs, again to avoid biasing the sample. The numbers of responding bar members and the response rates for the states, from least populous (where all members were invited) to most populous (where the described samples were invited), were 1,757 (13.0%), 2,692 (15.8%), 1,606 (14.6%), and 1,750 (8.8%). The aggregate total sample was N = 7,805, with an overall response rate of 12.7%. One state had a substantially lower response rate (8.8%) than the others (13.0% to 15.8%). The data collection in that state followed the others by several months, and the timing (for the CLE reporting cycle) may have been less ideal. Bar officials in that state also expressed concern early in the process about survey fatigue in the membership. It is unclear if these or other factors impacted the re
572 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554
sponse; nonetheless, almost 2,000 subjects participated in this state, providing a substantial sample. Of the 7,805 participants who responded to the survey, we established a working sample of N = 6,226 subjects. This included all participants who provided complete well-being data and who indicated that they were currently working as lawyers, judges, or in related positions. This working sample was employed for most analyses; for analyses in which a different sample was used, it is noted and explained in the relevant section of the report. Given the length of the survey and the notorious workloads of this group of professionals, we felt the overall response to be relatively robust. Considering the typically busy schedule and heavy email traffic of practicing attorneys, it is likely that most who declined did so for lack of time or present need for the type of CLE credit offered. Since variations in workload pressure and the cyclical nature of CLE needs are common among lawyers, we expected participants to be representative of their overall bar membership. Comparisons of the mean age, gender distribution, and racial and ethnic distribution of the respondents from each state with their state’s entire bar membership supported the conclusion of representative samples. Each of the variances between the state samples and bar totals was small;83 the variances also showed consistent patterns. The percentage of women responding in each state was greater than the corresponding state bar membership by 2–6%, and the percentage of non-Caucasian respondents was 3–5% greater than the non-Caucasian membership by state.84 In one state, the age means for the sample and overall membership were virtually identical (46.4 and 46.6); in the other three states, the sample mean was 2–4 years greater than the mean of the entire membership. We may speculate that, given the length of the survey, slightly older lawyers tended to have the autonomy and time to complete both the survey and the CLE program. It may also be that women, minority, and older lawyers were slightly
83 The form and availability of membership data varied among the four states, introducing some imprecision in the variance calculations presented here. One state did not collect age data. The other states had age data only in ten- to twenty-year increments, requiring approximation by assigning the mean age in each range to those members. One state had race and ethnicity data for only forty-five percent of its bar members, creating doubt as to whether the large number of members declining to respond were disproportionately in one or more of the groups. 84 We compared only the Caucasian/non-Caucasian ratios because in every state Caucasians constituted the overwhelming majority (ninety to ninety-six percent) of bar members as a whole and of the subgroups of respondents, leaving very small subsamples (and hence relatively greater sampling error) if the minority groups were treated individually for this purpose.
2015] WHAT MAKES LAWYERS HAPPY? 573
more drawn to the general description of the survey and CLE program (relating to “attitudes and experiences of lawyers”) than their counterparts. Regardless, based on analyses of demographic differences presented below, the slight over-representation of older, female, and minority subjects may mean that the sample differs very slightly from the aggregate total membership in the four states, in terms of marginally greater internal motivation and well-being.85 Such differences would have no significant bearing on the findings and conclusions of the study. A further check of major variables also revealed few statistically significant differences between states, and those differences were slight, reaching significance only because of the large sample sizes involved.86 Ultimately, the subsamples and overall sample provided substantial confidence that the data collected would generalize to lawyers in the United States. The samples were large, and each tracked the makeup of its state membership; the data showed negligible to nil differences between demographic groups on major variables; and the states participating were very diverse, as previously described. As reported throughout the findings, the consistency of patterns in the data ultimately adds confidence in the results.
V. HYPOTHESES The breadth and depth of the instrument permitted investigation of a number of primary and secondary hypotheses. The most fundamental inquiry in the study focuses on an expected substantial difference in the correlations with lawyer well-being of selected internal and external factors. Internal factors of interest were the psychosocial factors that previous SDT research would predict to most strongly impact well-being. External factors of interest were those phenomena that are exceptionally important, and generate great concern, for many law students and lawyers—law school grade performance, law review participation, law school debt, and attorney income. We describe five related hypotheses and report the relevant findings below. Secondary hypotheses and findings then address other categories of variables that we thought likely to impact well-being to a lesser extent—demographics and choices or accomplishments related to work and personal life.
85 Infra Part VII.D (indicating these trends, but little overall impact of demographic differences). 86 Again, it is not unusual in large samples for results to be statistically significant but realistically meaningless. Supra note 82 and accompanying text.
574 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554
(1) Our first hypothesis was that objective factors that often dominate the attention of law students and lawyers (and legal employers and teachers as well)—law school grade performance, law review membership, law school debt, and income after graduation—would only modestly predict attorney well-being and would therefore provide a contrast when compared to the expected stronger associations with well-being of the internal factors included in the study. This hypothesized contrast was provocative because, while research in general populations has shown external factors such as rewards to be quite secondary predictors of happiness,87 law students and lawyers appear to place great emphasis on them. If the correlations with wellbeing of these external factors were strong, or if the hypothesized contrast with the internal factors did not manifest in the data, it would provide evidence that lawyers are indeed different from other people regarding the sources of their well-being. If the data did show this contrast, it would suggest that the external factors are simply “overdone” in the legal community and are not as important as typically thought—challenging core assumptions that are important in their own right because they generate so much stress in law schools and law firms. (2) Our second hypothesis was that the frequency of experiences of autonomy (which includes authenticity), competence, and relatedness to other people would very strongly predict lawyer well-being. Any such findings could be particularly important, because lawyers may be specifically inhibited from satisfying these needs by training in legal analysis,88 habituation to adversarial tactics,89 demands to adopt imposed client goals and values, personal conflict on many levels,90 the need to prevail in zero-sum proceedings against other aggressive lawyers, billable hour requirements and other controlling supervision methods, and perhaps other concerns particular to the practice of law.91
87 See, e.g., Lyubomirsky, Happier than Others, supra note 8, at 240 (observing that objective circumstances, demographics, and life events are weak predictors of well-being and that wealth typically shows “remarkably small associations” with happiness); David G. Myers, The Funds, Friends, and Faith of Happy People, 55 AM. PSYCHOLOGIST 56, 59–60 (2000). 88 For a thorough report on the effects of traditional law school training, see MERTZ, supra note 14. See also Krieger, Human Nature, supra note 5, at 267–70 (discussing the consistency of the Mertz findings with other law student research). 89 For further discussion of the competitive and adversarial paradigm in legal education, see MERTZ, supra note 14, at 4, 6; Krieger, Human Nature, supra note 5, at 265–66. 90 See, e.g., MARY ANN GLENDON, A NATION UNDER LAWYERS: HOW THE CRISIS IN THE LEGAL PROFESSION IS TRANSFORMING AMERICAN SOCIETY 17–108 (1994). 91 For a discussion of many of these factors in the context of legal education, see Lawrence
2015] WHAT MAKES LAWYERS HAPPY? 575
(3) Third, we hypothesized that the extent to which subjects’ motivation for their work was internal (for interest, enjoyment, and meaning), rather than external (for money, status, or prestige, or imposed by others), would also strongly predict well-being. This finding would also be concerning in this career group, because internal motivation is experienced as autonomous, originating within one’s self rather than externally, and law school may tend to marginalize internal instincts and responses.92 As a corollary, we expected more external motivation would manifest a “payoff” in greater earnings, but would nonetheless predict decreased happiness compared to subjects with more internal motivation. This would clearly be important for lawyers and law students, given the emphasis typically placed on the external factors previously discussed. (4) Fourth, we expected lawyers who more strongly endorsed intrinsic values (for growth, intimacy, community, and altruism) to be happier than those who more strongly endorsed extrinsic values (for affluence, status, fame, and appearance). This again would be concerning in light of data showing erosion of healthy values after students enter law school.93 We refined this hypothesis after administering the survey in two states. The traditional values measure asks subjects to report their beliefs about what is important in their lives. We conceived a new measure of action taken to give effect to specific values, which we thought would predict well-being more accurately than measuring only belief.94 We therefore administered to subjects in the two remaining states both measures—addressing endorsement of different values and addressing action in daily life directed towards each value.95 We hypothesized that both measures would indicate greater
S. Krieger, What We’re Not Telling Law Students—and Lawyers—That They Really Need to Know: Some Thoughts-in-Action Toward Revitalizing the Profession from Its Roots, 13 J.L. & HEALTH 1, 18–20 (1998). 92 See MERTZ, supra note 14, at 98–99; see also Krieger, supra note 91, at 18–20, 26–27 (discussing the need for conscience and instincts for health and well-being). 93 Sheldon & Krieger, supra note 5, at 279–80. 94 Values likely produce effects on well-being primarily because actions guided by different values tend to fulfill (or not fulfill) basic needs. See Christopher P. Niemiec et al., The Path Taken: Consequences of Attaining Intrinsic and Extrinsic Aspirations in Post-College Life, 43 J. RES. PERSONALITY 291, 292 (2009). People may also overstate their positive aspirations more easily than their actual behaviors. 95 For example, with regard to “gaining prestige, influence, or power,” in the traditional measure subjects were asked to “indicat[e] how important it is to you that the goal be attained in the future,” and in the new measure they were asked to “indicat[e] how much you actually work on that goal in your life.” Attorney Survey, supra note 25.
576 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554
well-being benefits from intrinsic valuing, and also that actions would better predict well-being than would endorsements alone. If the data supported both hypotheses, the findings would confirm the importance of intrinsic versus extrinsic values for lawyers, and would suggest a new and perhaps more useful way to approach the values/wellbeing relationship that could be applied in research more generally. (5) Our fifth primary hypothesis was that attorneys who received autonomy-supportive—as opposed to controlling—supervision would thrive to a substantially greater extent than others.96 This finding would have overarching importance for various groups. First, it would demonstrate to students and lawyers seeking happiness in their work the importance of supportive mentoring and supervision. Second, since teachers and supervisors can be trained to provide autonomy support to others,97 it would provide a constructive direction for educators and employers seeking to enhance the morale and resulting performance of their charges.
VI. PRIMARY FINDINGS A. Grades, Law Review, and Money Issues 1. Law School Grades Grade performance is likely the single greatest concern of law students as a group.98 We asked subjects to provide their law school class rank (which is based on grade performance) rather than measuring grades directly, since law schools use many different grading scales that would unduly complicate the questions and undermine confidence in the data. The correlation of final law school class rank with current SWB of our bar members (N = 4,65099) was r = .12, (p < .01100). This was in the direction of the modest correlation101 we 96 See Sheldon & Krieger, Understanding Negative Effects, supra note 9, at 884, 894. 97 See generally Johnmarshall Reeve et al., Enhancing Students’ Engagement by Increasing Teachers’ Autonomy Support, 28 MOTIVATION & EMOTION 147, 150 (2004). 98 It is well accepted that grade competition in law schools is intense and generates substantial stress on many students. See, e.g., DAICOFF, supra note 2, at 143; Barbara Glesner Fines, Competition and the Curve, 65 UMKC L. REV. 879, 901–02 (1997); Glesner, supra note 5, at 657–58; Krieger, Human Nature, supra note 5, at 277; Sheldon & Krieger, supra note 5, at 276; Note, supra note 48, at 2033–37. 99 A number of subjects did not respond to this question, perhaps because they did not recall their rank or because their school did not compute or announce rankings. This resulted in a reduced, but still very large, sample size. 100 “P” values indicate the probability that a reported event or relationship occurred randomly or by chance. Findings are generally considered statistically significant when the likelihood of chance occurrence is less than one in twenty (p < .05). MOORE & MCCABE, supra note 82, at 405–07, 424–25. The large sample sizes in this study enhance the ability to rule out random 2015] WHAT MAKES LAWYERS HAPPY? 577 predicted, but weaker than expected considering the overarching importance attributed to class rank in law schools.102 In light of the preference of many selective employers for applicants with high grades, we investigated the relationship of class rank with attorney income. As expected, there was a positive correlation (r = .20, p < .001) of law school grade performance with earnings after graduation, a somewhat stronger relationship than the small, and more important, effect size103 of class rank for lawyer well-being. 2. Law Journal Membership The survey asked subjects to indicate if they had been members of a law review or law journal. Law review membership is a second primary focus for many law students, as it is considered to connote excellence as a student and potential lawyer. Students can become discouraged when not achieving this recognition.104 Because achieving a journal position is an external factor related to high grade performance, we expected journal membership to also modestly predict wellbeing. The data, however, were surprising, yielding a zero correlation (r = .00) based on statistically identical mean well-being (4.862 versus 4.863) of subjects who had and had not participated on a journal (N = 1,656 and 4,570 respectively). Much like class rank, there was a modest relationship (r = .15, p < .001) between journal membership and later earnings, a result certainly expected given the elite status and hiring preferences afforded journal members by most selective employers. As expected, higher grades were also associated with journal membership (r = .32, p < .001). When regressed with class rank, the relationship of journal membership to income dropped to .09, showing journal work to be an independent—but quite weak—posi events, so that very small correlations in the range of r = .03 are statistically significant (p < .05), correlations of r = .05 are highly significant (p < .01), and correlations of r = .06 are very highly significant (p < .001). 101 In studies with typically smaller sample sizes, correlations in this range might not reach statistical significance. As a rough guide, the relative strength (and practical value) of correlations of different sizes might generally be considered as follows: < .05, negligible; .05–.10, slight; .11–.20, small; .21–.30, moderate; .31–.40, substantial; .41–.50, strong; > .50, very strong. These are quite inexact ranges intended only to provide a sense of meaning to reported correlations. 102 To avoid biasing this surprisingly weak correlation, we also calculated the class rank to well-being relationship employing the largest potential sample (N = 5,330), including people not working in the law (and therefore whose low grades may have resulted in poor employment prospects). There was a negligible difference in the direction we predicted, with the correlation of class rank to well-being increasing to .13, still a very small correlation with well-being. 103 Recall that effect size does not presume a causal relationship. Supra note 18. 104 See Note, supra note 48, at 2033–37 (chronicling the emotional distress attending disappointing grades and law review decisions among Harvard law students).
578 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554
tive factor, even for income. Three important, though very small, inverse correlations with journal work appeared in the data. Compared to other subjects, journal participants reported lower internal motivation for their current law job (r = −.06, p < .01), suggesting that they chose jobs for income, status, or reasons other than interest and passion for the work. They also reported less autonomy need satisfaction (r = −.05, p < .01). These findings likely explain the absolute lack of a well-being benefit, despite the increased income and prestige associated with the law journal honor.105 A further note of interest appeared in the data: these particularly successful law students experienced no greater competence in law practice than non-journal members (r = −.01, inverse but not significant). 3. Law School Debt and Income After Graduation Previous studies confirm that financial affluence has a positive effect on well-being in general populations.106 This effect is generally modest, particularly among subjects with sufficient earnings to provide for basic life needs.107 Many lawyers and law students, like other people, are concerned about their income level, and competition for grades that will assure well-paying jobs appears to exert great stress on law students. Current law students also often incur $125,000 or more in educational debt,108 which increases their finance-related concerns.109 Many may decide to forego preferred service work because of their high debt loads,110 a particular concern because the most 105 These data are consistent with previous findings that high-performing law students tend to shift towards less internally motivated job preferences, and they appear to confirm the conclusion that academic honors could undermine future well-being if lawyers then choose higher pay rather than interest and meaning in their work. See supra note 48 and accompanying text. The data suggest an undermining effect on well-being and performance, wherein motivation for rewards displaces healthy autonomous motivation. See infra notes 208–09 and accompanying text. 106 See, e.g., Ed Diener et al., The Relationship Between Income and Subjective Well-Being: Relative or Absolute?, 28 SOC. INDICATORS RES. 195, 215–17 (1993); see also FREY & STUTZER, supra note 8, at 74, 75, 82; Myers, supra note 87, at 58–61. 107 See Myers, supra note 87, at 61. 108 Debra Cassens Weiss, Average Debt of Private Law School Grads Is $125K; It’s Highest at These Five Schools, ABAJOURNAL (Mar. 28, 2012, 10:29 AM), http://www.abajournal.com/ news/article/average_debt_load_of_private_law_grads_is_125k_these_five_schools_lead_to_m/. 109 Everyone would prefer to have little or no debt, but it is not a given that debt must generate great stress. A previous study showed that, despite substantially higher debt, students at a law school with a more supportive faculty were significantly happier than those at a contrasting school where much less debt was incurred. Sheldon & Krieger, Understanding Negative Effects, supra note 9, at 893. 110 See Gita Z. Wilder, Law School Debt and Urban Law Schools, 36 SW. U. L. REV. 509, 527 (2007). However, empirical analysis suggests that debt is not responsible, to the extent articulated, for students foregoing service work. Christa McGill, Educational Debt and Law Stu 2015] WHAT MAKES LAWYERS HAPPY? 579 prominent study of human needs to date found an inverse correlation between well-being and the emphasis that subjects placed on high earnings as a source of satisfaction.111 Because income can provide comfort and reduce financial stress, we expected increasing income (and decreasing law school debt, as an inverse wealth factor) in the current sample to modestly predict wellbeing. The data were supportive, showing almost identical, small-tomoderate correlations with well-being for both factors (income, r = .192; debt, r = − .189; both round to r = .19; p < .001). Further analysis showed that the negative association of debt with well-being was stronger for younger lawyers. This would certainly be expected, because law school costs and incurred debt would be less for older lawyers, and those lawyers would also have higher incomes as a result of more years in practice. B. Psychological Need Satisfaction Previous research demonstrated the central importance of experiences of autonomy and authenticity, relatedness, and competence for the well-being and performance of law students.112 Data from our attorney subjects confirmed the central importance of all three needs for their well-being. Correlations were exceptionally strong: autonomy, r = .66; relatedness, r = .65; and competence, r = .63 (all p < .001). Confirming their importance for mental health, the needs also bore strong inverse correlations with depression (r = −.51 to −.63; all p < .001). As hypothesized then, the needs were far more predictive of well-being in our subjects than were the external factors under consideration, with relationships to well-being approximately five times stronger than that of class rank and 3.5 times stronger than that of income or school debt. C. Motivation Data from the sample fully supported our hypothesis that internal (self-determined or autonomous) reasons for choosing work—interest, enjoyment, or effectuating core values—would be another critical dent Failure to Enter Public Service Careers: Bringing Empirical Data to Bear, 31 LAW & SOC. INQUIRY 677, 678 (2006). The motivation for prestige and affluence, coupled with a dearth of public service positions, is more explanatory than the need to repay debt. See id. at 704; Granfield & Koenig, supra note 52, at 517–18. 111 Sheldon et al., supra note 33, at 331–33. Note that affluence itself was not found to relate negatively with well-being, but the fact that subjects attributed importance to affluence did. 112 Sheldon & Krieger, Understanding Negative Effects, supra note 9, at 884–85. 580 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 factor for attorney well-being. The association with well-being was very strong (r = .55, p < .001), with a confirming inverse relationship to attorney depression (r = −.31, p < .001). This finding is particularly important, because law students have been found to turn away from internally motivated careers, often in favor of more lucrative or prestigious positions, after beginning law school.113 Because these data indicate that well-being is substantially impaired when law graduates emphasize external over internal factors in their career choices, we sought to clarify the importance for well-being of competing internal and external factors that could often affect the decisions of lawyers seeking jobs. We investigated the occurrence in the sample of interest and perceived meaning in work, higher earnings, and higher grades (which would typically tend to generate more, and more lucrative, employment opportunities). When motivation, class rank, and income were entered in a simultaneous regression equation with well-being, the independent association of healthy (internal) motivation with well-being remained at its full correlation strength ( b = .55). By contrast, after regression, the external factors lost some of their already modest value for predicting attorney wellbeing (for income, b = .13; for class rank, b = .05). This analysis further supported the importance of choosing interest and meaning in work rather than higher income when lawyers are faced with that choice in career decisions.114 D. Values As previously explained, we surveyed all subjects with an established measure of intrinsic and extrinsic aspirations, and subjects in two states were also asked the extent to which they acted to achieve each value.115 All data were consistent with findings in other populations, with both measures showing greater well-being for lawyers with more intrinsic, rather than extrinsic, values. The respective correlations with well-being were: for intrinsic aspirations, r = .21; for extrin 113 See ROBERT V. STOVER & HOWARD S. ERLANGER, MAKING IT AND BREAKING IT: THE FATE OF PUBLIC INTEREST COMMITMENT DURING LAW SCHOOL 22 (1989); Granfield & Koenig, supra note 52, at 517–18; Sheldon & Krieger, supra note 5, at 275, 281–82. We also found that supportive teaching could mitigate some of this negative effect. Sheldon & Krieger, Understanding Negative Effects, supra note 9, at 894–95. 114 Secondary findings show that, with age and time in career, it is increasingly possible to realize the ideal situation of higher pay and more internally motivated work in the same job. Infra Part VII.J. 115 To avoid response bias, we did not ask subjects directly if they “acted on their values.” They first responded to the traditional aspirations index, and then to the new measure directed to actual behaviors. 2015] WHAT MAKES LAWYERS HAPPY? 581 sic aspirations, r = .09 (N = 6,216, both p < .001).116 The new, actionoriented measure provided consistent but stronger associations with well-being: for intrinsic actions, r = .30, and for extrinsic actions, r = .19 (N = 2,523, both p < .001). Thus, our hypotheses regarding the primacy for lawyer well-being of intrinsic over extrinsic values, and of actions over aspirations, were both supported by the data. As with motivation, we had particular concerns about attorney values and well-being based on our findings of two distinct types of negative changes in student values during law school.117 The results here confirmed our concerns about the values shifts that occurred in law school, since these data showed that values operate in practicing attorneys as in other populations. Hence all of the shifts measured in law students would predict their decreased well-being both during and after law school. 1. Values and Professionalism Beyond the importance of values for attorney well-being, values have a crucial significance for professionalism and ethical behavior. Intrinsic values include self-understanding and improvement, positive interpersonal relationships, helping others, and building community;118 such values would logically lead to introspection, honesty, cooperation, respect, and altruistic behavior. These values, then, would tend to promote integrity, candor, dedication to a client or cause, and respectful interactions with clients, opposing parties, and others, thus elevating an attorney’s ethical and professional conduct. By contrast, none of the extrinsic values (achieving high earnings, status, appearance, or influence over others) appear to relate directly to professional or ethical behavior, and such values could actually lead to unethical or unprofessional conduct if perceived as helpful to attain the desired end. Previous articles have addressed this matter in some 116 Note that extrinsic valuing is not negative in itself, and indeed most people value affluence, influence, and recognition by others to some extent. When these values dominate and displace intrinsic values, however, negative effects occur. See Sheldon & Krieger, supra note 5, at 281–82 (finding that law students tended to lose both intrinsic and extrinsic valuing, raising particular concerns for lack of goal-driven behaviors of all kinds). 117 Our first longitudinal study found both that student values shifted towards less healthy extrinsic pursuits rather quickly after entering law school and that all values then diminished later in law school (a general “dulling” effect). See id. at 273, 279. Elizabeth Mertz notes specifically that one goal of law teaching is to change student values. See MERTZ, supra note 14, at 1. 118 See Deci & Ryan, supra note 30, at 244; Sheldon & Krieger, supra note 5, at 263–64; see also Sheldon et al., supra note 33, at 326. 582 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 detail.119 The scope of the current study did not permit direct testing of these propositions, but given the very definition of the intrinsic values, it would be surprising if they did not tend to promote ethical and professional behavior. Thus, if the findings here—that intrinsic values are positive well-being factors for lawyers—resulted in broader adoption of such values, it could have positive implications for attorney professionalism as well. E. Autonomy Support Autonomy support is an important construct in any relationship in which one person has less authority or power than the other (teacher-student, employer-employee, parent-child, mentor-mentee). The components of perceived autonomy support include the extent to which the person in authority (1) acknowledges the perspectives or preferences of the other; (2) provides meaningful choices to the other; and (3) when asserting control rather than providing choices, explains to the other the reasons why that is necessary.120 Previous research has shown autonomy support to have global benefits for need satisfaction,121 internal motivation,122 and performance123 of law students. If the current study demonstrated similar results, autonomy support could prove to be the single most important consideration for lawyers, and for their teachers and employers interested in fostering well-being and maximal performance. Investigation of autonomy support was necessarily limited to those lawyers who reported having one or more supervisors (N = 4,101).124 The findings were robust among our bar members, with autonomy support very strongly correlating with subjective well-being (r = .44) and correlating inversely with depression (r = −.30; both p < .001). Thus, the data clearly supported all primary hypotheses, showing all of the internal factors to more strongly predict lawyer well-being than any of the external “grades and money” factors. Figure 1 shows the associations of the primary factors in this study with subjective well-being; the darker bars represent the internal factors. 119 See, e.g., Krieger, Inseparability, supra note 21, at 429–30; Krieger, Most Ethical People, supra note 21, at 180–81. 120 See Sheldon & Krieger, Understanding Negative Effects, supra note 9, at 894. 121 See id. at 889. 122 See id. at 891. 123 See id. at 894. 124 For this smaller, but still very large, sample, correlations greater than r = .04 are significant (p < .05), and correlations of .06 are highly significant (p < .01). 2015] WHAT MAKES LAWYERS HAPPY? 583 FIGURE 1. SUBJECTIVE WELL-BEING CORRELATES 1. Replicating the Path Model for Autonomy Support, Motivation, and Well-Being Our previous research on law students revealed the critical role of autonomy support in a sequential path affecting their needs, motivation, academic performance, and well-being. We again employed multiple regression analyses of the current data to investigate those relationships in our attorney subjects. These analyses replicated the previous findings,125 indicating that autonomy support increased satisfaction of the lawyers’ three basic needs and thereby also increased well-being and internal motivation for their work. A path model representing these relationships for the current study is shown in Figure 2;126 all paths are substantial and highly significant (p < .001). These findings are particularly important for concerns about attorney wellness, satisfaction, retention, and performance,127 because supervisors, teachers, and mentors can be taught to provide autonomy support to 125 See Sheldon & Krieger, Understanding Negative Effects, supra note 9, at 892. Obviously, with this sample of working attorneys we could not seek to replicate the student finding regarding the third principle outcome—improved grades and testing performance. Also, because the path model was generated to test the student findings, it presents motivation as a cooutcome (with well-being) of need satisfaction, as was done in the previous study report. However, motivation is also a cause of well-being, as indicated in the current findings, and after regressions controlling for all the factors in the path model below, the independent correlation between well-being and internal motivation remained very strong (r = .45). 126 Because the path model relies on regression analyses, the coefficients show only unique effects, and hence are somewhat smaller than the standardized coefficients reported above. 127 Although performance was not measured in this study, we fully expect enhanced performance to result from autonomy support based on the law student findings, and also on the finding here that attorney motivation improved with perceived autonomy support. See infra notes 202–11 and accompanying text. 584 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 others.128 Seen from the contrary perspective, this also suggests that controlling supervisors who are not trained to be supportive will exert a number of avoidable negative effects on their employees and on organizational morale and efficiency. FIGURE 2. PATH MODEL RELATING AUTONOMY SUPPORT, NEED SATISFACTION, AND MOTIVATION/WELL-BEING OUTCOMES Supervisor Autonomy Support Autonomy Need Satisfaction Relatedness Need Satisfaction Competence Need Satisfaction SWB Self-Determined Job Motivation .39 .20 .32 .27 .32 .32 .23 .28 F. Brief Discussion of Primary Findings Findings broadly supported the five primary hypotheses, providing an empirically supported understanding of contrasting factors that predict attorney well-being. We particularly focused on the relative importance for well-being of subjective psychosocial factors compared to more objective, external factors typically of great concern to law students and lawyers (and to their teachers and employers as well). We expected that empirical results would, to some extent, contradict common assumptions about the importance of external factors such as earnings, debt, comparative grade performance, and honors or credentials. The data confirmed our hypotheses, revealing a pattern in which (1) the internal factors seen to erode in students during their initial law training were the precise factors most strongly predictive of 128 Johnmarshall Reeve, Autonomy Support as an Interpersonal Motivating Style: Is It Teachable?, 23 CONTEMP. EDUC. PSYCHOL. 312, 324–28 (1998); Reeve et al., supra note 97, at 159–61. For a fully developed article providing such training for law teachers, see Paula J. Manning, Understanding the Impact of Inadequate Feedback: A Means to Reduce Law Student Psychological Distress, Increase Motivation, and Improve Learning Outcomes, 43 CUMB. L. REV. 225, 245–57 (2012). See also ALFIE KOHN, PUNISHED BY REWARDS 186–87, 192–97 (1993) (instructing supervisors to increase internal (“authentic”) motivation and productivity through provision of understanding, perspective-taking, and choice to employees). 2015] WHAT MAKES LAWYERS HAPPY? 585 lawyer well-being and (2) the external factors emphasized in law school and by many legal employers were, at best, only modestly associated with lawyer well-being. This unfortunate pattern was somewhat stronger than we expected. One external factor of seemingly great importance to law students and legal employers—law review membership—had no measurable correlation with lawyer happiness and life satisfaction (r = .00); and income, the external factor most predictive of well-being (r = .19), was less predictive than the internal factor with the weakest association with well-being (intrinsic values, r = .30). Thus, the data established a distinct dichotomy of factors bearing on lawyer well-being, with correlations of external factors ranging from .00 to .19 (on a scale for which 1.0 is a perfect correlation) and correlations of internal factors from .30 to .66. In addition to clear implications for the universal search for life satisfaction and happiness, these findings have important implications for attorney ethics and professionalism and for the “bottom line” productivity and profits of legal employers. The most powerful predictors of well-being in these data—autonomy (r = .66), relatedness to others (r = .65), competence (r = .63), and internal motivation for work (r = .55)—are also sources of professional behavior and positive performance in lawyers; lawyers experiencing high well-being are also likely to produce more, remain longer, and raise the morale of others.129 VII. SECONDARY FINDINGS We analyzed many additional variables to gain further understanding of lawyer well-being and satisfaction. These included alcohol consumption, demographic differences (age, gender, race, and ethnicity), work variables (practice type, office setting, hours worked, position in law firm, and billable hours), family and personal choices that might contribute to life balance or “stress management” (relationship status, children, physical exercise or sports, vacations, and religious or spiritual practices), and the ranking of a lawyer’s law school. We investigated two supplementary issues that yielded concerning data: subjects’ perceptions of lawyers, judges, and the legal system, and the extent to which subjects’ early expectations for future income in their legal career were realized. Many results are interesting in their own right, and many confirm the primary findings regarding the overarching importance of the internal factors for well-being—particularly the 129 The relationships between well-being, professionalism, and performance are discussed more fully infra Part VIII.C. 586 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 three needs and internal motivation for work. A number of consistent patterns in the data also provide confidence in the study as a whole. Implications are discussed after presentation of the findings. A. Alcohol Consumption Abuse of alcohol by attorneys is a common concern.130 We viewed increasing alcohol use as a likely indicator of negative wellbeing and hypothesized that it would inversely correlate with wellbeing, as well as with the psychological variables most strongly associated with well-being. The survey instrument included established measures for frequency and quantity of consumption.131 Mean frequency of drinking reported in the sample was approximately once each week. Mean consumption on each occasion was 1.77 drinks.132 Frequency of drinking was, on first examination, marginally related to well-being (r =− .025, p = .051), while quantity consumed per occasion showed a small but more robust negative correlation across the sample (r = −.12, p < .001).133 However, because frequency correlated very strongly with quantity (r = .61, showing that those who drank more heavily also drank more often), we regressed both measures with well-being to determine independent effects. The result showed that frequency, independent of the influence of increasing quantity, was unrelated to (or even slightly positive for) well-being.134 Also as hypothesized, important psychological factors for well-being were inversely associated with quantity of drinking: intrinsic values orientation, r = −.13; internal 130 See Beck et al., supra note 2; Susan Daicoff, Lawyer, Know Thyself: A Review of Empirical Research on Attorney Attributes Bearing on Professionalism, 46 AM. U. L. REV. 1337, 1347, 1382 (1997); Eric Drogin, Alcoholism in the Legal Profession: Psychological and Legal Perspectives and Interventions, 15 LAW & PSYCHOL. REV. 117, 158 (1991). Lawyer Assistance Programs (“LAP”) were established in most states to address this problem. See Commission on Lawyer Assistance Programs, ABA, http://www.americanbar.org/groups/lawyer_assistance.html (last visited Mar. 1, 2015). LAP professionals provided critical assistance in the course of approval and administration of this survey by various bar associations, and they were instrumental in including measures for alcohol use in the instrument itself. 131 See Recommended Alcohol Questions, NAT’L INST. ON ALCOHOL ABUSE AND ALCOHOLISM, http://www.niaaa.nih.gov/research/guidelines-and-resources/recommended-alcoholquestions (last visited Mar. 1, 2015) [hereinafter Task Force]. 132 A detailed definition of an alcoholic “drink” is provided in the measure. See Task Force, supra note 131. Means from the analyses are restated here to express the actual number of drinks reported. 133 Quantity per occasion was similarly related to depressive symptoms (r = .10, p < .001). 134 The slight positive result after regression is not surprising, because frequent light drinking is often associated with salutary activities in leisure time, including shared meals or socializing with friends. 2015] WHAT MAKES LAWYERS HAPPY? 587 motivation for work, r = −.08; autonomy, relatedness, and competence need satisfaction, r = −.06, −.09, and −.09 respectively (all p < .001). Thus, our hypothesis regarding the inverse relationship between alcohol use and well-being was supported, but only for the quantity of drinking per occasion. Given these results, we report only the quantity measure in the remaining analyses of alcohol use.135 We then analyzed the level of alcohol use reported by various demographic groups within the sample. Between the genders, mean consumption for men was slightly greater than for women (M = 1.89, 1.56; p < .001);136 and among the racial and ethnic groups, Caucasian lawyers drank most and African Americans least (M = 1.80, 1.41 respectively; p < .001). Subjects earning more income drank slightly more (r = .04, p = .003). Lawyers in public service positions also drank less than private attorneys, particularly those in positions that typically provide the most income (M = 1.67 versus 1.90; p < .01).137 B. Attorney Well-Being in Contrasting Work Settings and Practice Types: Testing the Internal-External Factors Dichotomy The practice of law offers markedly different work settings, earnings, and lifestyle expectations. Would data indicate that some career paths were more likely than others to provide satisfaction and wellbeing to lawyers? If so, could we quantify specific factors impacting the well-being of lawyers in different settings or types of practice? The instrument asked subjects to identify both their work setting and the subject matter of their primary practice. We provided fifteen setting choices, including private firms (which ranged in size from solo practice to more than 750 lawyers), judicial chambers, and in-house work for public agencies, businesses, and other entities. We also included twenty-eight specific choices for type and subject matter of work, including criminal prosecution, public criminal defense, private 135 We also analyzed the product of consumption frequency and quantity. This expression, approximating total consumption of alcohol, was significant but slightly less predictive of wellbeing than the quantity measure alone. This was not surprising, because the product interaction of the two variables could occur in numerous ways, each having a different overall effect on wellbeing. For example (and using a comparable but simpler scale than that used in the survey): when multiplying the number of drinking occasions per week by the number of drinks per sitting, the product would be identical for two subjects differing greatly in their habits and likely their well-being—one drinking a single glass of wine with dinner each night for a week, and another drinking seven glasses of wine in a single sitting, one night per week. 136 This may represent actual equivalence, considering the greater body mass of men. 137 Detailed comparisons of this and other well-being factors among attorneys in different practice types are presented infra Part VII.C.3. 588 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 criminal defense, family law, corporate or commercial practice, intellectual property, and tax.138 In order to manage and analyze this multifaceted data in a comprehensible, meaningful way, we first organized respondents into two groups that might manifest the clearest contrasts (in terms of correlates of well-being) in their career paths. One group included subjects who had chosen jobs that tend to be highly sought after and most typically expected to produce high earnings (“prestige” positions).139 The contrasting group included the lawyers in public service positions typically perceived as providing more altruistic service but with modest income (“service” jobs). These groups of lawyers would likely differ in their motivation, values, and need satisfaction, and would certainly differ in external markers of success such as law school grade performance and current income. If well-being differed between these groups, it could provide a practical example of how SDT principles operate in actual law practice settings. Thus, the delineation of “service” and “prestige” groups was a theory-based approach both to organize much of the complex occupational data and potentially to view the applied effects of the dichotomy between internal and external contributors to well-being. We constituted the groups based on long experience both in law practice and working with law students. Choices were somewhat arbitrary but were confirmed by open inquiry to a listserv of approximately 1,000 clinical law teachers who also had substantial experience practicing law. The “service” attorney group included subjects in the following positions: public defender, criminal prosecutor, government agency, legal services to the indigent, and in-house counsel for nonprofit organizations. The typically lucrative “prestige” positions, all in private practice, included law firm settings of 100 or more lawyers;140 plaintiff’s tort/malpractice law; corporate, commercial, or transactional law; international business and commercial transactions; securi 138 Attorney Survey, supra note 25. 139 “Successful” law students and lawyers are often considered to be those who earn high grades and high income. Success and higher earnings are strongly identified together in the United States generally. See MYERS, supra note 19, at 31–34. We use prestige to describe this lawyer group, because other descriptive terms could have unintended negative connotations for these lawyers or the contrasting public service lawyers. We did not include other prestigious positions, such as judicial clerks, in the group, in order to maintain homogeneity in terms of law practice as commonly understood, client representation, and a tendency towards high earnings. 140 While we chose firms of more than 100 lawyers because of our sense that students and lawyers generally consider this to constitute a “large firm,” and consider large firm lawyers to be high earners, the data did show that firms of this size and larger do offer more pay than smaller firms. See infra Part VII.C.1.b. 2015] WHAT MAKES LAWYERS HAPPY? 589 ties or partnership law; and tax, estate planning, or patent and copyright. We also included a “judge” group, including judges and hearing officers, which we expected to distinguish itself based on a unique combination of both internal and external positive factors for well-being—high autonomy, internal work motivation, service values, and time for family and relationships, coupled with power, substantial income, and respect in the community. The fourth group, “other law practice,” included all other practicing lawyers, including practitioners in popular practice areas such as general practice, family law, private criminal defense, and many others not typically associated with either very high earnings or primary public service.141 The resulting subsamples by group were: “service,” N = 1,091; “prestige,” N = 1,434; “judges,” N = 141; and “other,” N = 2,852. Our hypotheses regarding these groups were organized around expected differences between internal/psychological factors and external factors such as income or status: (1) “Prestige” lawyers would report far higher income than “service” lawyers. (2) “Prestige” lawyers would have the highest mean law school class rank and greatest law review participation as students. (3) “Service” lawyers would report greater self-determined motivation for their work and more intrinsic values than “prestige” lawyers. (4) Because motivation and values are stronger factors for wellbeing than prestige or status, “service” lawyers would enjoy well-being equal to or greater than “prestige” lawyers. Our confidence in this hypothesis was somewhat compromised, because prestige positions could provide benefits for well-being in addition to high pay (such as work space, better furnishings, equipment, training, supervision, and secretarial and paralegal support). A greater confound might be that, if indeed higher grade achievers tended towards prestige positions, their achievements generally could reflect other positive attributes that would dispose them to well-being regardless of their position— such as health, energy, alertness, enthusiasm, or resilience. Nonetheless, if the “service” group did report more than nominally greater internal motivation, intrinsic values, or need satisfaction, Self-Determination Theory (and our primary findings reported above) would predict its members to be happier and more satisfied lawyers. 141 We excluded from this group those subjects identifying themselves as primarily law school teachers, bar administrators, mediators and arbitrators, and clerks or support staff for judges or lawyers to create a group of more typical lawyers engaged in client representation. 590 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 (5) As stated above, we expected the judge and hearing officer group to report the highest well-being. These subjects would likely have the best of both worlds, with substantial internal and external benefits for well-being. They also would not experience the stresses of client representation and vying against other parties in the adversarial process. Thus, the judges could constitute a group fundamentally different from the others, likely distinguishing themselves from the three lawyer groups in many ways. (6) Because the “other law practice” group was not created based on salient shared traits, we had no theory-based hypotheses regarding this group. However, we thought this group would experience less well-being than the “service” or “prestige” groups, because the “others” would likely have neither the financial benefits of the “prestige” lawyers nor the internal motivation and intrinsic values of the “service” group. 1. Findings The data broadly supported our hypotheses, yielding many results that conflict with common expectations about who among lawyers will enjoy the greatest well-being.142 Important findings included: (1) Judges were, indeed, clearly different from the three lawyer groups. The judges reported the greatest well-being on all measures— net affect, life satisfaction, and SWB (all p < .001).143 They also reported the greatest internal motivation and satisfaction of all three needs, combined with the second highest mean income—the predicted combination of internal and external well-being factors that likely produced their greater happiness. However, judges also reported a mean age almost ten years greater than any lawyer group, which could explain, at least in part, many of the apparent benefits noted.144 (2) As hypothesized, “prestige” lawyers had robustly higher income, law school grades, and law review participation than the “service” lawyers (all p < .001), but also had less internal motivation and 142 See AJD1, supra note 7, at 48 (including a similar finding about enhanced satisfaction of public service lawyers with regard to the work they do); see also Dinovitzer & Garth, supra note 28, at 22 (noting that survey “respondents working in state government are significantly more satisfied with their career choice”). 143 There are many statistically significant differences between these groups, but correlation strengths are typically small. Because of the complexity of numerous cross-comparisons of data among the groups, in this and other sections involving comparison of numerous subsamples (i.e., racial/ethnic and social support groups), correlation strengths are reported only for the more important analyses. 144 The various benefits of age for well-being are discussed infra Part VII.D.1. 2015] WHAT MAKES LAWYERS HAPPY? 591 intrinsic valuing (both p < .001). As we had hoped then, these groups provided a view of how internal verses external well-being factors play out in working professionals. As predicted, and of primary importance, “service” lawyers reported greater well-being than the more “elite,” highly paid “prestige” lawyers (r = .04, p < .05),145 despite substantially lower earnings (r = .33, p < .001). “Prestige” lawyers also reported more alcohol use (p < .001). A comparison of the components of subjective well-being adds detail to the picture. The “service” and “prestige” lawyers reported equal life satisfaction despite the greater affluence (presumably providing more expensive houses and cars, for example) of the latter group, suggesting that more desirable possessions are only marginally helpful, even for satisfaction with life. On the other hand, the “service” lawyers reported significantly higher day-to-day mood, likely from their sense of service and greater enjoyment and perceived meaning in their work. The rather striking net result was greater aggregate well-being for the lawyers in “service” positions. Another point of interest emerged in the data: although the “prestige” lawyers had substantially higher law school grades than any other group, they reported significantly lower satisfaction of the competence need (p < .01) than the group with the lowest grades and pay, the “service” lawyers. This suggests a core dissonance between “competence” as measured in law school (largely by grade performance) and a lawyer’s ability to feel competent in actual law practice.146 (3) The “other law practice” group did turn out to be the least happy group. These lawyers, as predicted, reported neither the high earnings of the “prestige” group (p < .001) nor the high service aspira 145 This well-being difference may also be slightly under-reported by the data, because “service” jobs are typically less competitive and easier to secure (as evidenced by the lowest mean class rank of the three attorney groups). Thus, some of the “service” jobs may be occupied by students who could not obtain other positions they preferred (i.e., this group likely includes a number of people not primarily motivated by service). 146 This difference in reported competence likely results from the fact that attorney competencies accrue from the development of practice skills, rather than from the more typical theoretical learning provided by most law schools, coupled with the fact that lawyers in public service positions are often provided case responsibilities sooner than those in “prestige” positions. The Carnegie Foundation’s report on legal education appears to recognize this gap between what law schools teach and what their students need to know in order to be competent. See WILLIAM M. SULLIVAN ET AL., EDUCATING LAWYERS: PREPARATION FOR THE PROFESSION OF LAW 173–76 (2007) (expressing concerns with a traditional over-emphasis on cognitive training, and noting that other fundamental lawyering skills require different methods of teaching and assessment); see also Harry T. Edwards, The Growing Disjunction Between Legal Education and the Legal Profession, 91 MICH. L. REV. 34, 34–36 (1992) (criticizing the increasingly academic and scholarly nature of legal education). 592 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 tions or internal motivations of the “service” group (both p < .001). However, despite their lower earnings, the “other” lawyers and the higher-earning “prestige” lawyers reported no difference in mood or affect, demonstrating that greater affluence has little effect on whether lawyers feel happy from day to day. As with the “prestige” lawyers, the “other” lawyers had higher class ranks and earnings than the “service” group (p < .01 and p < .001 respectively), but again reported less competence need satisfaction (p < .001). Ultimately, the “other” lawyers had significantly lower mood and less life satisfaction than the “service” lawyers (p < .001 and p < .01 respectively), resulting in lower overall well-being (p < .001). Taken together, these data continue to indicate the quite limited value of money, grades, and prestige for the well-being of professionals. They also call into question law school grades and honors147 as measures of competence and suggest that more attention be given to the well-being of those lawyers in the more typical practices who are neither highly paid nor in the public sector. The data should also provide some stress relief to law students and lawyers, and guidance to those trying to decide on a career focus. The competition and stress related to high earnings and high grades148—both zero-sum, limited resources—appear overdone. These data consistently indicate that a happy life as a lawyer is much less about grades, affluence, and prestige than about finding work that is interesting, engaging, personally meaningful, and focused on providing needed help to others. The data therefore also indicate that the tendency of law students and young lawyers to place prestige or financial concerns before their desires to “make a difference” or serve the good of others149 will undermine their ongoing happiness in life. This is a clear direction for increased education of law students and young lawyers. If a lawyer isn’t happy, “what is the point?”150 147 Recall that subjects with law review experience also reported no greater competence as attorneys. Supra note 146 and accompanying text. 148 See Glesner, supra note 5, at 637; Martin E.P. Seligman et al., Why Lawyers Are Unhappy, 23 CARDOZO L. REV. 33, 40, 47 (2001). See generally KRIEGER, HIDDEN SOURCES, supra note 16; Fines, supra note 98. 149 See generally ROBERT GRANFIELD, MAKING ELITE LAWYERS: VISIONS OF LAW AT HARVARD AND BEYOND (1992); STOVER & ERLANGER, supra note 113; Sheldon & Krieger, supra note 5; Note, supra note 48. 150 Gary A. Fenner, Robert M. Fenner & G. Michael Fenner, Three Perspectives of the Law: Advice to the Young Lawyer, 31 CREIGHTON L. REV. 665, 673 (1998). 2015] WHAT MAKES LAWYERS HAPPY? 593 Table 1 presents means for important variables within the four groups; the groups are listed from highest to lowest subjective wellbeing. For class rank, increasing means indicate lower class ranks. TABLE 1. VARIABLES BY PRACTICE GROUP N Age SWB Intrinsic Value Positive Affect/ Negative Affect Life Satisfaction Depression Judges 141 56 5.76 1.02 1.97 3.70 1.45 Service 1091 46 4.98 1.11 1.62 3.38 1.59 Prestige 1434 46 4.86 0.73 1.50 3.36 1.59 Other 2852 47 4.71 0.79 1.48 3.27 1.64 Motivation Income Class Rank* Autonomy Need Satisfaction Relatedness Need Satisfaction Competence Need Satisfaction # Drinks Judges 5.20 6.49 3.80 1.05 1.80 1.84 2.53 Service 4.55 5.74 3.67 0.73 1.43 1.61 2.67 Prestige 3.47 8.05 3.15 0.71 1.49 1.46 2.90 Other 3.62 6.30 3.47 0.71 1.40 1.37 2.86 * For class rank, increasing means indicate lower class ranks. 2. Do Attorney Preferences and Work Settings Affect the Factors That Promote Their Well-Being? We used these practice group distinctions to perform two more analyses, investigating the possibility that key factors for well-being might operate differently in groups of lawyers with different values, goals, or motivations. It might be that the impact of specific well-being correlates would vary based on differing personal priorities of individuals, their immersion at work with other people sharing (and hence reinforcing) the same motivations and values, or both. For example, in “prestige” law offices with many people focused on high income, the effect size with well-being of the collectively desired goal (income) might be greater than for service-oriented lawyers in serviceoriented settings. Similarly, in offices of predominantly service-oriented people, the apparent effect of service motivation might also be more important than in offices of lawyers who do not equally share this purpose. Analyses of income and work motivation in the “prestige” and “service” groups suggested that the factors retain their importance for well-being regardless of personal values or work setting. The two fundamentally different factors (income and internal motivation) pre 594 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 dicted well-being statistically to the same extent in both work groups (income to well-being: r = .19 for the “service” group, .21 for the “prestige” group; internal motivation to well-being: r = .55 for the “service” group, .57 for the “prestige” group; no significant differences). These data indicate that, regardless of people’s subjective preferences, their choices based on internal or external goals and motivations will predict their well-being to the extent reported in the primary findings. The consistency of these focused analyses is striking, and also adds confidence in the study methods and findings. C. Other Work Variables 1. Hours Worked, Firm Size, and Billable Hours The purportedly negative work environment of large law firms is legendary and typically includes long work hours and high billable hour requirements.151 We hypothesized that well-being would tend to decrease (1) as firm size increased, (2) as number of hours worked increased, (3) as the number of required billable hours increased, and (4) with less seniority and status within the firm. Based on the demonstrated importance of internal factors for well-being, we thought that billable hours would be particularly inimical, because the required record-keeping and reporting would keep lawyers constantly focused on an extrinsic goal (money), undermine their autonomy, and increase the sense of supervisory control rather than autonomy support. Similarly, because junior associates in firms often report low autonomy, we expected status in the firm to strongly correlate with wellbeing. We also hypothesized that, if income increased with firm size, lawyers working in the larger firms would be more externally motivated for those financial benefits, and their well-being would consequently suffer despite the increased income. All participants were asked to type in the number of hours they worked in an average week. Subjects who worked in a private firm were asked the number of attorneys in the firm, their position in the firm, whether billable hours were required, and if so, the number of such hours required on an annual basis. We then correlated these variables with the various well-being measures included in the survey. 151 See James J. Alfini & Joseph N. Van Vooren, Is There a Solution to the Problem of Lawyer Stress? The Law School Perspective, 10 J.L. & HEALTH 61, 62–64 (1995–96); Susan Saab Fortney, The Billable Hours Derby: Empirical Data on the Problems and Pressure Points, 33 FORDHAM URB. L.J. 171, 176–83 (2005); Schiltz, supra note 2, at 888–95. 2015] WHAT MAKES LAWYERS HAPPY? 595 a. Total Hours and Billable Hours Surprisingly, well-being did not vary significantly with absolute number of hours worked. The data did not support our expectations that the presumed additional stress of longer hours, and the decreased time for personal and family interests, would take a toll on lawyer happiness. We explored the data for further understanding in two ways. Analyzing the well-being within each of the four groups of subjects previously discussed (“judges” and “prestige,” “service,” and “other” lawyers), we found no significant correlation with hours worked in any group. We also generated subsamples within the primary working sample according to reported hours worked weekly. We determined mean well-being for those subsamples reporting from thirty to more than seventy work hours weekly.152 Table 2 shows the very small differences in mean SWB for each subgroup. TABLE 2. HOURS WORKED AND SWB Mean Hours SWB SD N 30–39 4.894 1.931 640 40–44 4.856 2.931 1382 45–49 4.854 2.012 1156 50–54 4.892 1.921 1317 55–59 4.764 1.968 437 60–64 4.761 2.062 444 65–69 4.716 2.276 96 70+ 4.859 2.205 46 These data do disclose a very slight tendency for SWB to decrease above fifty-four hours of work per week. However, the small number of subjects working seventy or more hours reported SWB equal to that of the subjects working the fewest hours, and none of the differences in SWB are significant.153 152 There were 672 subjects who reported 0 to 29 hours worked weekly; we did not analyze this group because the work hours were atypically low and likely attributable to a variety of reasons (such as retirement, unemployment, underemployment, illness, injury, disability, or independent sources of income) that would have different impacts on well-being. 153 Further perspective regarding these very small differences can be gained by recalling the varying SWB of the “judges” and the “prestige,” “service,” and “other” lawyers. The SWB of the 1,434 “prestige” lawyers (M = 4.86) was virtually equal to that of the lawyers reporting from 30 to 55 hours of work weekly (all M = 4.85–4.89). The SWB of the 2,852 “other” lawyers 596 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 Ultimately, the further analysis provided only detail regarding the overall lack of a relationship between hours worked and SWB. This null result may reflect beneficial correlates of increased hours that balance the expected negatives. First, private attorneys tend to realize significant additional income with increasing work hours (r ==.23, p=< .001). Second, for any lawyers who are enjoying their work as a result of strong engagement or interest, longer hours could manifest and perhaps increase that enjoyment. Further research would be required to explain this surprising null result with confidence. Although total hours worked had no bearing on well-being, the data concerning billable hours was telling. Compared to private attorneys reporting no billable hours requirement, those that had any billable hours requirement reported only slightly less well-being (r = −.06, p < .01). The hypothesized decrease in autonomy need satisfaction was supported (r = .16, p < .001), and billable hours were accompanied by increased alcohol use. Within the subsample reporting billable hours (N = 975), there were a number of meaningful correlations as billable hours increased. As expected, income increased (r = .22, p < .001), but important psychological predictors of well-being decreased—autonomy satisfaction (r = −.15), internal motivation (r = − .15), and relatedness satisfaction (r = −.11, all p < .001). The net result supported our hypothesis: subjects experienced less life satisfaction and lower net affect as billable hours increased. The decrease in SWB was small (r = −.10, p < .001), likely offset in part by the increasing income. Ultimately, the data show that each increase in billable hours brings moderately greater income and slightly less happiness. b. Size of Law Firm As predicted, with increasing firm size lawyers reported more external motivation for their work and less autonomy satisfaction (respectively, r = .14, r = −.13; both p < .001). On the other hand, income increased robustly with firm size (r = .46, p < .001). Both sets of changes were quite linear across the spectrum from sole practice to firms of 750 or more lawyers. Internal motivation was greatest for solo attorneys, but income was greatest for the largest firms. Considering all private attorneys, there was a very small decrease in wellbeing approaching significance as firm size increased (r = −.031, p = .122, N = 4,060), resulting from slightly less positive affect in larger firms (r = −.06, p = .02). Thus, the data indicate that in the larger firms (M = 4.71) was virtually equal to that of the lawyers reporting from 55 to 69 working hours (M = 4.71–4.76). 2015] WHAT MAKES LAWYERS HAPPY? 597 lawyers earn much more money but nonetheless experience no greater well-being and slightly more negative mood than those in smaller private firms. 2. Position Within Law Firm The instrument asked private firm attorneys to indicate whether they were a managing partner, senior partner, junior partner, senior associate, or junior associate. In order to capture meaningful results for these analyses, we focused on subjects in the primary working sample who were in firms large enough to be stratified by position, and thus excluded firms of fifteen or fewer attorneys. The resulting subsample for these analyses was N = 748. The number of subjects in four of the five positions did not vary greatly (N = 154–227); twentyeight subjects reported themselves as managing partners.154 Means for primary variables typically fell along a continuum, varying as expected based on seniority of position. The data provided noteworthy results, not all of which were expected. (1) The “word on the street” regarding diminished well-being among junior associates in firms was amply supported. The mean SWB for this group was very significantly lower than for senior associates, the next lowest mean (M = 4.13, 4.69 respectively; p = .01). The reasons were manifest, in that the means for most well-being factors in this group were significantly lower than those of the next lowest group (again, senior associates). Most notably, mean differences for each of the human needs critical for well-being exceeded one-half of a response interval: autonomy (M = 0.28, −0.22; p < .01), competence (M = 1.55, 0.78; p < .01), and relatedness (M = 1.50, 0.97; p < .01). (2) The second major finding was surprising. “Making partner” represents a milestone of success in law firm culture that presumably makes the new partners more secure, more affluent, and much happier. Although our data did confirm a major increase in compensation (M = 7.58, 10.47; p < .001—representing a difference between junior partners and senior associates of sixty-two percent, or $69,000 annually155), the partners experienced no greater happiness than the 154 Given the small number of managing partners and the similarity in data between them and the senior partners, for simplicity we considered these two as a single group for most purposes. 155 Calculated mean income for the two groups was $112,000 for senior associates and $181,500 for junior partners. The calculations are not fully precise, because each response choice represented a range of income. 598 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 associates. For the partners (N = 193), raw means for life satisfaction, positive and negative affect, and autonomy satisfaction showed tiny improvements compared to the associates (N = 163). None of these differences, however, was statistically significant despite the relatively large sample sizes, and perhaps most telling, aggregate well-being for the partners and associates was essentially identical (M = 4.72, 4.69 respectively; r = .01; p = .88156). These findings represent another indication of the limited ability of higher status and increased income to impact the well-being of professionals. Perhaps more importantly, it should give law graduates and lawyers pause when considering law firm positions. Although students and lawyers likely assume that making partner after several years of often difficult associate work will “pay off” with great happiness, such expectations are entirely unsupported by these data. (3) Taken together, lawyers in firms larger than fifteen were not happy when compared to the large lawyer groups previously reported. The “other” lawyers, the group that includes many typical private practice positions (general, family, and criminal defense, among others), offer a useful comparison, because they had the lowest mean well-being of the groups previously reported. Their mean well-being, however, was essentially identical to that of junior partners in the firms with more than fifteen lawyers (M = 4.73, 4.71, not significant). Other contrasts were also telling: firm junior associates had far lower well-being than the “other” lawyers (M = 4.07, 4.71; p < .001), and the overall well-being across the firm subsample was significantly lower than the well-being of both the “judges” (M = 4.79, 5.76; p < .01) and the “service” lawyers (M = 4.79, 4.98; p < .05). (4) Motivation in the firm subsample appeared to be the key factor undermining well-being. Income in the firms was notably high compared to the groups previously reported, which would be predicted to provide a modest increase in well-being. Indeed, lawyers in the lowest-earning firm position, junior associates, had earnings equivalent to the “judge” group and had greater earnings than “service” (M = 6.77, 5.74; p < .01) and “other” lawyers (M = 6.77, 6.30; p < .01). However, external motivation for gaining income was significantly higher in the law firm group157 than in all of the other lawyer and judge groups reported. As a result, overall motivation for work, a 156 This very high p value strongly suggests a random rather than meaningful difference in the means. 157 Income motivation is consistent at each position within the law firm subsample, with no significant differences. 2015] WHAT MAKES LAWYERS HAPPY? 599 first-tier factor for well-being, was very significantly less healthy for the law firm subsample than for the “judges,” “service,” or “other” groups (M = 1.30 versus 5.20, 4.55, and 3.62 respectively; all p < .01). Simply stated, lawyers appear to be choosing to work in medium and large firms for the wrong reasons in psychological terms, and hence their happiness is impaired despite their enhanced income. 3. Litigation Practice and Private and Public Attorneys Litigation inevitably generates stress on attorneys. Adversarial, zero-sum contests are stressful by nature,158 and the uncertain outcomes often have serious consequences for clients. We therefore hypothesized a modest negative well-being impact on subjects who litigated controversies (N = 3,097). The instrument asked subjects to indicate the percentage of their work that involved litigating cases (including any stage of litigation). As predicted, as litigation increased in subjects’ practices, there was an inverse, albeit very small, correlation with well-being (r = −.06; p < .01). We investigated the possibility that intrinsic purpose and internal motivation for the case work might moderate any negative impact of litigation stress by comparing subjects in the “prestige” and “service” groups. There was indeed evidence of this protective effect: the negative correlation of well-being with litigation was greater (although still small) in the “prestige” group (r = −.10, p < .01) than in the “service” group (r = −.04, not significant). Thus, the data suggest a small negative effect on wellbeing from litigation, with substantial moderation of the effect from adaptive motivation and values. Further research would be required to draw confident conclusions. 4. Pro Bono and Community Service Work Pro bono work is commonly encouraged in the legal profession and logically should promote well-being because it embodies the intrinsic values of altruism and community involvement and improvement.159 This benefit could be moderated among lawyers, however, because many are required to perform pro bono work by their bar associations or employers,160 and thus this virtuous work may often be 158 See Krieger, Human Nature, supra note 5, at 277–84 (describing the effects of an adversarial atmosphere on law students). See generally Seligman et al., supra note 148. 159 Many people derive “intense satisfaction” from unpaid service work. FREY & STUTZER, supra note 8, at 105. 160 See Scott L. Cummings & Rebecca L. Sandefur, Beyond the Numbers: What We Know—and Should Know—About American Pro Bono, 7 HARV. L. & POL’Y REV. 83, 84, 98 (2013). 600 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 experienced as externally compelled rather than internally motivated.161 The instrument asked subjects to report the number of “unpaid public service/pro bono/volunteer hours (whether legal or other community service)” that they “provide in a typical month.” Hours of unpaid service did correlate with greater positive affect (r = .13, p < .001), as well as SWB (r = .08, p < .001). Confirming the likely source of the well-being benefit, hours also related to internal motivation for work and intrinsic values orientation (r = .12, .06 respectively; both p < .001). D. Personal Demographics Subjects were asked to provide basic demographic data. Because there were few significant demographic differences in our law student data,162 we expected minimal well-being differences between the groups in our attorney sample as well. We did hypothesize, based on previous research in general populations,163 that increasing age would associate positively with well-being, as would a current marriage or similar primary relationship. 1. Age Previous research has shown that people generally tend to be happier as they age.164 We expected this result in our lawyer sample as well, because maturity would logically bring more autonomy, competence, and self-knowledge, and also provide more income and a better fit with personally preferred work. The mean age of the sample was 46.66 years, and data did show increasing age to predict well-being (r = .17, p < .001). To investigate likely reasons, we determined the correlations with age of income and internal work motivation. Both correlations with age were significant: income, r = .16; internal work motivation, r = .23 (both p < .001). Regression of age and these 161 Motivation for pro bono work was not central to this study and was not addressed by the survey. 162 Sheldon & Krieger, Understanding Negative Effects, supra note 9, at 889–90. See generally Sheldon & Krieger, supra note 5. 163 See Hyoun K. Kim & Patrick C. McKenry, The Relationship Between Marriage and Psychological Well-Being: A Longitudinal Analysis, 23 J. FAM. ISSUES 885 (2002). See generally MYERS, supra note 19, at 68–79; Kennon M. Sheldon & Tim Kasser, Getting Older, Getting Better? Personal Strivings and Psychological Maturity Across the Life Span, 37 DEVELOPMENTAL PSYCHOL. 491 (2001) (employing perceived maturity of personal goals and finding older individuals may be more psychologically mature than younger people and may be happier as a result). 164 See, e.g., MYERS, supra note 19, at 68–79; Sheldon & Kasser, supra note 163. 2015] WHAT MAKES LAWYERS HAPPY? 601 factors with well-being revealed that internal motivation and income were indeed the operative factors; age itself was unrelated to wellbeing. Thus, our hypotheses were supported in that age predicted greater well-being in our bar members, but other factors that increase with age accounted for the apparent well-being benefits.165 2. Gender Gender differences and dynamics are a matter of considerable interest in the legal profession.166 However, based on our studies of law students, we expected little or no difference in lawyer well-being by gender. We did expect men to have greater mean ages and therefore have more time in the profession, resulting in a well-being benefit at least from increased income and autonomy. We speculated that this benefit might be offset by well-being advantages from stereotypical, but perhaps somewhat accurate, positive internal qualities often attributed to women (e.g., intrinsic valuing of intimacy, community, and helping others, and more relatedness need satisfaction).167 The data supported these hypotheses, showing well-being between male (N = 3,740) and female (N = 2,340) lawyers to be virtually the same. Men were favored with a very small raw difference in mean SWB (4.89 versus 4.77; p < .05), with no difference in incidence of depression by gender. Supplemental analyses did reveal the kinds of offsetting differences we predicted. Men on average had been working in legal jobs longer, had slightly greater autonomy and competence satisfaction, and had higher income and lower loan balances upon graduation (likely resulting from the more recent mean graduation date of women). Men also tended more towards lucrative “prestige” positions. Women, however, were stronger in the kinds of salutary psychological variables predicted, showing more intrinsic values, greater relatedness satisfaction, and more affinity for service-oriented positions. Interestingly, each of these gender differences was 165 As noted earlier, debt incurred in law school was another factor that decreased with age. Regressions of age and law school debt with well-being showed that about one-third of the associations of each variable with well-being was shared. 166 See generally, e.g., Elizabeth H. Gorman, Work Uncertainty and the Promotion of Professional Women: The Case of Law Firm Partnership, 85 SOC. FORCES 865 (2006); Lani Guinier et al., Becoming Gentlemen: Women’s Experiences at One Ivy League Law School, 143 U. PA. L. REV. 1 (1994); Joan M. Krauskopf, Touching the Elephant: Perceptions of Gender Issues in Nine Law Schools, 44 J. LEGAL EDUC. 311 (1994); Adam Neufeld, Costs of an Outdated Pedagogy? Study on Gender at Harvard Law School, 13 AM. U. J. GENDER SOC. POL’Y & L. 511 (2005). 167 Even if these gender differences are true, there would, of course, be many exceptions. 602 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 very highly significant (p < .001), but ultimately they offset to generate virtual equality of well-being between the genders. 3. Race and Ethnicity Our main working sample included 5,810 Caucasians, 257 African Americans, 80 Asian/Pacific Islanders, and 78 Hispanic/Latinos. The findings overall were similar to those for gender. MANOVAs168 for mean differences between the groups showed no significant differences for either well-being or depression. Supplemental analyses showed other differences, however (p < .01 for all). Caucasians had higher earnings, but they also had less internal motivation for their work and consumed alcohol more intensely than others. African Americans reported the least alcohol use; Asians and Hispanics generally fell in the middle of the groups on most measures. Overall, then, the data supported our hypothesis of little or no difference between groups, and again demonstrated that higher earnings will not generate increased well-being if motivation for work is not also positive. 4. Marriage and Social Support As in our law student studies, we asked subjects to identify themselves as either “married (or in long-term, committed relationship),” “dating and in serious relationship,” “dating but not in serious relationship,” or “single/not dating.” We used these categories to create a scale of social support; MANOVAs revealed significant differences in well-being for each step of the support continuum. The data confirmed our expectation that subjects in a marriage-type relationship would show the greatest well-being of these groups. Depression findings were consistent, with married subjects showing the fewest (and single subjects the most) depressive symptoms. This is consistent with findings in other populations.169 Conversely, single/not dating lawyers showed the lowest well-being, although they also drank less than all other groups. It might be that lawyers who are single and not dating are more solitary or serious people, but further research would be needed to determine this. We then compared the group with the highest well-being (married, N = 4,690) with all unmarried subjects (N = 1,396), finding an effect size of marriage for well-being of r = .17 (p < .001). We investi 168 MANOVA, or multivariate analysis of variance, is used to investigate mean differences of multiple independent variables. TABACHNICK & FIDELL, supra note 18, at 22. 169 See Myers, supra note 87, at 62–64. 2015] WHAT MAKES LAWYERS HAPPY? 603 gated the relationship between marriage and satisfaction of the needs for autonomy, relatedness, and competence—the strongest predictors of well-being in the study. Married subjects reported higher satisfaction of all three needs, more so for relatedness as expected (r = .15, p < .001). Autonomy and competence showed very small, but also significant, correlations (r = .06 and r = .09 respectively, both p < .001).170 Regression of marriage and need satisfaction with wellbeing showed that about half of the apparent well-being benefit of marriage was related to increased need satisfaction, while the remainder of the effect was unrelated to these needs. Further analyses (not central to this study) would very likely show other well-being benefits, such as greater mean age, income, and internal work motivation, associated with marriage. As seen next, having children would be another important contributor to well-being that would strongly associate with marriage. 5. Children Having children results in powerful experiences that might exert opposing influences on well-being. Children generate ongoing, often major, stress on parents (in the literal sense of “stress” as a demand that requires a response171), and at the same time children are typically regarded as sources of deep joy and love. We thought it particularly interesting to research the well-being impact of children on busy professionals with demanding schedules. Would attorney life leave space for real enjoyment of children, or might children more often be experienced as another burden in the lives of this purportedly highly stressed occupational group? We had no hypothesis regarding this question. The survey instrument asked subjects about the number of children they had, offering choices from 0 to “more than 10.” There were 3,850 subjects who responded to this question, a very large sample despite the many who did not respond. (It may be that many who did not respond did not have children and hence ignored the question.) Of those responding, sixty-seven percent indicated that they had at least one child. We analyzed responses first to view the contrast between having children (any number) and having no children. Subjects with one or more children reported moderately greater well-being 170 The competence need is global in SDT theory and this instrument, rather than focused only on work or school tasks. Hence, responses may reflect subjects’ sense of competence in personal as well as work life. 171 See HANS SELYE, THE STRESS OF LIFE 74 (rev. ed. 1976). 604 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 (r = .20, p < .001). Perhaps surprisingly given the other demands on lawyers, well-being also increased as the number of children increased (r = .18, p < .001). Confirming these findings, depressive symptoms related inversely with number of children (r = −.11, p < .001), as did alcohol consumption (r = −.07, p < .001). Of course, the occurrence of children and marriage strongly tend to coincide, so regressions of the two would almost certainly show substantially overlapping relationships with well-being. We did not perform this and many other analyses that might be interesting but which were tangential to the purposes of this study. E. Law School Ranking We collected information on the primary law school172 attended by subjects, to determine the extent to which differences in their schools’ rankings predicted lawyer well-being. The instrument listed many regional schools and all schools located in or near the states whose bar associations participated in the study. Subjects were asked to select their primary law school from the list, and if not included, to type in the name of the school. A number of subjects declined to respond, perhaps because their school was not listed. Nonetheless, about 5,000 lawyers provided their school information, resulting in a large subsample for this analysis (N = 4,768 subjects responding to all related variables). The sample included graduates of 186 U.S. law schools, almost every institution ranked in the widely recognized U.S. News & World Report (“USNWR”) publication that we used for the study.173 Subjects were distributed from the first through fourth tiers of the rankings as follows: N = 1,926; 1,101; 1,321; and 420.174 The top 100 schools were coded with their specific USNWR ranking; third and 172 This was generally the school from which subjects graduated, unless they attended that school for only a short time while seeking their basic law degree. 173 U.S. NEWS & WORLD REPORT:AMERICA’S BEST GRADUATE SCHOOLS 28–32 (2011 ed.). Although the accuracy of these rankings is often questioned, they do generate a shared perception of relative ranking. There were other potential sources of inaccuracy in this analysis, however. Because our subjects spanned many decades of law practice and rankings vary from year to year, any rank chosen would be inexact for the entire sample. Nonetheless, because the relative standing of most law schools does not vary markedly across time, we proceeded with this analysis. Various rationales could lead to the use of rankings from different years during the decades covered by our subjects’ careers; we chose the last rankings published before our data were collected. All subjects would have entered practice after publication of the rankings, and we thought that the methodology used by the publishers for calculating the rankings might have been refined with time and experience. 174 The number of fourth tier graduates participating may reflect the fact that the participating states included few such schools, and where included, some were new and had few graduates to participate in the survey. 2015] WHAT MAKES LAWYERS HAPPY? 605 fourth tier schools were assigned ranking numbers at the midpoint of their tier (essentially all were treated as “average” for their tier), because schools in those tiers are ranked by USNWR only with a tier designation.175 Knowing the “resume value” that many selective employers place on law school ranking, we expected students from more elite schools to report substantially greater earnings.176 Because ranking and income are external factors, we further expected better ranking of a school to show a small positive relationship with well-being. Other factors might vary this result. Because more elite graduates would tend to have more work choices, ranking might well enhance internal motivation for work and therefore well-being. The positive correlation with well-being could also be greater if school quality or status resulted in admitting students with positive characteristics that would independently translate later to well-being or satisfaction (such as intrinsic values, alertness, clarity, perseverance, resilience, and interpersonal skills). On the other hand, any well-being benefit of higher ranking would be curtailed if the prestige or financial rewards of positions available to more elite school graduates attracted them away from lower-paying positions for which they felt more passion or interest. Given these mixed potential effects, we hypothesized a modest positive relationship of rank with lawyer well-being, largely based on increased income. Data were generally supportive, but to a surprisingly slight extent. The data showed an almost meaningless correlation between law school ranking and lawyer well-being (r = .05, p < .01), despite a modest correlation with greater income (r = .15, p < .001). Consistent with the very small well-being correlation, both depression and positive affect were unrelated to school ranking. Higher ranking was barely positive for internal motivation (r = .03, p < .1); this marginally significant outcome may result from the displacement dynamic suggested above, wherein a number of elite graduates are foregoing internally motivated work in favor of additional pay or other benefits. Satisfaction of 175 This lack of specificity would introduce another source of potential error for some purposes, particularly for determining quality of a school (if indeed the rankings themselves are an accurate measure of quality). However, because the rankings up to this year were expressed in this way, this approach to quantification would likely approximate the perception of these lower tier schools by students and employers (who would also have had only the same general tier designation to consider). 176 Increased earnings could, of course, also occur if attending a more elite school reflected other positive qualities that would independently increase earning ability. This study did not seek to address such questions. 606 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 each of the three basic needs also showed consistent, equally tiny increases with improving school rank (all r = .03 to .04; all p < .05). We investigated further, as the insubstantial findings contrasted markedly with the efforts among schools to improve their ranking status, and among students to attend schools with higher rankings. We subdivided the top 100 schools into segments of twenty-five schools each, and compared means of important variables within the six resulting groups (four groups within the top 100 schools, and the third and fourth tiers as before). The numbers of subjects in these six ranking groups were: top 25 schools, N = 804; schools ranked 26–50, N = 1,122; schools ranked 51–75, N = 832; schools ranked 76–100, N = 269; schools ranked 101–145, N = 1,321; schools ranked 145 and above, N = 420. A few patterns appeared, confirming and providing detail to the main findings177: (1) the means for well-being, income, and internal motivation for work declined fairly consistently from the top to the bottom of the rankings; (2) although the overall well-being variance was very small (r = .05), there was even less variance among the subjects in the middle rankings (seventy-five percent of the sample; schools ranked number twenty-six through the third tier), with means for schools in this range varying little from the means for the entire sample; (3) larger differences in means occurred at the extremes, above rank twenty-six and below the third tier, accounting for a disproportionate part of the variance across the entire sample; and (4) the third tier was anomalous, showing a positive “bump” for all variables—means for income and well-being exceeded those for the schools ranked seventy-five to one hundred, while internal work motivation nearly equaled that of the highest group (the top twenty-five schools). Further research would be needed to explain these phenomena with any confidence. The primary finding here is the very small association of law school ranking with lawyer well-being. It recalls the data on law review membership, another seemingly important external marker of high career potential among law students. With that prestigious achievement, subjects also enjoyed a modest increase in income, and in that case realized no associated well-being benefit. While such results are consistent with SDT principles and research, they are surprising because they challenge common assumptions regarding elite education, distinguished academic honors, and greater income as 177 As with the fine comparisons of the many differences between occupational and demographic groups, these comparisons are reported to provide a sense of patterns in the data only, and thus are not accompanied by significance calculations. 2015] WHAT MAKES LAWYERS HAPPY? 607 prime pathways to a happier working life. These and similar results in this study simply confirm the limited ability of external factors to generate well-being, and should encourage legal job seekers to weigh job offers carefully—particularly considering whether the work or work environment will promote or undermine the more powerful psychological correlates of well-being.178 F. Personal Life and Balance Choices “Work-life balance” and “stress management” have been common themes in the legal profession, suggesting that personal behavioral choices may provide answers to depression and other problems lawyers may face.179 Based on the primary findings here, choices that promote or express autonomy, relatedness, competence, internal motivation, or intrinsic values are likely to support lawyer well-being. We have reported positive results for two such choices in personal life—marrying and having children. We surveyed other personal life choices that appeared likely to impact lawyer well-being, asking subjects to indicate any of the following in which they typically engaged “at least weekly and for at least the past two months”: “exercise,” “meditation or mindfulness,” “yoga/tai chi,” “active sports or martial arts,” “prayer, affiliated with a religious organization,” “prayer, not affiliated with a religious organization,” and “personal discussion group (book club, support group, etc.).” We hypothesized that such practices would be associated with increased well-being. They would likely provide relaxation and respite from work, moderate stress, improve general health, and promote autonomy, self-understanding and growth, relatedness to others, and other subjective benefits. 1. Physical Activities: Exercise, Sports and Martial Arts, and Yoga and Tai Chi Subjects reporting regular exercise had greater well-being than others (r = .17, p < .001), supported by greater satisfaction of all three 178 See LAWRENCE S. KRIEGER, A DEEPER UNDERSTANDING OF YOUR CAREER CHOICES 12–14 (3d ed. 2013) (providing actual examples of law students that demonstrate the application of these principles as guides to a successful job search). 179 See generally AMIRAM ELWORK, STRESS MANAGEMENT FOR LAWYERS: HOW TO INCREASE PERSONAL & PROFESSIONAL SATISFACTION IN THE LAW (3d ed. 2007); Adele J. Bergin & Nerina L. Jimmieson, Explaining Psychological Distress in the Legal Profession: The Role of Overcommitment, 20 INT’L J. STRESS MGMT. 134 (2013); Barbara S. McCann et al., Hostility, Social Support, and Perceptions of Work, 2 J. OCCUPATIONAL HEALTH PSYCHOL. 175 (1997); Myers, supra note 87; Nerison, supra note 2; Deborah L. Rhode, Balanced Lives for Lawyers, 70 FORDHAM L. REV. 2207 (2002). 608 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 needs (r = .14 for autonomy and relatedness, r = .11 for competence; all p < .001). Regressions showed that roughly half of the effect size of exercise on well-being was related to increased need satisfaction; the remaining effect was independent of these needs and might be attributed to increased energy, clearer thinking, general health, or other factors often associated with physical exercise. Subjects participating in active sports or martial arts showed a very small well-being advantage over others (r = .08, p < .001). The more relaxing and meditative activities, yoga and tai chi, were not related to well-being—a surprising finding, because individuals commonly report a variety of benefits from these practices. Prayer and meditation practices showed similar results. These null findings are discussed together after reporting the data on prayer and meditation below. 2. Vacations We asked subjects to indicate the number of days they took in the past year for vacation (regardless of whether paid or unpaid, or whether they remained at home or took a vacation trip). As a matter of both theory and common sense, we hypothesized that vacation days taken would associate positively with attorney well-being. People tend naturally to enjoy “time off,” and to come back to work refreshed as a result. From an SDT perspective, vacations could (1) indicate personal autonomy and well-being/life balance as a personal goal; (2) reflect autonomy at the work place or internal motivation (doing things for the inherent enjoyment in them); and (3) associate positively with relatedness, because lawyers with friends or partners might well take more vacations, experience increased relatedness satisfaction on vacations with others, or both. The number of vacation days did correlate moderately with aggregate well-being (r = .23, p < .001) and related positively to important well-being factors as predicted: autonomy, relatedness, and competence satisfaction, r = .24, .19, and .15 respectively, and internal motivation for work, r = .12 (all p < .001). We considered whether more vacations taken might simply reflect more vacation days permitted by employers, another item measured in the survey. This yielded another surprising result: the correlation of vacation days used and vacation days permitted was inverse (r = −.10, p > .01).180
180 This negative relationship may result from many employers offering benefits that are not realistic or encouraged given the associated workloads, hourly and billable demands, management style, standards for promotion, or other matters. Such speculations would require further research for clarification.
2015] WHAT MAKES LAWYERS HAPPY? 609
The correlation strengths of vacation days and exercise with wellbeing are noteworthy, because they equal, and in some cases greatly exceed, the effect size for well-being of increasing income, decreasing debt, better grades, law review participation, or law school ranking. Such findings may provide helpful perspectives for law students anxious about debt or grades, or lawyers disappointed with their present earnings. Other simple, healthful life choices appear to be equally (or more) important for happiness and satisfaction in life.
3. Religious and Spiritual Practice The United States is known as a country with relatively high religious participation.181 There is current interest in the legal profession in spiritual practices as a way to alleviate stress and improve judgment.182 We thought that many lawyers involved with any such regular practice or observance would report greater well-being than others. We were also interested to see how such practices would compare to vacations and the surveyed physical activities in this regard, as all might well provide relief from stress and reflect intrinsic growth values, attention to autonomy, and a willingness to create life balance. As previously mentioned, the survey instrument asked respondents to indicate participation in meditation/mindfulness and prayer (distinguishing whether or not affiliated with a religious organization). Results provided virtually no support for the hypothesized benefits of these practices. The data showed slightly increased well-being (r = .07, p < .001), only for subjects practicing prayer and who were affiliated with a religious organization (N = 2,263), but not for the meditating subjects (N = 684), nor for those practicing prayer without organizational affiliation (N = 401). Given that prayer and meditation are fundamentally individual practices, we considered that most of the meditating subjects and those in the unaffiliated prayer group may pursue their practices by themselves, and that the fellowship afforded at worship services and as a congregant might account for much of the apparent benefit in the affiliated prayer group. It might also be that subjects in the latter group were generally more social and less solitary 181 See MYERS, supra note 19, at 177, 182. See generally BARRY A. KOSMIN & SEYMOUR P. LACHMAN, ONE NATION UNDER GOD: RELIGION IN CONTEMPORARY AMERICAN SOCIETY (1993). 182 See generally, e.g., STEVEN KEEVA, TRANSFORMING PRACTICES: FINDING JOY AND SATISFACTION IN THE LEGAL LIFE (1999); Rhonda V. Magee, Educating Lawyers to Meditate?, 79 UMKC L. REV. 535 (2011); Leonard L. Riskin, The Contemplative Lawyer: On the Potential Contributions of Mindfulness Meditation to Law Students, Lawyers, and Their Clients, 7 HARV. NEGOT. L. REV. 1 (2002). 610 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 than the others, thus enjoying greater relatedness satisfaction and well-being independent of their religious practice. Supplementary analyses offered support for this proposition. The affiliated prayer group reported greater mean relatedness need satisfaction than the other two groups, and, indeed, than the rest of the entire sample as well (r = .08, p < .001). In order to distinguish the impact of relatedness in this affiliated group from that of prayer itself, we entered both factors in a regression equation. After accounting for relatedness benefits, a very small but still significant residual correlation remained between well-being and prayer in the affiliated group ( = .05, p < .01). Thus, the data support only the conclusion that subjects engaging in prayer when affiliated with a religious group are slightly happier than others.183 As stated above, data showed yoga and tai chi practices to bear no relationship to well-being across the sample, much like unaffiliated prayer, meditation, and mindfulness reported here. These data do not mean that such practices are not helpful for increasing well-being, but they certainly offer no evidence that they are. The null result here is not entirely surprising. First, given the cross-sectional study design, we do not know the level of well-being of any of these subjects before they began their practices. It might be, for example, that lawyers who choose more self-sufficient approaches of meditation, mindfulness, and prayer tend to be more anxious or worried, less optimistic and trusting of established organizations, or more isolated and less attracted to joining with others in community, and thus need to “work harder” to maintain baseline well-being in the first place. Second, because of the length of the survey, this section of the instrument was very brief. We sought only to identify subjects with any level of recent and regular participation (specifically, “at least weekly” participation for at least “the past two months”). It is likely that many respondents had minimal relevant involvement and experience. The instrument also did not seek to define or distinguish among potentially very different practices, determine whether respondents had received training for their practice (and if so, the qualifications of their trainers), or determine the length of time devoted by subjects to each sitting or session. In short, our single, broad screening measure yielded no result, and more focused, nuanced investigation into these and other variables would be required to support 183 Discussion/book/support groups also showed a similar, very small positive relationship with well-being (r = .04, p < .01), perhaps representing another example of a self-enrichment activity with a social component. 2015] WHAT MAKES LAWYERS HAPPY? 611 findings with confidence. An experimental or longitudinal design would also undoubtedly help to gain insight into these questions. G. Smaller City Life and Practice We thought that smaller cities and towns, when compared to large cities, would provide more pleasant, less stressful work environments and, similarly, more cordial, less aggressively adversarial relationships within the community of lawyers. We asked subjects to identify the city population of their primary work location. Results supported our hypothesis, but with exceedingly small (though statistically significant) correlations. As city population decreased, subjects reported very slightly increased well-being (r = .03, p < .05). There were other very small effect sizes with decreasing city size: more positive perceptions of lawyers, judges, and the justice system (r = .04, p < .01),184 more internal work motivation (r = .08, p < .001), and more intrinsic values orientation (r = .04, p = .001). Smaller population centers also predicted moderately lower income (r = −.19, p < .001).185 Thus, our perhaps romanticized notion about “small town” life and practice was very modestly supported, with smaller population centers showing slight positive associations despite lower attorney income. H. Perceptions of Professionalism and Faith in the Justice System Measurement of professional behavior was beyond the scope of this study. We did, however, ask subjects for perceptions of the legal system and profession, factors that would likely impact the well-being of members of the profession. Questions focused on the fairness of case outcomes and the level of professionalism of judges and other lawyers (limited to those whom subjects observed regularly, to increase reliability). Fair outcomes and attorney behavior were each addressed by a pair of contrary statements, while behavior of judges was addressed by one statement.186 A factor analysis indicated that all five items related to a single factor in subjects’ thinking, probably reflecting the totality of their ex 184 Findings regarding perceptions of the profession are presented more fully in the next section. 185 Living expenses are likely somewhat lower as well, but this was not a subject of the study. 186 Items appeared as follows: “I don’t have much faith that our legal system produces fair outcomes;” “I am concerned by the low level of professionalism among lawyers;” “The lawyers I have encountered consistently exhibit appropriate, professional behavior;” “The judges I have encountered consistently exhibit professional behavior;” and “I believe the law most often produces just results.” 612 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 periences within the profession and justice system. The combined mean response of all subjects to all questions was 3.22 (after recoding negative statements to a positive orientation) on the 5-point scale, only slightly above the “neutral” response position.187 These thousands of lawyers, in other words, on average had only a very slightly positive sense of their peers and the legal system that provides their livelihood. We performed two targeted analyses. We first combined the responses to the items relating to faith in the law and in outcomes of the system. The mean response was 3.24, again slightly above neutral. We also analyzed the responses evaluating the behavior of the judges and other lawyers that subjects had encountered. As one should expect, mean responses were somewhat higher for judicial than attorney professionalism (M = 3.39, 3.06 respectively; p < .01). Nonetheless, while the mean rating of attorney professionalism was essentially neutral, the higher rating of judges did not reach the midpoint between “neutral” and the first response choice that represents any level of approval of judicial professionalism.188 These results suggest that there is much room for improvement in the professionalism of judges and lawyers.189 As reported in the preceding section, lawyers in smaller population centers reported a slightly more positive view, as assessed by combining the five items in one overall measure (r = .04, p < .01). But fundamentally, this large sample of professionals has a positive view of neither the justice in the justice system nor the professional behavior of professionals in the system—a very troubling finding and a call to action for legal educators and bar leaders.190 187 Responses were provided on a 5-point Likert scale; choices were “strongly disagree,” “disagree,” “neutral,” “agree,” and “strongly agree.” 188 The judges, analyzed separately, did have modestly higher opinions of their own professionalism and of the fairness of outcomes which, of course, they determine to a great extent (both p < .001). This pattern suggests an example of self-serving bias. 189 See Daicoff, supra note 130, at 1344; Schiltz, supra note 2, at 906–08 (noting prevailing public perception of lawyers as unethical). See generally Peter A. Joy, A Professionalism Creed for Judges: Leading by Example, 52 S.C. L. REV. 667 (2001); Ronald D. Rotunda, Lawyers and Professionalism: A Commentary on the Report of the American Bar Association Commission on Professionalism, 18 LOY. U. CHI. L.J. 1149 (1987). There is no lack of published concern about lawyer professionalism. 190 The finding here that lawyers’ perception of judges is little better than their perception of other lawyers may surprise bar leaders. These findings indicate important directions for inquiry by bar associations: how do bar members think just and fair outcomes might better be attained, and what can be done to improve the professionalism of judges as well as lawyers? 2015] WHAT MAKES LAWYERS HAPPY? 613 I. Expected Earnings Compared with Actual Earnings Long experience with law students indicates that their expected future earnings are a major factor in attracting many applicants to law school. Students also consistently appear to have an unrealistic, inflated sense of lawyers’ earnings.191 We therefore hypothesized that among the study participants earnings expectations when deciding to attend law school would not have been realized in practice. If this hypothesis were supported, the finding could have important implications for law student and lawyer well-being and for the overall tenor of the profession. Given that the critical factors for lawyer well-being are not related to earnings, encouraging new people to enter the profession for high earnings (even if expectations were accurate) could result in a pool of future attorneys predisposed to disappointment and lower well-being in their careers as well as in law school. This negative effect would tend to be exacerbated when externally motivated students or lawyers learn of their realistic earning potential. Other law school applicants who are motivated more for public service would also encounter more competition for entry to law school, perhaps resulting in overall diminution of a public service orientation among law graduates. We thought the most relevant data regarding the earnings expectations of prospective law students would be for their first years after graduation, when their economic circumstances would be least stable and their educational debt highest. The survey instruction therefore read: “Think back to when you were applying to law school, what your expectations were at that time for your future earnings in your first few years as a lawyer after law school.” Subjects were then asked: “Tell us how your actual earnings in your first few years working in law jobs compared to those expectations.”192 Response choices included “much more than expected,” “more than expected,” “about as expected,” “less than expected,” and “much less than expected.” Far more subjects, by a factor of almost nine to one, indicated that their earnings were much less than expected (N = 941) compared to much more than expected (N = 108). The combined number of subjects reporting earning either less or much less than expected 191 In my experience for over two decades, law students typically indicate expectations of earnings after graduation in the range of $100,000, far above the mean income for new graduates. See Starting Salaries—Class of 2012, NALP, http://www.nalp.org/starting_salaries_class_of _2012 (last visited Mar. 1, 2015); Salary Trends—A 15-Year Overview, NALP, http://www.nalp. org/2005julsalarytrends (last visited Mar. 1, 2015). 192 Emphasis in original survey. 614 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 (N = 2,190—55.3% of the sample) was more than four times greater than the combined number earning more or much more than expected (N = 501—12.7% of the sample);193 about one-third of the sample (32%) reported earning about as expected. The data therefore strongly supported our hypothesis about the unrealistic expectations of prospective law students for earnings following graduation. This finding points to an information gap with many potential negative consequences, and one that could readily be eliminated by clear disclosures from college career counselors and on law school and bar association websites. We recognize that such disclosures might impose difficult pressures on law schools, resulting from fewer applicants, less ability to select for high LSAT scores and undergraduate grades, and generation of less revenue. Nonetheless, it would seem that open disclosure of likely earnings would benefit the entire profession, and it would be fundamentally fair to the large percentage of apparently naive law school prospects indicated by these data. J. Brief Discussion of Secondary Findings Secondary inquiries focused on three groups of factors that were likely to impact well-being: personal life choices, work-related variables, and demographic differences. The data generally supported and supplemented the hierarchy of well-being factors revealed by the primary findings, while the repeating patterns in the data and the consistency of findings added confidence in the design and validity of the study as a whole. From a more practical perspective, analyses yielded specific findings that may guide lawyers towards more satisfying lives and careers. Specific secondary findings are summarized below; we note the apparent contributing factors (generally internal) that appear to contribute to each result. Comparisons of subjects in “service,” “prestige,” “other law practice,” and “judge” positions offered an applied example of the relative importance of internal and external well-being factors. The “judges” comprised the only career group with high scores for positive factors of both types and reported the greatest SWB of the four groups. Public service lawyers had the lowest grades and earnings of the lawyer groups, but nonetheless reported greater well-being than even the 193 Although there is a current employment problem within the legal profession, the sample spans several decades of graduation dates, and the measure is stated in terms of expectations and actual earnings for the “few years following law school graduation.” Thus, the data reflect long-term tendencies rather than short-term variations in the job market or lawyer pay. 2015] WHAT MAKES LAWYERS HAPPY? 615 “prestige” group, with the highest grades and earnings. This appeared to result from the more intrinsic values and internal work motivations of the “service” lawyers, combined with the greater importance for well-being of those internal factors (compared to grades and income). A related analysis of subjects in firms with sixteen or more lawyers confirmed that junior associates substantially lacked well-being despite ample compensation, apparently as a result of decreased satisfaction of the needs for autonomy, competence, and relatedness. Surprisingly, well-being was not enhanced at all for junior partners when compared to senior associates, despite the partners’ improved status and a sixty-nine percent increase in income. Various contrasts indicated that, overall, lawyers in these firms are very well paid but experience diminished happiness compared to other lawyers, largely because they are choosing medium or large firm work for external reasons that do not promote well-being. Among more specific work-related factors, increasing vacation days most strongly correlated with increasing well-being. Pro bono service hours, which embody the intrinsic value of altruistic service, also predicted increased well-being. By contrast, the specific practice factor that had the strongest negative relationship with well-being was required billable hours. This practice was associated with increasing income but decreasing autonomy, relatedness, and internal motivation, an apparent example of managers undermining workers’ self-determined motivation and well-being by promoting a focus on external rewards. Thus, as billable hours go up, income goes up and happiness goes down. Increasing law firm size presented another apparent example of the internal-external dichotomy, as firm size correlated with increasing income but decreasing autonomy, internal motivation for work, and well-being. Litigation activities bore a modest negative relationship to well-being; this correlation was substantially smaller for “service” than for “prestige” lawyers. Results for one work factor were surprising: the number of hours that lawyers worked each week showed no significant relationship with well-being. Possible explanations were suggested. Alcohol use proved to be an indicator of negative well-being, and was associated with decreased internal motivation and decreased experiences of autonomy, relatedness, and competence. Married (or similarly committed) subjects had the greatest well-being and fewest depressive symptoms of the social support groups, while single/not dating lawyers were the least happy. Regressions showed increased 616 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 need satisfaction, particularly relatedness, to largely account for the increased well-being in married lawyers. Subjects with children were also happier than those without children, and more children also predicted greater well-being. The predictive value for well-being of being married or having children was equal to, and often greater than, the predictive value of any of the external financial and status factors. Analyses of other personal choice factors focused on those that might provide life balance or stress relief. Physical exercise was related to increased satisfaction of all needs and predicted well-being to the same extent as higher income, lower debt, marriage, or children; active sports was a lesser but also positive factor. The number of vacation days taken was the strongest predictor of well-being among all specific activities measured in this study. It was associated with increased internal motivation for work and greater satisfaction of autonomy, competence, and relatedness. This provided a healthy overall picture of happier lawyers who tend to choose work for meaning and enjoyment, who enjoy the company of others, and who also tend to take personal breaks for rest and recreation. Analysis of the relationship between the USNWR rankings of law schools and the well-being of their graduates was striking. Ranking modestly associated with higher income, but only very slightly with well-being. These results were similar to those regarding another presumed marker of potential for success—selection to a law journal. Data regarding engagement in religious or spiritual practices and relaxing movement were also somewhat surprising. Subjects practicing meditation/mindfulness, prayer (unaffiliated with a religious congregation), and yoga or tai chi reported the same well-being as other subjects. Prayer for those associated with a congregation showed a small positive correlation with well-being, with increased relatedness need satisfaction accounting for part of this positive result. Possible explanations were suggested; the usefulness of these practices for lawyers would benefit from continued research. Demographics other than age showed little predictive value for lawyer well-being. There were reportable, modest differences between the genders and racial and ethnic groups, but ultimately wellbeing showed no or negligible differences between the groups. As expected, older lawyers were moderately happier than younger lawyers, with regressions showing the well-being benefit to relate to increasing internal motivation and, secondarily, increasing income. Two additional inquiries yielded concerning results. First, lawyers broadly reported that their expectations when entering law school for 2015] WHAT MAKES LAWYERS HAPPY? 617 their early career earnings were not realized in their actual careers. This information gap could be addressed by law schools and state and national bar associations, yielding important positive consequences both for individuals and the profession as a whole. Second, on questions related to professionalism and the judicial system, subjects did not agree to a meaningful extent that the legal system provides fair or just results, or that either judges or lawyers whom they had observed behaved professionally. Further study would be important to replicate and clarify the current findings, and to determine specific steps that might improve the profession.194 VIII. SUMMARY A. What Makes Lawyers Happy? Data from several thousand lawyers in four states allowed us to test hypotheses of primary importance for lawyer well-being, as well as subsidiary factors and practices providing further relevant insights. Well-being was quantified by combining life satisfaction with positive affect and then subtracting negative affect. We included a depression scale and measures of alcohol consumption, as checks on our wellbeing measures and also because of common concerns about depression and substance use among attorneys. Results were typically expressed as standardized (Pearson) correlations to permit comparison of the association strength of factors with subjective well-being and other important variables. Primary hypotheses addressed contrasting sets of variables: subjective/psychological factors established by Self-Determination Theory to promote well-being in general populations and objective/ external factors typically emphasized in legal populations—grade performance, law journal membership, law school debt, and income after graduation. The data supported all primary hypotheses, showing that psychological factors were far more important for the well-being of attorneys than the various external factors. Factors addressed by the primary hypotheses fell into three tiers of importance, based on their strength of association with well-being. Tier (1): Experiences of autonomy (including authenticity), relatedness to others, and competence most strongly predicted attorney well-being; correlations ranged from .63 for competence to .66 for autonomy. These large correlations indicate that well-being co-occurs 194 We explain below why improving the well-being of lawyers will likely improve their professionalism and productivity as well. Infra Part VIII.C. 618 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 with these factors so robustly that it may not be possible to experience thriving without relative satisfaction of all of these needs.195 Choosing work for internally motivated reasons (i.e., for enjoyment, interest, or meaning within subjects’ belief systems) was also very highly predictive of well-being, with a correlation of .55. Tier (2): Autonomy-supportive supervision of attorneys at the work place (provision of understanding, respect, and choices, as opposed to control) strongly predicted well-being (r = .44). Replicating law student research, autonomy support also appeared to increase the critical experiences of autonomy, competence, and relatedness, suggesting itself as an effective intervention for promoting well-being. Intrinsic values (for self-improvement, intimacy, and altruism/ community), as compared to extrinsic values (for affluence, power, or recognition) had the next highest correlation with well-being (r = .30), falling squarely between the preceding measures and the external factors. Tier (3): The external factors constituted a distinctly subordinate tier of apparent benefits for well-being, with correlations ranging from .00 for law review membership to .19 for law school debt at graduation and for attorney income. Class rank, perhaps the most emphasized and stress-inducing factor in law school, correlated rather weakly with well-being (r = .12). These results suggest the need for a core reorientation of priorities that deemphasizes grades, credentials, status, and money as foundations of happiness in the legal profession. Secondary analyses addressed many other factors hypothesized to impact attorney well-being. They further supported the principal conclusion regarding the primacy of internal over external factors for well-being, in many cases also suggesting that need satisfaction and internal motivation may be undermined by emphasizing external factors.196 Particularly salient examples included: (1) “prestige” job lawyers, with the highest grades and income of all groups analyzed, were not as happy as “service” lawyers, the group with the lowest pay and law school grades; (2) although income increased very strongly with law firm size (r = .46), well-being decreased at the same time; (3) billa 195 For example, the correlation in this sample between aspiring to values and acting on the same values was .70, very similar to the needs/well-being correlations. Perhaps more on point, the correlation between depression and its virtual mirror image, well-being, was -.69, hardly more strongly associated with well-being (though inversely) than the three needs. One prominent psychologist has included these needs in her expanded definition of well-being. See Carol D. Ryff & Corey Lee M. Keyes, The Structure of Psychological Well-Being Revisited, 69 J. PERSONALITY & SOC. PSYCHOL. 719, 720 (1995). 196 See infra note 209 and accompanying text. 2015] WHAT MAKES LAWYERS HAPPY? 619 ble hours, which manifest an external focus on generating revenues, were the strongest negative predictor of well-being studied despite a positive .22 association with increased income; (4) subjects in law firms with sixteen or more lawyers were very highly compensated but reported external motivations and broadly diminished well-being; (5) junior partners in the same firms reported identical well-being as senior associates, despite the greatly enhanced income and status of the partners; and (6) higher law school ranking was associated with increased income, but it correlated only negligibly with well-being. Other secondary findings refine and supplement the hierarchy of well-being factors suggested by the primary analyses. Each of the secondary factors meaningfully predicting well-being also correlated with important SDT factors, most particularly internal motivation for work or satisfaction of one or more of the psychological needs. In addition to the specific work-related variables discussed above, secondary findings included two new sets of well-being factors—personal life choices and demographics. The personal life choices showed generally stronger predictive power for well-being than the external grades, money, and credentials factors, becoming the third tier in the hierarchy of well-being factors and moving the external “success” factors to fourth position. Demographics had the least associations with wellbeing. Thus, the fuller study results may be represented by five tiers of factors that predict and appear to promote lawyer happiness, listed in descending order and shown in Table 3: (1) Needs for autonomy, relatedness, and competence, and internal motivation for one’s work (r = .66–.55). (2) Autonomy-supportive supervision and intrinsic values (r = .44–.30). (3) Personal life choices, including taking vacation days, having children, being in a marriage or similar relationship, and exercising at least weekly (r = .23–.17).197 (4) Affluence/prestige/“success” factors, including income, law school debt, class rank, law school rank, partnership in a firm, and law review participation (r = .19–.00). 197 As with the external factors, some of the personal variables bore zero correlation to well-being (and quantity of alcohol consumption was negative). They are not noted here because they were included for interest only, whereas law review and law firm partnership have central importance in the “success” paradigm of law students and lawyers. Table 3 includes a more comprehensive list of findings. 620 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 (5) Demographics, which had zero to negligible well-being correlations, other than age (r = .17, but shown by regression to be attributable to internal work motivation and income). TABLE 3. FIVE TIERS OF WELL-BEING FACTORS (STANDARDIZED CORRELATIONS) Tier SWB Tier 1 Autonomy need satisfaction .66 Relatedness need satisfaction .65 Competence need satisfaction .63 Internal work motivation .55 Tier 2 Supervisor autonomy support .44 Intrinsic values .30 Tier 3 Vacation days taken .23 Children .20 Married/long-term committed .17 Exercise .17 Prayer (affiliated/congregation) .07 Alcohol use (quantity) −.12 Tier 4 Income .19 Law school debt (decreasing)198 .19 Class rank .12 Law school ranking .05 Partnership in firm .00 Law journal .00 Billable hours −.10 Tier 5 Age .17 Other demographics .00−.03 198 Debt is, of course, an inverse affluence factor. The r value here reflects decreasing debt balance at the time of graduation from law school. The impact of debt is actually less than indicated, as regression showed approximately one-third of the correlation to be attributed to the younger age of subjects with higher debt on graduation. 2015] WHAT MAKES LAWYERS HAPPY? 621 B. Lawyers Are Not Different from Other People with Regard to Their Happiness and Satisfaction As we expected, this question was essentially answered as analyses uncovered the determinants of attorney well-being. However, we wanted to address the question explicitly because of the special importance of happiness in the range of human experiences,199 because lawyers are often considered to think and act differently than others, and because lawyers are apparently trained to do so in law school (i.e., to “think like lawyers”). Further, our previous studies revealed core changes in student values and motivations during law school,200 and a linguistic analysis of basic law training found consistent undermining effects on student values, interpersonal caring, and moral and ethical decisionmaking.201 All of this suggested the possibility that lawyers, whether by nature or through training, may respond differently than other people to psychological and external factors that typically generate well-being in the general population. In addition, such training might convince lawyers that the usual sources of human well-being do not relate to them. If this belief were false, it could lead to life choices quite inimical to well-being. The data were consistent and clear when viewed from this perspective, comparing the correlates of happiness in lawyers and in other people. The tenets of SDT established by decades of research in the general population appeared to apply without qualification to this large sample of legal professionals. The relative strength of different factors was also essentially as expected, with fundamental needs and self-determined (internal) motivations more strongly predicting wellbeing than values, and with subjective psychological factors appearing substantially more important than external factors. Simply stated, there is nothing in these data to suggest that attorneys differ from other people with regard to their prerequisites for feeling good and feeling satisfied with life. Thus, it would appear that lawyers, and their teachers and employers, should banish any notions that lawtrained people are somehow special in this important regard. In order to thrive, we need the same authenticity, autonomy, close relationships, supportive teaching and supervision, altruistic values, and focus on self-understanding and growth that promotes thriving in others. 199 See Lyubomirsky, Happier than Others, supra note 8, at 239. 200 See Sheldon & Krieger, supra note 5, at 279; Sheldon & Krieger, Understanding Negative Effects, supra note 9, at 890, 893–94. 201 See MERTZ, supra note 14. 622 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 C. Improved Well-Being Implies Improved Productivity, Ethics, and Professionalism Performance was not measured in this study, but previous research indicates that the benefits of well-being extend to improved accomplishment of complex mental tasks,202 generally improved work performance, and greater culturally valued success.203 Specifically among law students, increased well-being and internal motivation (resulting from enhanced autonomy support and need satisfaction) were accompanied by better grades and bar exam performance.204 Numerous other studies show well-being to correlate with performance and productivity in general populations, with substantial evidence that well-being promotes health, energy, optimism, creativity, altruism, and work performance.205 Happier employees also tend to remain with employers longer and raise the morale (and hence performance and retention) of others in the organization;206 less happy employees impose high costs on employers in terms of increased absence and turnover and poor work performance.207 The current data demonstrate that lawyers who find interest and meaning in their work are much more likely to be happy than others; such engagement also makes high productivity more likely.208 Conversely, previous research indicates that motivation based on external factors such as increased financial incentives can actually result in decreased performance and productivity, likely by displacing (“crowding out”) more salutary internal motivation for work.209 These facts, coupled with the current 202 See Lyubomirsky et al., Positive Affect, supra note 8, at 840 (noting, in a meta-analysis of hundreds of related studies, some conflicting results among studies, but an overall positive effect size between positive affect and mental performance of r = .25). 203 See id. at 840, 846; MYERS, supra note 19, at 127–41. 204 See Sheldon & Krieger, Understanding Negative Effects, supra note 9. 205 See Huang & Swedloff, supra note 2, at 337; see also FREY & STUTZER, supra note 8, at 105. 206 See Huang & Swedloff, supra note 2, at 337 nn.9–17. 207 See FREY & STUTZER, supra note 8, at 105. 208 See DANIEL H. PINK, DRIVE: THE SURPRISING TRUTH ABOUT WHAT MOTIVATES US 86–88 (2009) (discussing internal motivation and productivity); see also FREY & STUTZER, supra note 8, at 105 (regarding the connectedness of well-being, internal motivation, and productivity in the work place); KOHN, supra note 128, at 186. 209 See FREY & STUTZER, supra note 8, at 105 (referring to “hundreds” of laboratory experiments and actual work settings documenting the phenomenon of external incentives undermining work productivity). A meta-analysis of 128 related experiments concluded that “tangible rewards tend to have a substantially negative effect on intrinsic motivation.” Edward L. Deci et al., A Meta-Analytic Review of Experiments Examining the Effects of Extrinsic Rewards on Intrinsic Motivation, 125 PSYCHOL. BULL. 627, 658–59 (1999). See generally KOHN, supra note 128, at 119–41 (discussing the effect of financial incentives on workplace productivity). 2015] WHAT MAKES LAWYERS HAPPY? 623 data showing a very large (r = .55) correlation of internal motivation with well-being, support the conclusion that increased well-being and productivity will associate with each other, mediated in large part by the extent of workers’ sense of autonomy and internal (versus external) motivation. The survey also did not seek to measure professionalism or ethics, but it did measure psychological factors that are virtually certain to be important sources of ethical and professional behavior for lawyers—authenticity (which is essentially identical to integrity),210 competence, relating well to others, helping and community values, and valuing self-understanding and growth.211 These factors also include the strongest predictors of well-being in our subjects, suggesting that one powerful approach to raising the level of professional behavior among lawyers is to teach law students and lawyers to maximize their own happiness. D. What the Findings Mean for Lawyers and Their Teachers and Employers While many lawyers, their teachers, and their employers attribute great importance to grades, rankings, honors, and financial rewards, earlier research on general populations revealed basic flaws in the “American Dream” paradigm that regards money, status, and other external markers of success as foundations of a happy life.212 The current study provides data from a very large sample of lawyers that repeatedly support the same conclusion—there were no external rewards or status factors that strongly, or even moderately, predicted attorney well-being. This research quantifies and highlights the subordinate importance of external considerations that often dominate law schools and law practice, and further highlights the greater importance of personal and interpersonal considerations that are commonly subordinated in law schools and practice. The data contradict beliefs that prestige, income, and other external benefits can adequately compensate a lawyer who does not regularly experience autonomy, integrity, close relationships, and interest and meaning in her 210 See generally Krieger, Most Ethical People, supra note 21, at 174–75 (pointing to SDT well-being factors as sources of professionalism, and discussing the essential identity of integrity and the autonomy/authenticity need). 211 These connections recall the Mertz findings, supra notes 64–70 and accompanying text, that replacing values, connection to self, and caring for others with competitive success results in eroding the personal foundations of ethical decisionmaking. 212 See, e.g., Kasser & Ryan, supra note 36, at 921–22; Kasser & Ryan, supra note 76, at 286; Sheldon et al., supra note 33, at 335–36. 624 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 work. The data therefore suggest fundamental changes in the belief system shared by many law students, lawyers, and their teachers and employers. In particular, the shared understanding of “success” needs to be amended so that talented students and lawyers consistently avoid choices in the pursuit of material success that will undermine their happiness. We offer only brief comments on ways these findings might be applied by various groups. For pre-law students, the data suggest choosing a law school for its sense of fit with their personal values and personal learning goals and styles, rather than focusing on school prestige and USNWR rankings. Law students and lawyers would realize greater well-being from culturing their sense of self, internal purpose, and positive relationships with other people in personal and professional life than from focusing intensely on rewards and recognition. For teachers and employers, the findings repeatedly suggest the need for a systematic effort to recast perceptions of “success” in law school and the profession, by shifting institutional emphases from competition, status, and tangible benefits to support, collaboration, interest, and personal purpose. The research suggests particularly important responsibilities for law teachers. They impact students early in the formation of professional attitudes and identities, and that impact is apparently negative for many students, particularly with regard to the kinds of internal psychological factors found here to be the primary correlates of lawyer well-being.213 First, educating law students about these findings should decrease anxiety, stress, and excessive competition, because grades, honors, and the other zero-sum competitive factors measured in the study had limited to nil associations with well-being. By contrast, none of the factors found to bear strongly on well-being involve limited resources; all are products of a student’s or lawyer’s individual choices. A second important strategy for law teachers would be to approach the task of teaching legal analysis with humility, clearly conveying to students that, although this skill will enable them to dispassionately analyze and argue legal issues while setting aside their own instincts, values, morals, and sense of caring for others, such a skill must be narrowly confined to those analytical situations. This is not a superior way of thinking that can be employed in personal life, or even in most work situations, without suffering psychological consequences.214 For private sector employers, shifting the 213 See MERTZ, supra note 14, at 132, 134; Sheldon & Krieger, supra note 5, at 280, 282; Sheldon & Krieger, Understanding Negative Effects, supra note 9, at 893–94. 214 See MERTZ, supra note 14, at 97–137. 2015] WHAT MAKES LAWYERS HAPPY? 625 external rewards paradigm (including billable hours) towards psychological well-being would likely need to be addressed in steps, but should result in improved productivity, morale, and retention. Public sector employers may increase the satisfaction and retention of their lawyers by educating them about the relative well-being of “service” lawyers compared with those in the private sector. One specific cost-effective strategy supported by the data for application in both school and work settings is the provision of autonomy-supportive, rather than controlling, teaching, mentoring, and work supervision. This practice can be learned215 and, as stated, has been shown to promote broad improvements in well-being, motivation, and performance.216 Teachers and employers may consult detailed guides for providing autonomy-supportive mentoring and teaching, and for implementing additional approaches to achieve improved well-being and performance outcomes.217 The generation of more such guides, including with a specific focus on the private law firm, would be beneficial. Two additional concerns for law teachers and employers were raised by the data. First, subjects had only neutral opinions of the professionalism of lawyers they encountered and only slightly above neutral opinions about appropriate outcomes in the legal system. They also did not approach agreement that judges’ behavior was appropriate. Second, subjects broadly reported entering law school with inflated expectations for their earnings as lawyers. While these concerns would benefit from more focused research, given their importance they would seem to merit immediate attention from law teachers, employers, and bar leaders. IX. LIMITATIONS AND FUTURE DIRECTIONS The data were generated by self-report rather than objective observation, and lawyers from only four states were surveyed, with a relatively small percentage of such lawyers responding. While sam 215 See Manning, supra note 128, at 232–33, 235 for a thorough and clear example of teaching others how to provide autonomy support. This example focuses on law teachers providing written critique of law student work, but both the general concepts and many of the specific recommendations and examples would equally apply to attorney supervisors. 216 See supra notes 57–70, 113–14 and accompanying text. 217 These strategies and many others are described in some detail, with recommendations for step-by-step implementation by law schools, in Krieger, Human Nature, supra note 5, at 284–308. Such recommendations would broadly apply to employers as well. See PINK, supra note 208, at 83–106 (describing provision of autonomy-supportive management in work places); see also FREY & STUTZER, supra note 8, at 103–05. 626 THE GEORGE WASHINGTON LAW REVIEW [Vol. 83:554 ples were large and results were highly consistent internally, the possibility remains that the results do not generalize broadly to lawyers in the United States. As previously noted, certain limitations were also inherent in the study design. In particular, while this study employed an extensive survey instrument and its results were consistent with previous longitudinal findings, the cross-sectional design did not permit confident causal conclusions. Finally, standardization of effect sizes for factors expressed in different metrics was accomplished in accordance with accepted practice, but such standardization does not eliminate the conceptual challenge of directly comparing factors measured in different units or by different methods. Thus, while the relative power of the internal and external factors for predicting wellbeing was repeatedly supported by secondary analyses, caution is advised when assuming mathematical precision in comparing effect sizes of factors initially measured in different units. Future research would benefit from longitudinal design. Although the scope of such studies and the diversity of subjects is likely to decrease, such results could be considered together with this and future cross-sectional studies to deepen understanding of the issues. This study did not seek to compare lawyers with other groups of working adults. Future studies of other occupational groups would further illuminate the extent to which lawyers, law training, and legal work may be unique, potentially leading to improvements in educational, hiring, and management practices. Specific unexpected findings may be of interest to particular researchers and would benefit from targeted study. Such findings might include the inverse relationship between vacation days permitted and days actually taken, and the relationships with well-being of law journal participation, hours worked at a job, partnership in a law firm, and self-improvement practices such as meditation or yoga. CONCLUSION These data from a large and diverse sample of practicing attorneys establish that the processes governing the well-being and life satisfaction of people generally, as elaborated by Self-Determination Theory research, fully apply to lawyers. Psychological factors related to self, others, meaningful and personally engaging work, and supportive work supervision were far more predictive of well-being than external “success” factors relating to competitive standing, honors, status, or financial rewards. Striking examples included highly competitive and prized achievements such as law review membership and 2015] WHAT MAKES LAWYERS HAPPY? 627 making partner in a law firm, neither of which bore any relationship to the well-being of subjects. Secondary analyses also showed that public service lawyers were happier and more satisfied than other lawyers, including those in the most prestigious, highly paid positions. Further, across the sample, a number of personal routine and lifestyle choices matched or exceeded the power of income, honors, and credentials as predictors of lawyer well-being. Informing law students, lawyers, and their teachers and employers about these findings could serve a number of important goals, including improved well-being, performance, and ethical behavior across the profession. [table id=9 /]