NYCLA-CLE I n s t i t u t e Supervising JusticeofMatrimonialMatters,Bronx County; University ofRochester; toward certification inciviltriallaw, criminaltriallaw, workerscompensation lawand/ormatrimonial law. hours oftotalCLE credit.Ofthese,1qualify ashoursofcreditforEthics/Professionalism, and0qualify ashoursofcredit This programhas beenapprovedbytheBoard ofContinuingLegalEducation oftheSupremeCourt NewJerseyfor3 Board foramaximumof3 Transitional &Non-Transitional credithours:1Ethics;2PP This coursehasbeenapproved inaccordancewiththerequirementsofNew Hon. DennyChin, T in Prepared inconnectionwithaContinuingLegalEducationcoursepresented at New York CountyLawyers’ Association, 14 Vesey Street, New York, NY 3 he TRANSITIONAL &NON-TRANSITIONAL J U.S. Courtof Appeals, 2dCircuit; R udicial Hon.Betty Weinberg Ellerin, 1st Dept.andcurrentlyat Alston &Bird,LLP P Hon. JudithGische, presented on Thursday, May2,2013. r g o r ole M P NYCLA’s JudicialSection r g o r P a r g o r m C aking a h

m a I A C of m r F o & App.Div., 1stDept. - C A

S M p formerPresidingJustice, Appellate Division, D u o Hon.La Tia Martin, d o

l n t E r o s y MichaelS.Ross,Esq.; e : r ecision a : mpa r o t York State ContinuingLegalEducation MCLE CREDITS: : t NYS Sup. Ct. and NYSSup.Ct.and hy Prof. Maya Sen, Prof.MayaSen,

Information Regarding CLE Credits and Certification The Role of Empathy in Judicial Decision Making May 2, 2013; 6:00 PM to 9:00 PM

The New York State CLE Board Regulations require all accredited CLE providers to provide documentation that CLE course attendees are, in fact, present during the course. Please review the following NYCLA rules for MCLE credit allocation and certificate distribution.

i. You must sign-in and note the time of arrival to receive your course materials and receive MCLE credit. The time will be verified by the Program Assistant.

ii. You will receive your MCLE certificate as you exit the room at the end of the course. The certificates will bear your name and will be arranged in alphabetical order on the tables directly outside the auditorium.

iii. If you arrive after the course has begun, you must sign-in and note the time of your arrival. The time will be verified by the Program Assistant. If it has been determined that you will still receive educational value by attending a portion of the program, you will receive a pro-rated CLE certificate.

iv. Please note: We can only certify MCLE credit for the actual time you are in attendance. If you leave before the end of the course, you must sign-out and enter the time you are leaving. The time will be verified by the Program Assistant. Again, if it has been determined that you received educational value from attending a portion of the program, your CLE credits will be pro-rated and the certificate will be mailed to you within one week.

v. If you leave early and do not sign out, we will assume that you left at the midpoint of the course. If it has been determined that you received educational value from the portion of the program you attended, we will pro-rate the credits accordingly, unless you can provide verification of course completion. Your certificate will be mailed to you within one week.

Thank you for choosing NYCLA as your CLE provider!

New York County Lawyers’ Association Continuing Legal Education Institute 14 Vesey Street, New York, N.Y. 10007 • (212) 267-6646

The Role of Empathy in Judicial Decision Making Thursday, May 2, 2013; 6:00 PM to 9:00 PM

Program Chair & Moderator: Hon. Judith Gische, App.Div., 1st Dept. Program Co-sponsor: NYCLA's Judicial Section Program Chair & Moderator: Hon. Judith Gische, App.Div., 1st Dept. Faculty: Hon. Denny Chin, U.S. Court of Appeals, 2d Circuit; Hon. La Tia Martin, NYS Sup. Ct. and Supervising Justice of Matrimonial Matters, Bronx County; Michael S. Ross, Esq.; Prof. Maya Sen, University of Rochester; Hon. Betty Weinberg Ellerin, former Presiding Justice, Appellate Division, 1st Dept. and currently at Alston & Bird, LLP

AGENDA

5:30 PM – 6:00 PM Registration

6:00 PM – 6:10 PM Introductions and Opening Remarks

6:10 PM – 9:00 PM Panel Discussion • What is empathy? • Do judges actually use empathy in their roles as decision makers? Anecdotal evidence; Empirical evidence? • Are there particular areas of the law that call for empathy in judicial decision making? Is there a role in sentencing? Is there a role in child custody decisions? • Are there ethical considerations that either restrict or require a judge to be an empathic decision maker? Should empathy always yield to the rule of law?

New York County Lawyers’ Association

Continuing Legal Education Institute 14 Vesey Street, New York, N.Y. 10007 • (212) 267-6646

The Role of Empathy in Judicial Decision Making Thursday, May 2, 2013 6:00 PM to 9:00 PM

Program Co-sponsor: NYCLA's Judicial Section

Program Chair & Moderator: Hon. Judith Gische, App.Div., 1st Dept.

Faculty: Hon. Denny Chin, U.S. Court of Appeals, 2d Circuit; Hon. La Tia Martin, NYS Sup. Ct. and Supervising Justice of Matrimonial Matters, Bronx County; Michael S. Ross, Esq.; Prof. Maya Sen, University of Rochester; Hon. Betty Weinberg Ellerin, former Presiding Justice, Appellate Division, 1st Dept. and currently at Alston & Bird, LLP

Table of Contents

Examining Empathy: Discrimination, Experience and Judicial Decisionmaking By Jill D. Weinberg & Laura Beth Nielsen, 85 Southern California Law Review 313 (2012)

Empathetic Judging and the Rule of Law By Susan Bandes, 2009 Cardozo L. Rev. 133

Sentencing: A Role for Empathy By Hon. Denny Chin, 160 U. Pa. L. Rev. 1561

Identifying Judicial Empathy: Does Having Daughters Cause Judges to Rule for Women’s Issues By Adam Glynn and Maya Sen, July 28, 2012

Faculty Biographies

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

EXAMINING EMPATHY: DISCRIMINATION, EXPERIENCE, AND JUDICIAL DECISIONMAKING

JILL D. WEINBERG*

LAURA BETH NIELSEN†

TABLE OF CONTENTS I. INTRODUCTION ...... 314 II. ASSESSING DISCRIMINATION: THREE PERSPECTIVES ON JUDICIAL DECISIONMAKING ...... 321 A. THE EMPATHETIC PERSPECTIVE ...... 324 B. THE LIBERAL LEGAL PERSPECTIVE ...... 327 C. THE POLITICAL PERSPECTIVE ...... 328 III. AN EMPIRICAL ANALYSIS OF FEDERAL DISTRICT COURT JUDGES AND DISCRIMINATION ...... 330 A. DATA COLLECTION ...... 330 B. STATISTICAL MODELS AND HYPOTHESES ...... 333 IV. RESULTS ...... 336 A. LOGISTIC REGRESSION MODEL ...... 339 B. INDIVIDUAL-LEVEL ANALYSES...... 342 1. Judge-Plaintiff Minority Status ...... 343

* Ph.D. (candidate), Northwestern University; J.D. 2008, Seattle University; M.A. 2009, University of Chicago; M.A. 2010, Northwestern University. † Research Professor, American Bar Foundation. Associate Professor of Sociology and Law & Director of Legal Studies, Northwestern University. J.D. 1996, Berkeley School of Law (Boalt Hall); Ph.D. Jurisprudence and Social Policy 1999, University of California, Berkeley. This research was supported by the American Bar Foundation, the National Science Foundation (#SES-0417389), and the Searle Foundation. The authors would like to thank Ellen C. Berrey, Patti Ewick, Ryon Lancaster, Robert L. Nelson, and Christopher W. Schmidt, as well as the participants at the Conference of Empirical Legal Studies (New Haven, 2010) who provided feedback on earlier versions of this Article, especially Bert Huang who provided great comments as our discussant.

313 WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

314 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

C. SUMMARY ...... 345 V. EMPATHY, EXPERIENCE, AND A NEW VISION OF JUDICIAL DECISIONMAKING ...... 346 A. DIVERSITY OF THE JUDICIARY ...... 347 B. EMPATHY AND THE EFFICACY OF THE ADVERSARIAL SYSTEM ...... 349 C. ACCOUNTING FOR EMPATHY AND THE RULE OF LAW ...... 349 VI. CONCLUSION ...... 351

I. INTRODUCTION There are moments when the law is not enough. In Virginia v. Black,1 a normally silent Justice Clarence Thomas interjected with what one commentator called a “Luke-I-am-you-father”2 voice. The case involved a First Amendment challenge to a Virginia law that prohibited cross burning. During a deputy U.S. solicitor general’s oral argument in favor of the law, Justice Thomas condemned him for not going far enough. Justice Thomas, who grew up in the segregated South and was the only black Justice on the bench, posed a very potent question: “Aren’t you understating the . . . effects of . . . the burning cross” given that crosses were “symbol[s] of [a] reign of terror” during the “100 years of lynching . . . in the South?” He continued, “I think that what you’re attempting to do is to fit this into our jurisprudence rather than stating more clearly what the cross was intended to accomplish.”3 Similarly, during the Senate Judiciary Committee hearing on the nomination of then-nominee, now-Justice Sonia Sotomayor, Senator Jeff Sessions challenged her prior representations that she could be an impartial

1. Virginia v. Black, 538 U.S. 343 (2003). The specific issue before the Court was whether an anti-cross-burning statute violated the First Amendment right to symbolic expression. Id. at 351–52. Symbolic expression, or “symbolic speech,” refers to actions that convey a particular message. Generally, symbolic expression has been protected by the First Amendment; most notably, flag burning, wearing armbands, or sit-ins have been a form of protest. For a broad discussion of these cases, see Frederick Schauer, Intentions, Conventions, and the First Amendment: The Case of Cross-Burning, 2003 SUP. CT. REV. 197; Timothy Zick, Cross Burning, Cockfighting, and Symbolic Meaning: Toward a First Amendment Ethnography, 45 WM. & MARY L. REV. 2261 (2004). 2. Dahlia Lithwick, Virginia Burning: Are Cross-Burnings Speech or Violence?, SLATE (Dec. 11, 2002, 6:21 PM), http://www.slate.com/id/2075301/. 3. Transcript of Oral Argument at 20–21, Virginia v. Black, 538 U.S. 343 (2002) (No. 01- 1107).

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 315 judge by quoting remarks she made the day before: “You have repeatedly made this statement: ‘I accept the proposition that a difference there will be by the presence of women and people of color on the bench, and that my experiences affect the facts I choose to see as a judge.’”4 Without hesitation, Sotomayor responded, “the point that I was making was that our life experiences do permit us to see some facts and understand them more easily than others.”5 Ultimately, Senator Sessions made his stance clear that empathy and judicial decisionmaking can and should be mutually exclusive, saying, “Call it empathy, call it prejudice, or call it sympathy, but whatever it is, it is not law. In truth, it is more akin to politics, and politics has no place in the courtroom.”6 These stories involving Justices Thomas and Sotomayor raise (again) the issues of whether judges are, or can be, “impartial” and whether empathy is compatible with judicial reasoning.7 The discussion of whether judges can,8 or should,9 set aside their personal feelings and experiences is

4. Confirmation Hearing on the Nomination of Hon. Sonia Sotomayor, to Be an Associate Justice of the Supreme Court of the United States Before the S. Comm. on the Judiciary, 111th Cong. 84 (2009) (statement of Sen. Sessions, Ranking Member, S. Comm. on the Judiciary) [hereinafter Confirmation Hearing on Judge Sonia Sotomayor]. 5. Id. (statement of Hon. Sonia Sotomayor). 6. Id. at 7 (statement of Sen. Jeff Sessions). 7. John Paul Rollert, Reversed on Appeal: The Uncertain Future of President Obama’s “Empathy Standard,” 120 YALE L.J. ONLINE 89, 90 (2010) (“To the Right, empathy was nothing less than a code word for judicial activism, a dog whistle to the Democratic base that the President would choose judges who would put the counsel of a bleeding heart above the demands of impartial justice.”); Kim McLane Wardlaw, Umpires, Empathy, and Activism: Lessons from Judge Cardozo, 85 NOTRE DAME L. REV. 1629, 1631 (2010) (“No sooner had President Barack Obama uttered the word ‘empathy’ in connection with judicial appointments that the word took on a life of its own. It became a code word for judicial overreaching, and it served as the blank slate onto which politicians painted doomsday scenarios of a judiciary run amok.” (footnote omitted)). 8. Legal commentators are skeptical of this possibility. See, e.g., Donald C. Nugent, Judicial Bias, 42 CLEV. ST. L. REV. 1, 3 (1994) (“[J]udges disserve themselves and the system if they presume that bias and prejudice do not enter the decisionmaking process to some degree.”); Jeffrey M. Shaman, The Impartial Judge: Detachment or Passion?, 45 DEPAUL L. REV. 605, 605 (1996) (“Pure impartiality is an ideal that can never be completely attained. Judges, after all, are human beings who come to the bench with feelings, knowledge, and beliefs that cannot be magically extirpated.”). 9. The notion that judges should be impartial has a long-standing tradition within American government. This philosophy dates back to Alexander Hamilton’s Federalist Paper No. 78 in which he said the judiciary “is the best expedient which can be devised in any government, to secure a steady, upright and impartial administration of the laws.” THE FEDERALIST NO. 78 (Alexander Hamilton). The current rules governing judicial conduct adopted Hamilton’s stance of the impartial judge. This declaration is featured statutorily. See 28 U.S.C. § 455(a) (2006) (“Any justice, judge, or magistrate judge of the United States shall disqualify himself in any proceeding in which his impartiality might reasonably be questioned.”). This proposition is also featured in the American Bar Association’s Model Code of Judicial Conduct. The provision that addresses bias defines it solely in terms of an individual’s personal background. MODEL CODE OF JUDICIAL CONDUCT 2.3(B) (2007) (“A judge shall

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

316 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

not new; most legal academics (at least since the Legal Realists of the early twentieth century), recognize that judicial decisionmaking involves some discretion10 and that decisionmaking does not occur in a vacuum. More recently, Critical Race Theorists have begun to ask if identity plays a role specifically in judicial decisionmaking, and more broadly within the law.11 Although there has been a recent return to legal formalism and original not . . . manifest bias or prejudice, . . . including but not limited to bias [or] prejudice based upon race, sex, gender, religion, national origin, ethnicity, disability, age, sexual orientation, marital status, socioeconomic status, or political affiliation . . . .”). Although judges are prohibited from unethical and inappropriate political activity, see MODEL CODE OF JUDICIAL CONDUCT 4 (2007), there is no explicit text that warns judges about having his or her political attitudes influence their judicial duties. Notwithstanding the Model Code of Judicial Conduct, some legal commentators argue that citizens are ambivalent about having impartial triers of fact. Martha Minow, Stripped Down Like a Runner or Enriched by Experience: Bias and Impartiality of Judges and Jurors, 33 WM. & MARY L. REV. 1201, 1217 (1992) (“We want judges and juries to be objective about the facts and the questions of guilt and innocence but committed to building upon what they already know about the world, human beings, and each person's own implication in the lives of others.”). We believe this apparent ambivalence comes from confusion concerning commentary on the role of experience. Specifically, critics often conflate experience with bias and impartiality. In other words, a judge can bring his or her personal experiences to the bench but they should not undermine the law’s purpose of objective and detached decisionmaking. See Sherrilyn A. Ifill, Judging the Judges: Racial Diversity, Impartiality and Representation on State Trial Courts, 39 B.C. L. REV. 95, 98 n.12, 99 (1997) (stating that the contributions judges from minority backgrounds can make to judicial decisionmaking actually foster impartiality by “diminishing the possibility that one perspective dominates”). 10. See, e.g., VIRGINIA A. HETTINGER, STEFANIE A. LINDQUIST & WENDY L. MARTINEK, JUDGING ON A COLLEGIAL COURT 30 (2006) (commenting that Legal Realists are credited for asserting that judges insert their own beliefs and values into their decisionmaking); Brian Leiter, Rethinking Legal Realism: Toward a Naturalized Jurisprudence, in NATURALIZING JURISPRUDENCE: ESSAYS ON AMERICAN LEGAL REALISM AND NATURALISM IN LEGAL PHILOSOPHY 15, 15–16 (2007) (same). For a more detailed discussion, see infra note 31. 11. See, e.g., PATRICIA HILL COLLINS, BLACK FEMINIST THOUGHT: KNOWLEDGE, CONSCIOUSNESS, AND THE POLITICS OF EMPOWERMENT, at xiv (1990) (discussing “the need to reconcile subjectivity and objectivity in producing scholarship”); PATRICIA J. WILLIAMS, THE ALCHEMY OF RACE AND RIGHTS 6–7 (1991) (illustrating one critical race theorist’s attempt to challenge traditional concepts of “objective” legal scholarship by writing as “black, female and [a] commercial lawyer”); Angela P. Harris, Race and Essentialism in Feminist Legal Thought, 42 STAN. L. REV. 581, 583–84 (1990) (commenting that although most legal thinkers prefer to speak from a position of “objectivity” and “neutrality” rather than “subjectivity” and “bias,” there are theorists who would advance a more self-referential voice); Mari J. Matsuda, Public Response to Racist Speech: Considering the Victim’s Story, in WORDS THAT WOUND: CRITICAL RACE THEORY, ASSAULTIVE SPEECH, AND THE FIRST AMENDMENT 17, 19 (Mari J. Matsuda et al. eds., 1993) (“This description ties law to racism, showing that law is both a product and a promoter of racism.”); Girardeau A. Spann, Pure Politics, in CRITICAL RACE THEORY: THE CUTTING EDGE 21, 24 (Richard Delgado & Jean Stefancic, eds., 2000) (“[E]ven if a justice makes strenuous efforts to compensate for his or her known prejudices, the justice will still be vulnerable to those biases and predispositions that continue to operate at a subconscious level . . . .”).

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 317 intent among some legal theorists,12 fundamentally, legal scholars share the recognition that there is some level of discretion inherent to judicial decisionmaking. These questions have emerged explicitly in the most recent Senate confirmation hearings of U.S. Supreme Court nominees.13 There have been questions about whether the nominee’s social background would affect her ability to adjudicate cases fairly. Concerns central to the rule of law, such as predictability and consistency of the judiciary, have led politicians to question the role of personal identity in judicial decisionmaking. While these political ideology and constitutional theory questions are asked of all nominees, identity characteristics like race, sex, and the like typically are only asked of white female judges and minority judges. No senator asked Judge Roberts whether his social position (wealthy, white man) would affect his ability to fairly decide cases. This criticism may also be part of what motivates advocacy of other measures that would limit the scope of judicial power, such as passing legislation that strips jurisdiction from federal courts14 or overrides the

12. Richard H. Pildes, Forms of Formalism, 66 U. CHI. L. REV. 607, 607 (1999) (discussing the various modes of modern legal formalism); Lawrence B. Solum, The Supreme Court in Bondage: Constitutional Stare Decisis, Legal Formalism, and the Future of Unenumerated Rights, 9 U. PA. J. CONST. L. 155, 159 (2006) (proposing a “neoformalis[t]” theoretical framework that would require judges to follow their legal precedent). 13. This statement is not meant to suggest that judicial activism and political motives have entirely disappeared from confirmation debates. For example, Senator Al Franken of Minnesota asked Sotomayor to define the phrase “judicial activism” because “in political discourse about the role of the judiciary, that is almost the only phrase that is ever used. And I think that there has been an ominous increase in what I consider judicial activism of late . . . .” Confirmation Hearing on Judge Sonia Sotomayor, supra note 4, at 383 (statement of Sen. Franken, Member, S. Comm. on the Judiciary). Another example from the hearing is featured in the statements of Linda Chavez, President and Chairman of the Center for Equal Opportunity. Chavez claimed that Sotomayor “drunk deep from the well of identity politics” and continued to provide examples in which her policy preferences guided her decisions in cases involving race-based government contracts, bilingual education, racial profiling, and affirmative action. Confirmation Hearing on Judge Sonia Sotomayor, supra note 4, at 493 (statement of Linda Chavez, President & Chairman of the Ctr. for Equal Opportunity). We are suggesting that there is an increased interest in the social and personal background of individuals that carries significant implications for decisionmaking and the opinions rendered. In fact, the statement made by Senator Sessions shows an interesting conflation between empathy and politics, see Confirmation Hearing on Judge Sonia Sotomayor, supra note 4, in which being empathetic makes a judge political and one who seeks to inject those preferences in cases. 14. Although there is no express provision regarding jurisdiction-stripping, the Constitution provides some (albeit more elusive) guidance regarding this matter. U.S. CONST. art. III, § 1 (“The judicial Power of the United States, shall be vested in one supreme Court, and in such inferior Courts as the Congress may from time to time ordain and establish.”). Not surprisingly, scholars have written extensively on the interpretation of Article III and the constitutionality of limiting federal jurisdiction.

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

318 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

holdings of the Supreme Court.15 These criticisms are frequently theoretical and are generally based on the untested assumption that a judge’s personal background influences case disposition. For all the political posturing and jurisprudential discussion about whether and to what extent identity is relevant16 for judicial appointment or judicial decisionmaking, there has been a remarkable dearth of empirical data analyzing whether judges of different social backgrounds decide cases differently. This Article is motivated by empirical and normative questions. First, do judges’ personal backgrounds affect their case outcomes?17 We sought

See, e.g., Gerald Gunther, Congressional Power to Curtail Federal Court Jurisdiction: An Opinionated Guide to the Ongoing Debate, 36 STAN. L. REV. 895, 901–16 (1984); John Harrison, The Power of Congress to Limit the Jurisdiction of Federal Courts and the Text of Article III, 64 U. CHI. L. REV. 203, 209–10 (1997); Henry M. Hart, Jr., The Power of Congress to Limit the Jurisdiction of Federal Courts: An Exercise in Dialectic, 66 HARV. L. REV. 1362, 1372–73 (1953); Robert J. Pushaw, Jr., Congressional Power Over Federal Court Jurisdiction: A Defense of the Neo-Federalist Interpretation of Article III, 1997 BYU L. REV. 847, 856–94; Martin H. Redish, Text, Structure, and Common Sense in the Interpretation of Article III, 138 U. PA. L. REV. 1633, 1636–41 (1990). 15. William N. Eskridge, Jr., Overriding Supreme Court Statutory Interpretation Decisions, 101 YALE L.J. 331, 335–53 (1990) (discussing the increasing trend of congressional overrides of Supreme Court decisions); Anna Harvey & Barry Friedman, Pulling Punches: Congressional Constraints on the Supreme Court’s Constitutional Rulings, 1987–2000, 31 LEGIS. STUD. Q. 533, 555 (2006). 16. Discussion of empathy became a part of the Supreme Court confirmation process in recent years primarily because President Obama said he would seek out judges who were empathetic while following the rule of law. He said: I will seek someone who understands that justice isn’t about some abstract legal theory or footnote in a casebook; it is also about how our laws affect the daily realities of people’s lives . . . . I view that quality of empathy, of understanding and identifying with people’s hopes and struggles, as an essential ingredient for arriving as [sic] just decisions and outcomes. President Barack Obama, Remarks on the Retirement of Justice David Souter (May 1, 2009), available at http://www.gpo.gov/fdsys/pkg/DCPD-200900317/pdf/DCPD-200900317.pdf. However, the role of empathy and judging has played a role in confirmation hearings, but, as one commentator points out, Senators only express concern when the nomination is made by a Democratic president. See Glenn Greenwald, Justice Sam Alito on Empathy and Judging, COMMON DREAMS (May 27, 2009), http://www.commondreams.org/view/2009/05/27-13 (comparing the treatment of Republican judicial nominee Alito and Democratic nominee Sotomayor). Greenwald points to a number of instances in which then-nominee, now-Justice Samuel A. Alito somewhat embraces a judicial philosophy that includes empathy and experience: “When I get a case about discrimination, I have to think about people in my own family who suffered discrimination because of their ethnic background or because of religion or because of gender. And I do take that into account.” Id. (citing Confirmation Hearing on the Nomination of Hon. Samuel Alito, to be an Associate Justice of the Supreme Court of the United States Before the S. Comm. on the Judiciary, 111th Cong. (2006) (statement of Hon. Samuel Alito)). 17. We are not seeking to make a causal argument—for example, minority status does not cause a judge to vote a certain way—but rather, our primary research question is whether judges with a particular set of social characteristics systematically evaluate cases differently.

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 319 to test whether federal judges with different identity characteristics make systematically different decisions, using a comprehensive dataset on federal employment civil rights cases. We believe employment civil rights cases give substantive intellectual purchase for this analysis. Aside from being one of the largest categories of civil court filings on the federal docket in the United States,18 litigation by private parties is arguably the most common form in which discrimination claims are adjudicated.19 In addition, these contests often are emotionally charged for the parties. It may well be that if judges have been targets of discrimination, the experience will shape their perceptions about the presence or absence of discrimination. Second, if personal background does influence case outcome, what are the implications for the justice system? We examine these questions empirically, focusing on case disposition at the summary judgment phase.20 Summary judgment is an interesting trial moment to study because it is typically the first21 opportunity and one of the only points in litigation

18. According to United States Courts 2010 caseload statistics, 32,125 private civil rights cases commenced in federal district court or federal circuit court of appeals, which is the largest case category in the federal court system. U.S. COURTS, Federal Judicial Caseload Statistics (Mar. 31, 2010), http://www.uscourts.gov/Viewer.aspx?doc=/uscourts/Statistics/FederalJudicialCaseloadStatistics/2010/t ables/C03Mar10.pdf. The universe of potential claims is even higher if we consider charges filed with the Equal Employment Opportunity Commission (“EEOC”). In 2010, the EEOC received 99,922 charge filings. While this figure reflects multiple claims of discrimination—for example, race and gender discrimination—these data suggest that there are more aggrieved employees who perceive themselves to be victims of discrimination, but who are not captured by court-filing statistics. Charge Statistics: FY 1997 Through FY 2010, U.S. EQUAL EMP’T OPPORTUNITY COMM’N, http://www.eeoc.gov/ eeoc/statistics/enforcement/charges.cfm (last visited Jan. 8, 2012). 19. See SEAN FARHANG, THE LITIGATION STATE: PUBLIC REGULATION AND PRIVATE LAWSUITS IN THE U.S. 94–128 (2010) (discussing the historical trend of reliance on private litigation of civil rights issues). 20. There is variation in summary judgment rates across time, district, and case category, but these measures have been descriptive and have not investigated issues of judicial decisionmaking patterns from this perspective. Stephen B. Burbank, Vanishing Trials and Summary Judgment in Federal Civil Cases: Drifting Toward Bethlehem or Gomorrah?, 1 J. EMPIRICAL LEGAL STUD. 591, 592–93 (2004); Joe S. Cecil et al., A Quarter-Century of Summary Judgment Practice in Six Federal District Courts, 4 J. EMPIRICAL LEGAL STUD. 861, 863 (2007); Marc Galanter, The Vanishing Trial: An Examination of Trials and Related Matters in Federal and State Courts, 1 J. EMPIRICAL LEGAL STUD. 459, 483–84 (2004); William P. McLauchlan, An Empirical Study of the Federal Summary Judgment Rule, 6 J. LEGAL STUD. 427, 435–48 (1977); Paul W. Mollica, Federal Summary Judgment at High Tide, 84 MARQ. L. REV. 141, 164–77 (2000); Theodore Eisenberg & Charlotte Lanvers, Summary Judgment Rates over Time, Across Case Categories, and Across Districts: An Empirical Study of Three Large Federal Districts, CORNELL LAW FACULTY PUBL’N 1, 2 (Aug. 5, 2008), http://scholarship.law.cornell.edu/lsrp_papers/108. 21. FED. R. CIV. P. 56. Although there are several pretrial dispositive motions such as early dismissal, a dismissal in those instances may be the result of a procedural defect and not necessarily

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

320 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

when parties call upon the trial judge to assess the merits of the plaintiff’s claim and determine whether the case—in part or in whole—should proceed to trial.22 In addition, and as importantly, the Federal Rules of Civil Procedure give judges considerable discretion to retain or dispose of cases at this phase, so it provides a unique opportunity to reveal the circumstances under which judges give a nonmoving party (which typically in litigation is the plaintiff) the benefit of the doubt.23 The Article is divided into four parts. Part II discusses the juridical, psychological, and political science literature to assess what we know about how personal background can influence judicial decisionmaking. Part III describes the data and methods we use to address our research question. Our study investigates questions concerning judge-plaintiff minority status using logistic regression to estimate predicted effects of minority status on outcome, controlling for other potentially relevant variables. Prior studies largely examine judicial behavior at a more macro level and do not probe the relationship between plaintiff characteristics and a judge’s personal background.24 By focusing on detailed case characteristics, this study provides evidence about whether the claims asserted or the plaintiffs’ characteristics affect summary judgment rates in civil rights matters. Part IV presents the empirical results. Our data show variation across judges— based on the merits. These defects include lack of subject matter jurisdiction, lack of personal jurisdiction, improper venue, insufficient process, insufficient service of process, and failure to join a party. FED. R. CIV. P. 12(b)(1)–(5), (7). Moreover, one merits-based defense under Rule 12—failure to state a claim upon which relief can be granted pursuant to 12(b)(6)—permits plaintiffs to amend their complaints. 5B CHARLES ALAN WRIGHT & ARTHUR R. MILLER, FEDERAL PRACTICE & PROCEDURE CIV. § 1357 (3d ed. 2010). Dismissal of a case is only permitted if amending the complaint would be futile. See, e.g., Platt Elec. Supply, Inc. v. EOFF Elec., Inc., 522 F.3d 1049, 1060 (9th Cir. 2008) (finding the amendment of complaint futile because the statute of limitations had run); Bonano v. Southside United Hous. Dev. Corp., 363 F. Supp. 2d 559, 562–63 (E.D.N.Y. 2005) (rendering leave to amend futile when proposed new claim cannot withstand motion to dismiss for failure to state a claim). 22. See 10A CHARLES ALAN WRIGHT, ARTHUR R. MILLER & MARY KAY KANE, FEDERAL PRACTICE AND PROCEDURE CIV. § 2712 (3d ed. 2010). 23. FED. R. CIV. P. 56(a) (“The court shall grant summary judgment if the movant shows that there is no genuine dispute as to any material fact . . . .”). See also WRIGHT, MILLER & KANE, supra note 22, § 2728 (The summary judgment standard supplies courts with the discretion to deny a motion even when the moving party could show there is no genuine issue of material fact). However, this discretion was briefly curtailed with the 2007 revision to the Rule 56 in which all references to “shall” were changed to “should” only to be changed back in 2010. WRIGHT, MILLER & KANE, supra note 22, §§ 2711, 2728. See also Steven S. Gensler, Must, Should, Shall, 43 AKRON L. REV. 1139, 1147–49 (2010) (replacing “shall” with “should” allows for some, but not carte blanche, discretion to deny a motion for summary judgment even if there is no genuine issue of material fact). 24. See infra Part II.B.

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 321 namely, that white judges tend to dismiss cases for summary judgment at a higher rate than minority judges. These findings also show no political party effects, suggesting judicial decisionmaking may be less influenced by political ideology than some political scientists suggest and more influenced by experience than we have previously considered. Part V discusses the findings and concludes by considering how personal background may affect judicial decisionmaking. We then propose future research that can unpack these findings in a more systematic way.

II. ASSESSING DISCRIMINATION: THREE PERSPECTIVES ON JUDICIAL DECISIONMAKING Most empirical scholarship on judicial decisionmaking focuses on case outcomes in the appellate and Supreme Courts. Our research diverges from this pattern because we study district court judges and their decisions at the motion for summary judgment phase of trial. Rather than focusing on the mechanisms that drive judicial behavior, this Article contributes to the burgeoning discourse on federal district court judicial decisionmaking.25 Most empirical work on the federal judiciary examines decisionmaking at the appellate and Supreme Court levels in part because the decisions judges render there carry significant precedential power and because they are

25. See generally NANCY SCHERER, SCORING POINTS: POLITICIANS, ACTIVISTS, AND THE LOWER FEDERAL COURT APPOINTMENT PROCESS (2005) (arguing that the appointment of lower court judges has become more and more politicized in the past few decades); Christina L. Boyd, Lee Epstein & Andrew D. Martin, Untangling the Causal Effects of Sex on Judging, 54 AM. J. POL. SCI. 389 (2010) (exploring how gender affects a judge’s individual decisionmaking as well as that of the other judges on the panel); Christina L. Boyd & James F. Spriggs II, An Examination of Strategic Anticipation of Appellate Court Preferences by Federal District Court Judges, 29 WASH. U. J.L. & POL’Y 37, 38 (2009) (discussing the “complicated relationships within the federal judicial hierarchy”); Stephen J. Choi, Mitu Gulati & Eric A. Posner, What Do Federal District Judges Want? An Analysis of Publications, Citations, and Reversals, 26 J.L. ECON. & ORG. 1 (2011) (hypothesizing that federal district judges write opinions hoping to minimize their workload while maximizing their reputations and chances for advancement to higher courts); David A. Hoffman, Alan Izenman & Robert J. Jeffrey Lidicker, Docketology, District Courts, and Doctrine, 85 WASH. U. L. REV. 681 (2007) (discussing empirical data from trial courts that show that judges who write opinions are influenced mainly by procedure rather than by attempts to appeal to a wider audience or advance their own careers); Charles A. Johnson, Law, Politics, and Judicial Decisionmaking: Lower Federal Court Uses of Supreme Court Decisions, 21 LAW & SOC’Y REV. 325 (1987) (testing two competing theoretical models of lower federal court reactions to Supreme Court decisions); Kirk A. Randazzo, Strategic Anticipation and the Hierarchy of Justice in the U.S. District Courts, 36 AM. POL. RES. 669 (2008) (presenting empirical analysis offering support for the claim that district courts are constrained by the anticipated responses of appellate courts); Christina L. Boyd, The Impact of Courts of Appeals on Substantive and Procedural Success in the Federal District Courts (July 14, 2009) (working paper), available at http://ssrn.com/ abstract=1434076 (arguing that appellate courts impact district courts regarding which parties are victorious and which methods of decisionmaking take place).

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

322 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

easily accessible for study via online databases. These studies, however informative, do not capture the additional constraints26 placed on lower court judges; namely, appellate review and the more strict application of the law to cases.27 And, they tell us very little about the day-to-day functioning of district courts. A second contribution of this analysis is that we study judicial decisionmaking midstream in the litigation process,28 while empirically testing whether legal and extralegal variables have any explanatory power in predicting case survival at the summary judgment phase. Given the critical role judges play at summary judgment,29 it is important to

26. Researchers have recognized that judges may be political but their voting must take into account the constraints placed on them, including any institutional constraints placed by the courts such as stare decisis. LEE EPSTEIN & JACK KNIGHT, THE CHOICES JUSTICES MAKE 10 (1998). See also Lee Epstein, Jack Knight & Andrew D. Martin, The Supreme Court as a Strategic National Policymaker, 50 EMORY L.J. 583, 591 (2001) (arguing that justices who vote according to their individual preferences and against the majority’s interests risk congressional reversal and replacement); Max M. Schanzenbach & Emerson H. Tiller, Strategic Judging Under the U.S. Sentencing Guidelines: Positive Political Theory and Evidence, 23 J.L. ECON. & ORG. 24, 24 (2007) (explaining that while judicial decisionmaking is influenced by “judges’ policy preferences,” it is “constrained by the prospect of higher court review”). 27. See infra Part II.A. 28. Most of the empirical work in this area focuses almost exclusively on settlement rates. See, e.g., Theodore Eisenberg & Charlotte Lanvers, What Is the Settlement Rate and Why Should We Care? 6 J. EMPIRICAL LEGAL STUD. 111, 125–46 (2009) (analyzing aggregate settlement rates in two federal districts); Galanter, supra note 20, at 481–84 (2004); Marc Galanter & Mia Cahill, “Most Cases Settle”: Judicial Promotion and Regulation of Settlements, 46 STAN. L. REV. 1339, 1339–40 (1994) (“[C]ourts and policymakers should approach settlement with a more critical eye, distinguishing ‘good’ settlements from less desirable ones”); Jason Scott Johnston & Joel Waldfogel, Does Repeat Play Elicit Cooperation? Evidence from Federal Civil Litigation, 31 J. LEGAL STUD. 39, 40 (2002) (“[S]ettlement rates for some type of cases—such as torts—exceed[] 90 percent.”); Herbert M. Kritzer, Adjudication to Settlement: Shading in the Gray, 70 JUDICATURE 161, 161–62 (1986) (examining judges’ roles in settlement). However, there is significant literature that focuses on various factors that influence whether a case does go to trial. See, e.g., Cecil et al., supra note 20, at 863 (finding that summary judgment motions increased between 1975 and 2000, but that this did not influence outcome); Eisenberg & Lanvers, supra at 129–35 (finding that summary judgment rates varied across case categories and districts, particularly with matters concerning civil rights); Laura Beth Nielsen, Robert L. Nelson & Ryon Lancaster, Individual Justice or Collective Legal Mobilization? Employment Discrimination Litigation in the Post Civil Rights United States, 7 J. EMPIRICAL LEGAL STUD. 175, 184–94 (2010) (finding that only 6 percent of cases actually make it to trial and that the success of plaintiffs depends on whether the individual is represented by an attorney). 29. See Burbank, supra note 20, at 616 (finding that in 2000, judges in the Eastern District of Pennsylvania terminated 4.1 percent of cases by summary judgment); Cecil et al., supra note 20, at 883 (finding that 7.8 percent of cases ended at summary judgment in 2000); Eisenberg & Lanvers, supra note 20, at 13–17 (finding that the difference in summary judgment rates depending on district court was highly statistically significant).

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 323 understand how and why judges reach the decisions they do when evaluating the merits of a case. In the context of employment civil rights cases, the determination of discrimination—and more specifically, its presence or absence—depends largely on the judge’s perception of an employer’s actions against a plaintiff. Because illegal discrimination can operate through implicit bias rather than overt harassment,30 the facts and evidence in these cases often are ambiguous and open to interpretation. Although our analysis cannot capture the actual decision-making processes of a judge, we draw from a number of theories which, at the macro level, provide descriptive and explanatory power to our data. Three perspectives31 are useful for considering judicial decisionmaking: (1) the empathethic perspective, a seldom-discussed view that suggests judges decide cases based on their lived experiences and interactions with macro- social factors that contain systemic social barriers to people of color and women; (2) the legal perspective or the legal model of decisionmaking, which takes the position that judges mechanistically apply the law to facts; and (3) the political perspective or the empirical explanation, which contends that judges tend to render opinions based on political, ideological, or strategic preferences.

30. See, e.g., Marianne Bertrand, Dolly Chugh & Sendhil Mullainathan, Implicit Discrimination, 95 AM. ECON. REV. 94, 94 (2005) (“[S]ometimes . . . discrimination may be unintentional and outside of the discriminator’s awareness”); John F. Dovidio, Kerry Kawakami & Samuel L. Gaertner, Implicit and Explicit Prejudice and Interracial Interaction, 82 J. PERSONALITY & SOC. PSYCHOL. 62, 65–66 (2002) (analyzing the effect of implicit bias on interracial interactions); Christine Jolls & Cass R. Sunstein, The Law of Implicit Bias, 94 CALIF. L. REV. 969, 971–73 (2006) (explaining that implicit bias demonstrated in tests is predictive of implicit bias in actual behavior); Jerry Kang, Trojan Horses of Race, 118 HARV. L. REV. 1489, 1514 (2005) (“There is now persuasive evidence that implicit bias against a social category . . . predicts disparate behavior toward individuals mapped to that category.”); Linda Hamilton Krieger & Susan T. Fiske, Behavioral Realism in Employment Discrimination Law: Implicit Bias and Disparate Treatment, 94 CALIF. L. REV. 997, 1004 (2006) (“[M]any scholars have drawn on advances in the empirical social sciences to demonstrate that what the law refers to as ‘intentional discrimination’ can just as easily result from the uncontrolled application of implicit, unconscious, or automatic stereotypes and other subtle ingroup preferences as from the operation of conscious discriminatory designs.”); Lincoln Quillian, New Approaches to Understanding Racial Prejudice and Discrimination, 32 ANN. REV. SOC. 299, 314–20 (2006) (“[A]n implicit attitude is an attitude that can be activated without conscious awareness and, when so triggered, influences judgments and actions”). 31. This is not a full and exhaustive list of theories, or perspectives, on judicial decisionmaking. For example, scholars have offered pragmatic, economic, strategic, and organizational theories of judicial behavior. We believe that with the specific case of civil rights claims, legal, political, and ideological empathy best describe how judges go about assessing these types of claims. For a review of theories not presented in this article, see RICHARD A. POSNER, HOW JUDGES THINK 19–56 (2008).

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

324 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

A. THE EMPATHETIC PERSPECTIVE Although judges may decide cases mechanically or politically, the empathetic perspective suggests that judges do not completely abandon their experiences when deciding cases. There are instances in which an individual’s personal background influences how she perceives social situations. As Justice Sotomayor suggests, living in the United States as a Latina tends to include different life experiences, resultant attitudes, and world views. Perhaps as proponents and critics suggest, judges acknowledge the importance of law, yet add life experience resulting from shared kinds of experiences.32 Given this, we contend “empathy” (by which we mean the world views of judges that are formed, at least in part, by the social location they occupy), plays a crucial role in cases that are more emotionally charged and morally consequential, as viewed differently from controversies that are not (e.g., federalism and takings clause matters).33 We believe that empathy and politically motivated decisionmaking are not synonymous, but scholars are quick to conflate the two. The empathethic perspective also comes from the Critical Race Theory tradition, which suggests that a diverse judiciary greatly shapes judicial decisionmaking, legal analysis, and, by extension, the law itself.34 Specifically, judges who hail from different social or cultural backgrounds may provide a more nuanced understanding of facts, evidence, and credibility determinations than judges who lack such experience. While the jurisprudential discourse in this area is significant,35 there has been little empirical work to move beyond the anecdotal or doctrinal accounts. To support this claim, we draw from the extensive literature in psychology that examines the influence of individual background on how

32. See, e.g., Richard L. Allen, Michael C. Dawson & Ronald E. Brown, A Schema-Based Approach to Modeling an African-American Racial Belief System, 83 AM. POL. SCI. REV. 421, 435 (1989) (finding that a black individual’s socioeconomic status, religion, and media structure influence that individual’s cognitive processing of information). 33. This is empirically consistent with research that shows judges deciding cases less ideologically if the subject matter of the case is less political or morally charged. See Cass R. Sunstein, David Schkade & Lisa Michelle Ellman, Ideological Voting on Federal Courts of Appeals: A Preliminary Investigation, 90 VA. L. REV. 301, 325–30 (2004). 34. See Edward M. Chen, The Judiciary, Diversity, and Justice for All, 91 CALIF. L. REV. 1109, 1117–19 (2003) (discussing how judges’ diverse backgrounds inform their judicial decisionmaking); Sherrilyn A. Ifill, Racial Diversity on the Bench: Beyond Role Models and Public Confidence, 57 WASH. & LEE L. REV. 405, 449–57 (2000) (connecting judicial diversity with diverse judicial decisionmaking). 35. See supra note 11.

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 325 individuals respond to questions about the presence or absence of discrimination. While psychologists define empathy in a variety of ways, the general definition suggests that empathy is an emotion that becomes activated by imagining or observing another person’s particular situation.36 Often (but not always), by observing or imagining the other person’s emotions in a particular context, an empathetic response is generated because of “perspective taking.”37 In other words, empathy is a stand-alone emotion, but there can be an element of vicarious emotion as well. Empathy is different from sympathy, in which a person merely imagines the experiences of another. In contrast, a person experiencing empathy is aware and actually sensitive to the state or condition of another.38 Empathy has been shown to have a substantial impact on decisionmaking. For example, previous research shows that individuals who empathize with victims have a greater willingness to help that individual.39 Whether a person is more or less empathetic depends on common membership in a social category or group.40 Often these

36. E.g., C. DANIEL BATSON, THE ALTRUISM QUESTION: TOWARD A SOCIAL-PSYCHOLOGICAL ANSWER 86 (1991) (defining empathy as a “set of congruent vicarious emotions, those that are more other-focused than self-focused, including feelings of sympathy, compassion, tenderness, and the like”); MARK H. DAVIS, EMPATHY: A SOCIAL PSYCHOLOGICAL APPROACH 12 (1996) (describing empathy as an inner imitation when observing another person or object); Martin L. Hoffman, Is Altruism Part of Human Nature?, 40 J. PERSONALITY & SOC. PSYCHOL. 121, 128 (1981) (defining empathy as a “vicarious affective response to others: that is, an affective response appropriate to someone else’s situation rather than one’s own”). 37. DAVIS, supra note 36, at 14. In our research, an example of empathy is a judge who identifies with the feelings of a victim of discrimination by putting himself or herself in the place of the victim as a way to understand the emotional evidence that is typically presented in employment civil rights matters. 38. Id. at 3–5. 39. See, e.g., C. Daniel Batson, et al., An Additional Antecedent of Empathic Concern: Valuing the Welfare of the Person in Need, 93 J. PERSONALITY & SOC. PSYCHOL. 65, 70 (2007) (finding that when an individual’s welfare was valued more highly, empathic concern for the individual was also higher, and this increased empathy was “associated with increased helping”); C. Daniel Batson et al., Immorality from Empathy-Induced Altruism: When Compassion and Justice Conflict, 68 J. PERSONALITY & SOC. PSYCHOL. 1042, 1052 (1995) (“Knowing and feeling empathy for the person in need . . . led many participants to forsake justice in the interest of benefiting the person for whom they felt empathy . . . .”); Christine A. Smith & Irene H. Frieze, Examining Rape Empathy from the Perspective of the Victim and the Assailant, 33 J. APPLIED SOC. PSYCHOL. 476, 476 (2003) (“Empathy with a rape victim or perpetrator might influence perceptions, judgments, and blame of a rape victim or perpetrator.”). 40. C. Daniel Batson et al., Empathy, Attitudes, and Action: Can Feeling for a Member of a Stigmatized Group Motivate One to Help the Group?, 28 PERSONALITY & SOC. PSYCHOL. BULL. 1656, 1657 (2002) (“There is considerable evidence that feeling increased empathy for a person in need increases the readiness to help that person.” (citation omitted)); Mark Tarrant & Aimee Hadert, Empathic Experience and Attitudes Toward Stigmatized Groups: Evidence for Attitude Generalization,

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

326 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

differences are based on race41 and gender42 categories. Since empathy involves understanding the emotional states of other people, minority judges may be more open to the perspectives of members of subordinated groups, which posit that discrimination itself exists and remains a social problem. This proposition is consistent with research that suggests that individuals are more or less likely to perceive the presence of discrimination based on their identification with a stigmatized social group.43 Group identification theory suggests that the more an individual identifies with a social group, the greater the likelihood that he or she will interpret interpersonal interactions in terms of group-based attitudes and beliefs.44 Consistent with this theory, a number of studies reported strong correlations between group identification and perceptions of discrimination among members of devalued groups.45 Perceptions of discrimination between people who more or less identify with a social group are most pronounced in ambiguous situations.46 For example, when there are ambiguous prejudice cues, women who highly identified with gender as their social group were significantly more likely to attribute a negative evaluation from a male as sex discrimination.47 By

40 J. APPLIED SOC. PSYCHOL. 1635, 1652 (2010). 41. James D. Johnson et al., Rodney King and O.J. Revisited: The Impact of Race and Defendant Empathy Induction on Judicial Decisions, 32 J. APPLIED SOC. PSYCHOL. 1208, 1208 (2002). 42. Smith & Frieze, supra note 39, at 493. 43. Cheryl R. Kaiser & Clara L. Wilkins, Group Identification and Prejudice: Theoretical and Empirical Advances and Implications, 66 J. SOC. ISSUES 461, 462–63 (2010); Brenda Major, Wendy J. Quinton & Shannon K. McCoy, Antecedents and Consequences of Attributions to Discrimination: Theoretical and Empirical Advances, in 34 ADVANCES IN EXPERIMENTAL SOCIAL PSYCHOLOGY 251, 258–59 (Mark P. Zanna ed., 2002). 44. Major, Quinton & McCoy, supra note 43, at 280, 308. 45. Brenda Major et al. Perceiving Personal Discrimination: The Role of Group Status and Legitimizing Ideology, 82 J. PERSONALITY & SOC. PSYCHOL. 269, 273 (2002); Robert M. Sellers & J. Nicole Shelton, The Role of Racial Identity in Perceived Racial Discrimination, 84 J. PERSONALITY & SOC. PSYCHOL. 1079, 1088–89 (2003). See also Nyla R. Branscombe, Michael T. Schmitt & Richard D. Harvey, Perceiving Pervasive Discrimination Among African Americans: Implications for Group Identification and Well-Being, 77 J. PERSONALITY & SOC. PSYCHOL. 135 (1999) (finding that blacks who perceive prejudice more closely align themselves with the minority group); Michael T. Schmitt et al., Perceiving Discrimination Against One’s Gender Group Has Different Implications for Well-Being in Women and Men, 28 PERSONALITY & SOC. PSYCHOL. BULL. 197, 198–99 (2002) (“[D]isadvantaged groups are likely to perceive prejudice against them as occurring across a wider variety of contexts than do privileged groups.”). 46. E.g., Brenda Major, Wendy J. Quinton & Toni Schmader, Attributions to Discrimination and Self-Esteem: Impact of Group Identification and Situational Ambiguity, 39 J. EXPERIMENTAL SOC. PSYCHOL. 220, 228–29 (2003). 47. Id. at 228.

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 327 contrast, women with low gender identification did not attribute ambiguous interpersonal cues to sex discrimination.48 These findings suggest that in ambiguous circumstances, individuals who highly identify with their groups are vigilant for discrimination, whereas individuals with low group identification did not perceive discrimination. These processes may be particularly important in the context of employment discrimination cases, where indicators of discrimination in the workplace may be subtle.

B. THE LIBERAL LEGAL PERSPECTIVE A second theory of judicial decisionmaking is the liberal legal model, which eschews “empathy” or “life experience” in favor of a model whereby judges act consistently to apply facts to law.49 The liberal legal model is the most traditional (albeit aspirational) type of judicial decisionmaking. This theory suggests that judges decide cases by determining the relevant legal rule or principle and then apply it mechanistically to the facts of a case or controversy.50 This view posits that judges interpret the law “in light of the plain meaning of statutes and the Constitution, the intent of the framers, . . . precedent,” and a balancing of societal interests.51 Judges do not consider their personal or political views when adjudicating cases.52 This model has more explanatory power at the trial court level than at the appellate court level because of the additional constraints placed on trial court judges.53 Not only can their decisions be reviewed by two levels of appellate courts, but also trial judges do not have control of their dockets to selectively accept cases to further their political objectives.54 Even though

48. Id. 49. Robert N. Wilentz, Judicial Legitimacy—Finding the Law, 8 SETON HALL LEGIS. J. 221, 228 (1985). 50. Frederick Schauer, Formalism, 97 YALE L.J. 509, 510 (1988); Wilentz, supra note 49, at 228 (“The [mechanical] approach posits that all a judge need do is apply predetermined rules to the facts of a case.”). But see Ernest J. Weinrib, The Jurisprudence of Legal Formalism, 16 HARV. J.L. & PUB. POL’Y 583, 583 (1993) (rejecting the argument that legal formalism is “the mechanical application of determinate rules”). 51. JEFFREY A. SEGAL & HAROLD J. SPAETH, THE SUPREME COURT AND THE ATTITUDINAL MODEL REVISITED 48 (2002). 52. Under this model of judicial behavior, judges serve merely as the mechanism to discharge the relevant law to a case or controversy. POSNER, supra note 31, at 42 (“Since the rules are given and have only to be applied . . . the legalist judge is uninterested professionally in the social sciences, philosophy, or any other possible sources of guidance for making policy judgments, because he is not engaged, or at least he thinks he is not engaged, in making such judgments.”). 53. Denise M. Keele et al., An Analysis of Ideological Effects in Published Versus Unpublished Judicial Opinions, 6 J. EMPIRICAL LEGAL STUD. 213, 233 (2009). 54. Id.

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

328 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

appellate courts have similar restraints—namely, a higher reviewing court —rarely are cases granted certiorari and taken up to the Supreme Court.55 However, for many legal thinkers, such as legal realists and critical legal scholars, the notion that decisionmaking can and does follow a mechanical pattern of legal reasoning is questionable. Because these scholars believe the law is indeterminate and open to interpretation, they believe there is little possibility for a judge to mechanistically apply the law.56 “In every legal system,” according to H.L.A. Hart, “a large and important field is left open for the exercise of discretion by courts and other officials in rendering initially vague standards determinate, in resolving the uncertainties of statutes, or in developing and qualifying rules only broadly communicated by authoritative precedents.”57

C. THE POLITICAL PERSPECTIVE Given that the liberal legal model serves as more of an ideal type of decisionmaking, scholars recognized and sought to capture what judicial behavior really looks like using empirical attitudinal models.58 This perspective suggests that judges have political preferences and seek to embed them in the opinions they render.59 In other words, legal rules express the preferences of judges, and how they interpret the rules reflect

55. Id. 56. Brian Leiter, American Legal Realism, in THE BLACKWELL GUIDE TO THE PHILOSOPHY OF LAW AND LEGAL THEORY 50, 51 (Martin P. Golding & William A. Edmundson eds., 2005). This critique comes from multiple schools of thought. For legal realists, the law is indeterminate not only because the law on the books (statutes, case law, etc.) is a system of contradictory rules, but also because decision makers must look at extralegal considerations such as evidence, which harbors no standardized outcome. Id. See, e.g., JEROME FRANK, LAW AND THE MODERN MIND 46–52 (1930) (explaining legal realism); Karl N. Llewellyn, A Realistic Jurisprudence—The Next Step, 30 COLUM. L. REV. 431, 464–65 (1930); Roscoe Pound, The Call for a Realist Jurisprudence, 44 HARV. L. REV. 697, 710 (1931) (recognizing “a plurality of elements in all situations and . . . the possibility of dealing with human relations in more than one way”). A more extreme interpretation from the critical legal studies movement views the law’s indeterminacy as a way for judges to embed political agenda into the decision. See DAVID KAIRYS, INTRODUCTION TO THE POLITICS OF LAW 1, 4 (1983) (“[T]he law usually embraces and legitimizes many or all of the conflicting values and interests involved in controversial issues . . . . Judges then make choices, and those choices are most fundamentally value based, or political.”); SEGAL & SPAETH, supra note 51, at 87–88 (“[T]he legal realists argued that lawmaking inhered in judging.”). 57. H.L.A. HART, THE CONCEPT OF LAW 136 (2d. ed. 1994). 58. See generally SEGAL & SPAETH, supra note 51 (discussing the use of the attidudinal model to explain judicial decisionmaking). 59. See id. at 86 (“Simply put, Rehnquist votes the way he does because he is extremely conservative; Marshall voted the way he did because he was extremely liberal.”).

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 329 their social, economic, and political outlook. The attitudinal model shows empirical evidence of voting driven by political ideology (although most studies have relatively low explanatory power, meaning that political ideology can explain a relatively small percentage of the variation observed). Based on the assumption that votes can be directly correlated to judicial attitudes by coding and tracking votes, not only do studies find that judges display attitudinal patterns by consistently favoring conservative or liberal laws,60 but there is also support for the proposition that precedent does not constrain judges from voting based on their policy preferences.61 Despite the empirical power of many studies working within the attitudinal model, serious critiques have been made against this model. Social science research on judicial behavior may accurately measure what (some) judges do, but it is speculative about the judges’ motives and incentives for acting ideologically or strategically, particularly at the trial court level.62 While political and social considerations undoubtedly play a role throughout the realm of judicial decisionmaking,63 the likelihood that judges behave consistently within attitudinal categories depends on institutional incentives and disincentives, including but not limited to institutional constraints,64 such as standard of review, appellate court

60. See, e.g., id. at 323 (finding a very strong correlation of 0.76 between ideological values and voting); Jeffrey A. Segal & Albert D. Cover, Ideological Values and the Votes of the U.S. Supreme Court Justices, 83 AM. POL. SCI. REV. 557, 561–62 (1989) (finding a very strong correlation of 0.80); Sunstein, Schkade, & Ellman, supra note 33, at 314–15 (finding a 13 percent difference in overall voting patterns between democrat and republican circuit court judges, even before considering panel effects). 61. SEGAL & SPAETH, supra note 51, at 111. However, the one caveat to this proposition is that most of these studies examine judges who reside in courts of appellate jurisdiction. 62. See C.K. ROWLAND & ROBERT A. CARP, POLITICS AND JUDGMENT IN FEDERAL DISTRICT COURTS 17 (1996) (“[T]he axiomatic interpretation of aggregate outcomes as the product of extralegal policy preferences assumes away important questions about the cognitive process by which legalistic procedures produce politicized outcomes . . . .”). 63. Numerous studies have found links between lower court judges’ policy preferences, as measured by the judges’ prior party affiliations, the party of the appointing president, and the judges’ voting decisions. See, e.g., id. at 24–57; C.K. Rowland & Robert A. Carp, A Longitudinal Study of Party Effects on Federal District Court Policy Propensities, 24 AM. J. POL. SCI. 291, 300 (1980). However, these studies did not sufficiently account for the possibility that even policy-maximizing lower court judges will find their behavior constrained by the threat of reversal on appeal. 64. Going beyond the attitudinal model, this theory contends that preferences cannot be viewed in isolation, but should be considered in the way they interact with the collegial institutional context, including stare decisis and other judges. As such, judicial decisionmaking cannot be explained without accounting for these interactive effects and the strategic behavior they occasion. Put simply, judges behave strategically to achieve those goals, given the institutional context, structures, and constraints. EPSTEIN & KNIGHT, supra note 26, at 10–11.

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

330 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

review, and the law itself. On the other hand, federal district court judges may have ambition for higher office (for example, the U.S. Court of Appeals or the Supreme Court), and thus may have significant incentives to placate others who aid in the nomination of judges. However, we also believe that there are instances in the litigation process in which trial judges have opportunities to legitimately exercise discretion, conditions not all that different from those upon which the attitudinal model is premised.

III. AN EMPIRICAL ANALYSIS OF FEDERAL DISTRICT COURT JUDGES AND DISCRIMINATION This analysis investigates whether and to what extent the three perspectives play a role in explaining how federal district court judges decide employment civil rights cases. Although scholars suggest that judges tend to decide cases politically, we hypothesize that judges’ decisions may vary according to their personal characteristics, demographics, and life experiences (perhaps without intention). Social- psychological research demonstrates that if there is ingroup bias, it is most likely to occur when race or gender are “primes,”65 making employment civil rights cases a theoretically rich location for empirical study of these questions. Part III.A describes the data collection process; Part III.B explains our statistical models and our predictions.

A. DATA COLLECTION The data analyzed in this Article comes from a large random sample of federal district court filings of employment civil rights disputes filed between 1988 and 2003. The sample comprises employment civil rights

65. In social psychology, “primes” refer to cues that are introduced to individuals and can produce and reproduce stereotypes such as race and gender. Anthony G. Greenwald, Debbie E. McGhee & Jordan L. K. Schwartz, Measuring Individual Differences in Implicit Cognition: The Implicit Association Test, 74 J. PERSONALITY & SOC. PSYCHOL. 1464, 1477 (1998). These primes can often influence the interpretation of events. For example, individuals primed with racial stereotypes are more likely to find a black criminal defendant guilty than a white criminal defendant. Bernd Wittenbrink & Julia R. Henly, Creating Social Reality: Informational Social Influence and the Content of Stereotypic Beliefs, 22 PERSONALITY & SOC. PSYCHOL. BULL. 598, 603 (1996). These primes are implicit and seemingly automatic responses to different forms of cues such as word or visual images. The most famous study is the Implicit Association Test where participants are presented with visual images or words that they have to sort based on association. For example “Black” and “White” would be on the same screen as “safe” or “dangerous.” Research indicates that these primes—although unconscious— are very racialized and reveal stereotypic associations to individuals based on their race. See generally Greenwald, McGhee & Schwartz, supra (discussing the usefulness of implicit association tests based on the findings of three experiments).

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 331 cases filed in seven regionally diverse districts: Atlanta, Chicago, Dallas, New Orleans, New York City, Philadelphia, and San Francisco, which account for the districts in which roughly 20 percent of all federal district court cases are filed.66 Researchers traveled to these district courts and federal records centers, where cases are archived, to ensure a random sample of case filings (rather than using databases like Westlaw, LexisNexis, or Public Access to Court Electronic Records (“PACER”), which contain only some of these cases).67 Our elaborate coding form asked over one hundred questions concerning case characteristics as well as outcome. For this paper, however, we will primarily be relying on the variables for the outcome of summary judgment motions and the race of the presiding judge. We cross-referenced the judge and district court to obtain information about judges using the Federal Judicial Center’s biographical directory, including age, years on the bench, race, gender, and the President who appointed the judge to the bench.68 Given our research question, we retained only those cases in which the judge presiding over the case was appointed by the President of the United States, to control for possible political party biases and also because magistrate judges are appointed by a majority vote of federal district judges of a particular district and are not appointed for life.69

66. The research design replicated and greatly expanded on the earlier research project of John Donohue and Peter Siegelman, who investigated the changing nature of federal employment discrimination law from the 1970s through 1990. See generally John J. Donohue III & Peter Siegelman, The Changing Nature of Employment Discrimination Litigation, 43 STAN. L. REV. 983 (1991) (analyzing the effects of the business cycle on employment discrimination litigation); John J. Donohue III & Peter Siegelman, Law and Macroeconomics: Employment Discrimination Litigation Over the Business Cycle, 66 S. CAL. L. REV. 709 (1993) (same); Peter Siegelman & John J. Donohue III, The Selection of Employment Discrimination Disputes for Litigation: Using Business Cycle Effects to Test the Priest-Klein Hypothesis, 24 J. LEGAL STUD. 427 (1995) (same). Our data were a random sample of federal district court cases and 100 interviews with plaintiffs, defendants, and their attorneys. Nielsen, Nelson & Lancaster, supra note 28, at 181. 67. Research on whether a judge’s social background truly plays a role in case outcomes has yielded mixed results within social science. Some scholars believe that any apparent differences stem from which cases are published, suggesting that judges recognizing that their opinions will be cited style them differently. See Keele et al., supra note 53, at 218–19. 68. For example, only twenty-six federal district court judges are black females, or 4.4 percent. Federal Judicial Center, Biographical Directory of Judges, http://www.fjc.gov/history/home.nsf/page/ judges.html (last visited Jan. 16, 2012). 69. Our decision to remove magistrate judges from our dataset was a conscious one. Federal district court judges are nominated by the President, see U.S. CONST. art. II, § 2, and confirmed by the Senate, see U.S. CONST. art. III, § 1. Magistrate judges are appointed by a majority vote of federal district judges in a particular district. 28 U.S.C. § 631(a) (2006). Given this distinction, we believed that there is no ideology measure for these individuals.

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

332 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

Although our dataset features a number of measures that capture various points within litigation—pretrial dismissal under Rule 12, early and late settlement, and trial—we limited our analysis for this Article to matters in which a judge decided a motion for summary judgment. This decision has practical and theoretical justifications. Practically, our data show that few cases reach trial (100 of our 1672 observations), and within that group only a handful were bench trials,70 which limits the universe of cases we can rely on to evaluate differences across judge and plaintiff backgrounds. Summary judgment is a good choice for analyzing judicial decisionmaking because it is a merits-based motion, but a stage in litigation where the judge possesses considerable discretion. Rule 56 of the Federal Rules of Civil Procedure authorizes a federal district court judge to authorize judgment in whole or in part in a case when the record establishes that a party is entitled to that judgment as a matter of law.71 Put simply, this rule permits a judge to look at the merits of the case and the presence or absence of “genuine dispute as to any material fact.”72 Although the Rules Advisory Committee modified Rule 56, judges still have some discretionary power to retain cases that do not satisfy the standard of the rule. In 2007, Rule 56 was briefly amended so that in Rules 56(c), (d), and (e), all references to when summary judgment “shall” be entered were changed to “should.”73 However, in 2010, the Advisory Committee reinstated “shall” for these provisions.74 This marked an acknowledgment that a judge can “deny summary judgment even when it appears that there is no genuine issue of material fact.”75 In other words, even when a plaintiff does not produce sufficient evidence to survive a summary judgment motion, a judge can choose to retain a case even if the nonmoving party does not satisfy the burden of production or persuasion. The dependent variable we analyzed was whether a case survived, or was disposed of, in summary judgment. We coded this variable dichotomously—“0” representing a case that survived in whole or in part and “1” representing a case that was dismissed in its entirety. Because our theoretical questions focused on whether a judge perceived the plaintiff as

70. Nielsen, Nelson & Lancaster, supra note 28, at 187. 71. FED. R. CIV. P. 56. 72. FED. R. CIV. P. 56(a). 73. FED. R. CIV. P. 56 advisory committee’s note on 2007 amendments. 74. FED. R. CIV. P. 56 advisory committee’s note on 2010 amendments. 75. See WRIGHT, MILLER & KANE, supra note 22.

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 333 the victim of illegal discrimination, this coding scheme is the most accurate.76 In addition, as a practical matter, some plaintiffs may assert multiple claims not knowing what form of discrimination was operating in the workplace (for example, gender versus race discrimination). We analyzed 522 motions for summary judgment decided by 431 federal district court judges determining the fate of 520 plaintiffs. The demographic characteristics of the judges are described and discussed below, but the proportion of minority judges in our sample is lower than the representation of minority judges on the federal bench generally.

B. STATISTICAL MODELS AND HYPOTHESES To assess the relationship between judges’ demographic characteristics and summary judgment outcomes, we analyze the effect of various case characteristics on the probability that a case is dismissed at summary judgment. Because the dependent (or outcome) variable is binary, we specify three different logistic regression models.77

Model 1: ln[P/(1-P)] = β0 + β1Xi_case + ei

Model 2: ln[P/(1-P)] = β0 + β1Xi_judge + ei

Model 3: ln[P/(1-P)] = β0 + β1Xi_judge + β2Xi_case + ei In each model, ln[P/(1-P)] represents the outcome of case i at summary judgment.78 As noted above, all the cases in the dataset involve a determination at this phase of litigation and are coded as “1” if the case is dismissed and “0” if the case survives in whole or in part.79 The first model includes only plaintiff and case characteristics. This model is meant to consider whether case disposition at summary judgment depends not on the judge presiding over the matter, but rather on elements such as the type of claim asserted (sex versus race discrimination); the plaintiff’s

76. Our data features cases of plaintiff-initiated and cross-motions for summary judgment (albeit very few). To account for this, we coded our outcome variable to indicate whether the plaintiff’s motion was dismissed or retained. 77. For a more detailed discussion about logistic regression, see generally, FRED C. PAMPEL, LOGISTIC REGRESSION: A PRIMER (Michael S. Lewis-Beck ed., 2000) (addressing the logic of logistic regression, the interpretation of results, estimation procedure, and probit versus logit analysis). 78. The outcome is the probability of presence of a particular outcome. This figure is represented as logged odds. PAMPEL, supra note 77, at 10. In this case, we are interested in the probability that a case would be dismissed at summary judgment. 79. We coded our variable in this fashion because it is typical for plaintiffs to include multiple claims and plead in the alternative, which is permitted under the Federal Rules of Civil Procedure. See FED. R. CIV. P. 8(d)(2)–(3).

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

334 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

characteristics such as occupation and tenure on the job; and the plaintiff’s representational status. This model is similar to a model in our previous research80 but eliminates a number of variables that were insignificant.81 We constructed this model to test the legal model because it takes into account case-specific variables that would influence the outcome at summary judgment. In other words, if we find this model to best describe our data, a judge’s political or personal preferences do not play a role in case disposition. The second model contains variables for judges’ demographic characteristics. The variable β1 is a binary variable coded as “2” if the judge is identified by the Federal Center as a racial minority and “1” if the judge is white.82 We included a number of covariates concerning a judge’s background that have previously been found to influence case outcomes including: race,83 gender,84 and the political party of the appointing

80. See generally Nielsen, Nelson & Lancaster, supra note 28. 81. We removed two variables that we constructed and used in previous work. Although these variables were meant to measure the “quality” of the case (and by extension the “strength” of the case), they were found to be statistically insignificant and somewhat imprecise. One variable was the index of legal effort. See Nielsen, Nelson & Lancaster, supra note 28, at 182. This measure ranged from 0 to 3 and points were assigned if a case file contained depositions, expert testimony, or statistical evidence at summary judgment. Id. From a doctrinal perspective, the quantity of evidence proffered does not and should not influence the ruling of a summary judgment motion. See WRIGHT & MILLER, supra note 21. The second variable we eliminated from our analysis was the treatment of the charge that preceded the lawsuit by the EEOC, as well as the EEOC priority code for cases filed after 1995. See Nielsen, Nelson & Lancaster, supra note 28, at 182. The EEOC established a priority case handling process in which an EEOC complaint processing specialist assigned each case an A, B, or C priority code. Id. at 191. The specialist also decided if further investigation will “probably” result in a cause finding (an “A” case), will “likely” result in a cause finding (a “B” case), or has “uncertain merit” (a “C” case). Id. This variable not only was insignificant, but research also reveals that the assignment of priority codes is somewhat inconsistent. See id. at 191–92; C. Elizabeth Hirsh, Settling for Less?: Organizational Determinants of Discrimination-Charge Outcomes, 42 LAW & SOC’Y REV. 239 (2008) (discussing the EEOC “Charge Handling Priority System” and finding that “[t]he intensity of investigations varies considerably across cases”). 82. Our coding scheme for this variable was a conscious one. First, we had few observations involving different ethnic and racial minority judges that when we attempted to create a variable for each group based on white, black, Asian, Hispanic and other, many of these variables dropped out of our model. Theoretically, our decision to make this a binary variable (minority/nonminority) based on literature suggesting that minority status leads to increased sensitivity to situations involving discrimination. 83. Pat K. Chew & Robert E. Kelley, Myth of the Color-Blind Judge: An Empirical Analysis of Racial Harassment Cases, 86 WASH. U. L. REV. 1117, 1156–63 (2009) (finding that both the race and political affiliation of judges impacts the outcome in discrimination cases); Darrell Steffensmeier & Chester L. Britt, Judges’ Race and Judicial Decisionmaking: Do Black Judges Sentence Differently?, 82 SOC. SCI. Q. 749, 761 (2001) (“[B]lack judges are somewhat more likely to incarcerate defendants

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 335

President.85 This model presumes that a judge’s personal background will correlate with the outcome of summary judgment. The third model is a combination of the first two. This model takes into account a judge’s personal background while controlling for variables that may affect summary judgment outcome, particularly plaintiff and case characteristics. This model not only controls for all judge-specific and case- specific variables, but it also accounts for a type of judicial behavior in which legal, political, and empathetic perspectives are operating. If judges’ demographic characteristics have an effect on summary judgment motion outcomes, we should see variation across judges according to minority status or gender, even holding constant the political party of the judge. We also might expect to see variation across plaintiff characteristics. If judicial variables have negative coefficients in models 1 and 3, it means that minority judges are less likely to dismiss their cases at summary judgment. If the political theory of judicial decisionmaking is operating, we would expect to see judges appointed by Democratic presidents to be less likely to dismiss cases. If psychological theories of empathy and identity group theory are operating, the plaintiff’s personal background in tandem with a judge’s personal background will predict and hence are more punitive in their sentencing decisions.”); Thomas M. Uhlman, Black Elite Decisionmaking: The Case of Trial Judges, 22 AM. J. POL. SCI. 884, 891 (1978) (finding little difference between black and white judges “in determining guilt and assigning punishments,” but noting that “[r]ace-related defendant disparities appear in sentencing”). 84. See, e.g., Boyd, Epstein & Martin, supra note 25, at 406 (“observ[ing] consistent and statistically significant individual and panel effects in sex discrimination disputes”); Sue Davis, Susan Haire & Donald Songer, Voting Behavior and Gender on the U.S. Courts of Appeals, 77 JUDICATURE 129 (1993) (finding statistically significant differences between men and women judges in their support for employment discrimination claimants and for criminal defendants in search and seizure cases); John Gruhl, Cassa Spohn & Susan Welch, Women as Policymakers: The Case of Trial Judges, 25 AM. J. POL. SCI. 308, 320 (1981) (finding that “[w]omen are about twice as likely to sentence females to prison as men are”); Jennifer L. Peresie, Note, Female Judges Matter: Gender and Collegial Decisionmaking in the Federal Appellate Courts, 114 YALE L.J. 1759, 1761 (2005) (finding that plaintiffs in sexual harassment and sex discrimination cases “were twice as likely to prevail when a female judge was on the bench”). 85. See, e.g., SEGAL & SPAETH, supra note 51, at 217–22 (discussing presidential impact on the Supreme Court based on how many Justices they appointed and who those Justices replaced); Orley Ashenfelter, Theodore Eisenberg & Stewart J. Schwab, Politics and the Judiciary: The Influence of Judicial Background on Case Outcomes, 24 J. LEGAL STUD. 257, 276–77 (1995) (finding that “[c]ases before judges appointed by Republican presidents are more likely to have settled and won or settled than cases before judges appointed by Democratic presidents, although the result [was] not statistically significant”); Lee Epstein et al., Ideological Drift Among Supreme Court Justices: Who, When, and How Important?, 101 NW. U. L. REV. 1483, 1521–28 (2007) (finding that Justices behave according to the appointing President’s expectations in the early years of their tenure, but that as time passes, the influence of the appointing President decreases).

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

336 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

whether a case survives summary judgment.

IV. RESULTS We first ran a number of descriptive and bivariate analyses of summary judgment outcomes with key explanatory variables. The dependent measure (outcome of motion for summary judgment) was close to even, but that defendants were slightly more likely to be successful at this stage of litigation than were plaintiffs. In the 522 cases that reached summary judgment, 293 (or 56%), terminated the plaintiffs’ case, while all or some portion of the plaintiffs’ case survived in 229 (or 44%) of the cases we analyzed. Next, we analyzed key judicial characteristics. Table 1 below presents these descriptive statistics. These data show that, in the aggregate, minority judges have served on the bench for virtually the same period as white judges (9.95 years versus 9.88 years for white judges). The number of Republican-appointed judges was higher (54.5% versus 44.5% for Democrat appointees)86 but on a proportionate basis, minority judges were 53.2% Republican and 46.8% Democrat. Roughly 10% of the judges in our sample were racial/ethnic minorities and 22% were female. Thus, our sample includes slightly fewer women judges than the national composition (22% in our sample versus 29% nationally), and many fewer minority judges (10% in our sample versus 23% nationally).87 The differences between our sample and the broader judiciary are the result of drawing from selected jurisdictions and not the entire federal district court judiciary.

86. Within the comprehensive dataset, there was a higher proportion of minority judges who were elected by Democrat presidents (67.59% minorities versus 32.41% white). 87. Federal Judicial Center, supra note 68.

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 337

TABLE 1. Descriptive Statistics on Judges at Summary Judgment Variable Frequency Percent (%)

Race White 432 85.35 Minority* 47 9.81 Gender Male 374 78.1 Female 105 21.9 Political Party of Appointing President Republican 261 54.5 Democrat 218 45.5 Years on Bench White Mean = 9.88 σ = 7.8 Minority* Mean = 9.95 σ = 7.5

*Minority includes Black, Hispanic, and Asian. Minority status used for analysis based on Federal Judicial Center’s classification. After analyzing the simple distributions of case outcomes and judicial characteristics, we ran bivariate analyses of three categories of variables against the dependent variable: plaintiff characteristics, claim characteristics, and judicial characteristics. This revealed two statistically significant relationships.88

88. The flipside is that we found a number of variables for which the relationship with our dependent variable was insignificant; most notably, the judge’s political party affiliation (χ2 (1) = 0.0555, p = .814).

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

338 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

TABLE 2. Cross Tabulation Representational Status by Summary Judgment Outcome Case Dismissed at Summary Judgment

Representational Status No Yes Total Pro Se 29 (23.20%) 298 (76.8%) 382 Representation 180 (47.12%) 202 (52.88%) 125 TOTAL 209 298 507 (r = 0.209; χ2 (1) = 22.4, p = .000) A plaintiff’s representational status had a large, significant effect on summary judgment outcomes (r = 0.209; χ2 (1) = 22.4, p = .000). In table 2 above, the proportion of pro se plaintiffs who had their cases dismissed at summary judgment was 0.768 (67.79 percent), while plaintiffs with attorneys had their cases dismissed only 52.88 percent of the time. This reveals nearly a 24 percent difference in summary judgment outcome based only on whether the plaintiff has a lawyer. However, other plaintiff characteristics, such as gender and age, were not significantly associated with summary judgment outcome. This finding is consistent with previous work with this dataset89 and existing access to justice research.90

TABLE 3. Cross Tabulation Judge Minority Status by Summary Judgment Outcome Case Dismissed at Summary Judgment

Judge Minority Total No Yes Status White 169 (39.12%) 263 (60.88%) 432 Minority 29 (61.7%) 18 (38.3%) 47 TOTAL 198 281 479 (r = -0.1364; χ2 (1) = 8.91, p = .003) Table 3 above shows a strong and significant difference in the

89. See Nielsen, Nelson & Lancaster, supra note 28, at 188–92 (finding that legal representation is critical to plaintiff success in court generally, but also to avoid losing at summary judgment). 90. See, e.g., WOLF HEYDEBRAND & CARROLL SERON, RATIONALIZING JUSTICE: THE POLITICAL ECONOMY OF FEDERAL DISTRICT COURTS (Richard H. Hall ed., 1990); DEBORAH L. RHODE, ACCESS TO JUSTICE 14–15 (2004); Rebecca L. Sandefur, Access to Civil Justice and Race, Class, and Gender Inequality, 34 ANN. REV. SOC. 339, 346–52 (2008).

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 339 outcomes handed down by minority judges versus white judges (r = -0.1364; χ2 (1) = 8.91, p = .003). Overall, white judges are far more likely to grant a motion for summary judgment for the defendant (61 percent of cases), than are their counterpart minority judges (38 percent of the time, or some 23 percent less than white judges). This finding is consistent with previous work concerning the role of race and decisionmaking.91 However, there was no statistically significant difference in dismissal rates based on a judge’s gender, political party affiliation, or years on the bench.

A. LOGISTIC REGRESSION MODEL The focal point of our analysis examines the independent main effects of plaintiff, claim, and judicial characteristics and how they influence summary judgment outcomes. We present the results of our logit model in table 4.

91. See Chew & Kelley, supra note 83, at 1156–58 (finding that judges’ race impacts the outcome in discrimination cases).

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

340 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

TABLE 4. Binary Logit Models Predicting Summary Judgment Outcome, 1988–2003 Independent Variable Litigation Judicial Complete Characteristics Characteristics Model Plaintiff Characteristics -0.157 -0.083 -1.376* Minority 0.267(0.491) 0.244 (0.524) Female -0.719(0.511) -0.831(0.535) Manager, Professional -1.836**(0.699) -1.658*(0.735) Sales, Service, Office Occ. -1.915**(0.727) -1.64*(0.764) Age -0.0228(-1.09) -0.012(0.0244) Member of Union 0.285(0.57) -0.217(0.612) Statutory Basis / Type of Discrimination Title VII – Race -0.596(0.533) -0.532(0.576) Title VII – Sex 1.658**(0.591) 1.63*(0.605) Title VII – Other 0.102(1.209) -1.206(-0.92) ADEA – Age 0.242(0.528) 0.348(0.551) ADA – Disability -0.309(0.508) -0.145(0.543) 42 USC 1981 0.482(0.516) 0.558(0.536) 42 USC 1983 -1.841*(0.76) -1.890*(0.776) Constitutional Case 0.624(0.874) 0.529(0.907) Other Statutory Basis of Suit -0.349(0.449) -0.446(0.480) Alleged Discriminatory Practice Hiring -2.037**(0.764) -1.97*(0.789) Firing -0.805(0.436) -0.749(0.460) Retaliation -1.053*(0.418) -.0950*(0.454) Sexual Harassment -1.048(0.791) -0.600(0.863) Conditions of Employment -0.309(0.471) -0.424(0.499) Pay -1.155*(0.589) -1.17(0.617) Litigation/Representational Status Pro Se 2.081***(0.565) 2.024***(0.58) Multiple Plaintiffs -1.781*(0.727) -1.705*(0.735) Judicial Characteristics Minority -2.581***(0.773) -2.247779 Female -0.179(0.388) -0.436(0.522) Democrat -0.141(0.332) -0.0227(0.429) Years on Bench 0.0088(0.0205) -0.018(0.0273) Constant 6.186***(1.746) 3.035***(3.31) 8.65***(2.158) Observations 184 184 184 AIC 227 243.9 226.4 BIC 304.2 260 316.4 As model 1 shows, a number of variables predict the dismissal of a

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 341 case at summary judgment, including a plaintiff’s representation status (unrepresented plaintiffs’ cases are more likely to be dismissed), and the number of plaintiffs (the more plaintiffs in a lawsuit, the less likely it will be dismissed). These findings are consistent with our bivariate analyses. Pro se plaintiffs are statistically more likely to have their cases dismissed entirely at summary judgment (b = 2.081, p < .001). In addition, a case with more than one plaintiff is less likely to be dismissed entirely at summary judgment (b = -1.78, p < .01). There were no significant differences based on a plaintiff’s minority status or gender, but we suspect that controlling for claim type (race discrimination versus sex discrimination) accounted for any potential variation. In this model, sex discrimination claims were more likely to be dismissed at summary judgment (b = 1.658, p < .01). Model 2 examines only judicial characteristics, with the hypothesis that judges may decide cases politically or personally, regardless of claim characteristics. Although this model is the weakest of our three models, it shows that a judge’s minority status significantly predicts case disposition. Consistent with our bivariate analysis and other literature, a white judge is more likely to dispose of a case at summary judgment than a minority judge (b = -2.581, p < .01). This holds true even when controlling for judges’ gender, political party, and tenure effects.92 For purposes of this analysis as well as for the remainder of the Article, we focus on model 3 featured in table 2 because the fit statistics indicate that this model best fits our data.93 From this model, we find a

92. We did not include district effects in our analysis for a number of reasons. First, we examined whether minority judges were evenly distributed within the various jurisdictions. We found that minority judges were evenly distributed within the seven jurisdictions, although there were slightly more judges in the Eastern District of Pennsylvania and fewer in the Northern District of Texas: Eastern District of Pennsylvania (18.04%); Northern District of Illinois (16.22%); Southern District of New York (14.23%); Northern District of California (13.78%); Northern District of Georgia (9.72%); Eastern District of Louisiana (8.94%); and Northern District of Texas (7.2%). We then conducted a bivariate analysis to determine whether there were statistically significant differences in summary judgment outcome based on district. There were significant differences across jurisdictions (χ2 (6) = 32.76, p = .000), but the actual distribution showed that only the Eastern District of Pennsylvania retains cases at summary judgment at a higher rate, whereas the other districts have a higher proportion of dismissing cases entirely. This is the same jurisdiction with a higher proportion of minority judges than the other federal district courts. 93. We conducted Akaike Information Criterion (“AIC”) and Bayesian Information Criterion (“BIC”) tests to judge which model best fit the values of our data. These tests select the model that maximizes the log likelihood (the likelihood that the true values from our data “fit” the values produced from the logistic regression), given the number of parameters (independent variables). See generally Kenneth P. Burnham & David R. Anderson, Multimodel Inference: Understanding AIC and BIC in

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

342 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

number of effects that influence summary judgment disposition. With respect to plaintiff and claim characteristics, cases involving sex discrimination are more likely to be dismissed (b = 1.63, p < .01). Not surprising, in this case, the number of plaintiffs and representational status predicted whether a case was dismissed on summary judgment. When a case has more than one plaintiff, it is significantly less likely to be dismissed (b = -1.705, p < .05). When a plaintiff is unrepresented, a case is more likely to be dismissed (b = 2.024, p < .001). In model 3, there is only one variable that predicts whether a case will be dismissed entirely at summary judgment. Consistent with our bivariate analysis and model 2, a white judge is more likely to dispose of a case at summary judgment than is a minority judge (b = -2.339, p < .05). This holds true even when controlling for judges’ gender, political party affiliation, and tenure effects.

B. INDIVIDUAL-LEVEL ANALYSES While the effects presented in table 4 provide an overall analysis of the determinants of outcomes in employment cases at the summary judgment phase, we also were interested in examining the interaction between certain characteristics of the plaintiffs and the minority status of the judge. To get at this analysis, we constructed predicted probabilities, which allowed us to understand the exact impact of a variable on our outcome.94 We also considered representational status because of its statistical significance in our logit models and because it may capture more

Model Selection, 33 SOC. METHODS & RES. 261 (2004) (discussing AIC versus BIC model selection and multi-model inference, including methods of model averaging). When we compared our first and third models, we found that the third model—which included claim, plaintiff, and judicial characteristics—best fit our data. Even though district did not appear to influence summary judgment outcome, we ran a logistic regression that accounted for district effects. Aside from the Eastern District of Pennsylvania remaining significant, district effects proved to be insignificant. We removed district effects from our final model because they also weakened the fit of our model to our data. The McFadden’s R2 went up from 0.321 (32.1 percent) to 0.371 (37.1 percent), suggesting that the variance increased as we took into account district effects. Moreover, although the AIC value went down as we added our district variables, the BIC statistic (which imposes a more stringent standard), was smaller when we removed them from the model. This is consistent when running the likelihood ratio test in which district variables did not improve the model (p = .0520). 94. J. SCOTT LONG & JEREMY FREESE, REGRESSION MODELS FOR CATEGORICAL DEPENDENT VARIABLES USING STATA 119–20 (1st ed. 2001). We controlled for judge and plaintiff minority status as well as representational status and held the other variables at their means.

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 343 latent forms of racial and socioeconomic disadvantage.95 What follows is a presentation of race and pro se estimates.

1. Judge-Plaintiff Minority Status Table 5 shows that white and minority judges dispose of race discrimination summary judgment motions at different rates. And, when judges hear cases brought by plaintiffs who are the same minority status as the judge, the cases survive motions for summary judgment at a much higher rate. For example, when a white judge decides a case involving a white plaintiff, the plaintiff’s case (or some portion of it) has a 40 percent predicted probability of surviving a motion for summary judgment. When a white judge adjudicates a case involving a minority plaintiff, however, the predicted probability of the plaintiff’s case surviving summary judgment drops to roughly 34.43 percent.

TABLE 5. Predicted Probabilities Estimated for Minority Status Using Logit Model Independent Variable SJ Survival SJ Dismissal Pr(y = 0|x) Pr(y = 1|x)

White Judge White Plaintiff 0.4014 0.5986 Minority Plaintiff 0.3443 0.6557

Minority Judge White Plaintiff 0.8742 0.1258 Minority Plaintiff 0.8448 0.1552 Minority judges are more likely to allow employment civil rights cases to continue past motions for summary judgment regardless of the race of the plaintiff. When a minority judge presides over a case involving a minority plaintiff, the plaintiff’s case (or some portion of it), has an 84.48

95. Both the more comprehensive data and summary judgment sample show that minorities are more likely to pursue these matters pro se. In our larger sample, 21.48 percent of plaintiffs remained pro se throughout litigation versus 8.13 percent of white plaintiffs. Robert L. Nelson et al., Rights on Trial: Race and Representation in Employment Civil Rights Litigation 55 (Apr. 11, 2011) (unpublished manuscript) (on file with authors). Our summary judgment subsample shows that of the cases that reached summary judgment, 32.27 percent were cases brought by minority pro se plaintiffs, while 13.66 percent of cases were brought by white pro se plaintiffs.

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

344 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

percent predicted probability of surviving summary judgment. And, consistent with prior research,96 our data demonstrates that a minority judge is likely to allow some portion of a white plaintiff’s case to continue at nearly the same predicted probability (87.42 percent). Overall, there is a 47.28 percent difference between white and minority judges who adjudicate claims involving minority plaintiffs, and roughly a 50 percent difference between white and minority judges who adjudicate claims involving white plaintiffs. 2. Judge-Plaintiff Minority and Representational Status Because our logit model revealed that unrepresented plaintiffs are more likely to have their cases dismissed at summary judgment, we constructed hypothetical cases to examine the predicted probabilities of case dismissal based on race and representational status.

TABLE 6. Predicted Probabilities Estimated for Minority and Pro Se Status Using Logit Model Independent Variable SJ Survival SJ Dismissal Pr(y = 0|x) Pr(y = 1|x)

White Judge White Pro Se Plaintiff 0.1331 0.8669 Minority Pro Se Plaintiff 0.1073 0.8927

Minority Judge White Pro Se Plaintiff 0.6141 0.3859 Minority Pro Se Plaintiff 0.5549 0.4451

As we have shown, pro se plaintiffs are more likely to have their cases dismissed at summary judgment. These differences become more pronounced depending on the judge’s race. Table 6 above reveals similar trends in minority status and disposal of cases. White judges tend to dismiss minority pro se plaintiffs at a slightly higher rate than white pro se plaintiffs. They will dismiss a case involving a pro se minority plaintiff 89.27 percent of the time, whereas they will dismiss a case involving a

96. See Uhlman, supra note 83.

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 345 white pro se plaintiff 86.7 percent of the time. Minority judges dismiss cases at a much lower rate than white judges. Cases involving a minority judge and a minority pro se plaintiff have a 44.51 percent predicted probability they will be dismissed. And cases involving a minority judge and a white pro se plaintiff are even less, where the predicted probability of a dismissal is 38.59 percent. This is nearly a 44.76 percent difference between white and minority judges who adjudicate claims involving minority pro se plaintiffs, and a 48.1 percent difference between white and minority judges who adjudicate claims involving white pro se plaintiffs.

C. SUMMARY These results provide support that empathetic decisionmaking plays a role in employment civil rights cases. The logistic regression model that included plaintiff, claim, and judicial characteristics shows that the minority status of the judge predicts whether a case will be dismissed entirely at summary judgment. When we conducted a fine-grain analysis— examining plaintiff and judge minority status—we found that white judges were less likely to dismiss a case at summary judgment than minority judges. Even though white plaintiffs had a lower predicted probability of having the cases dismissed, this probability was even lower if a minority judge adjudicated a claim involving a white plaintiff. Our findings found some support for the liberal legal model, but also clearly demonstrate that a judge’s minority status still explained a great deal. A liberal legal model of decisionmaking would suggest that claim or plaintiff characteristics would make little difference but rather the merits of the claim would predict case outcome. The two representation variables would provide support for this claim because both predicted case disposition. When we accounted for representational effects in our predicted probabilities, we saw variation based on plaintiff-judge minority status. The predicted probabilities did change—namely, that pro se plaintiffs had a higher predicted probability of having their case dismissed—but our data still showed that white judges were more likely to dismiss cases involving minority plaintiffs while minority judges were less likely to dismiss cases involving white plaintiffs. Finally, our findings did not provide support for the political model of decisionmaking. In the two models that accounted for the political party affiliation of the judge, there was no significant difference when a judge was appointed by a Democratic or Republican president or when a judge resided in the Northern District of Texas versus the Northern District of

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

346 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

California. Although we cannot say for sure what affects judicial decisionmaking at the summary judgment phase, we suspect that judges were acting less politically because of the discretionary nature of the summary judgment motion and because the cases we examined were at the district court level (and hence there was less at stake politically or professionally). Our findings are both strong and provocative. For sociolegal scholars to better understand the dynamics of personal experience and empathy, we require more research on judicial decisionmaking at different points in pretrial and trial process. And, these analyses would need to span different areas of legal decisions. The “prime” associated with employment discrimination cases may make judicial characteristics and empathy more salient than it would be in say, bankruptcy or contract cases. Methodologically, the regression models and predicted probabilities yielded interesting findings about real cases; however, every case is unique. In other words, our current data cannot identify what specific information influences whether a judge sees (or does not see) discrimination. The use of an experimental research would be the most appropriate approach because we would have greater control scenarios presented to individuals and to manipulate variables that we believe shape an individuals’ interpretation of events.

V. EMPATHY, EXPERIENCE, AND A NEW VISION OF JUDICIAL DECISIONMAKING These results suggest that judges’ assessments of employment discrimination cases vary. We contend that this variation is the result of the different attitudes, opinions, and experiences that stem from being white or a person of color. White judges are far more likely to dispose of any employment discrimination case at the summary judgment phase than are minority judges. Our data also show that even when we take into account pro se status—believing that the economic and legal resources may influence the viability of a claim—white judges tend to dismiss cases involving minority plaintiffs at a much higher rate than cases involving white plaintiffs. Equally compelling is the finding that there is a higher predicted probability that minority judges dismiss cases involving minority plaintiffs than cases involving white plaintiffs, even when we take into account representational effects. These results raise a number of issues concerning the diversity of the judiciary, the mission of the adversarial

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 347 system, and the role of social science research to aid in judicial decisionmaking.

A. DIVERSITY OF THE JUDICIARY A diverse judiciary is essential to the administration of justice and to the retention of faith in the courts.97 The law recognizes that litigants are entitled to a jury of their peers from the community,98 yet they often have a judge who comes from a different racial, ethnic, or socioeconomic background. In other words, while the Constitution does not provide for a “judge of her peers,” they are very much the face of justice that litigants will associate when perceiving the fairness of the legal system. Judges are gatekeepers. They are instrumental in adjudicating discrimination claims—particularly because so few cases reach jury trial.99 As such, they serve as the primary triers of fact in these cases. For this reason, it is important to have a diverse judiciary—accounting not only for traits we typically associate with “diversity” such as race, gender, sexual orientation, and so forth. Instead, we should embrace a broader understanding of diversity that captures a broader array of characteristics such as geographical, socioeconomic, professional, and intellectual backgrounds. According to the Federal Judicial Center, diversity within the federal judiciary is minimal.100 Only 136 of the 597 active federal district judges are a member of a racial/ethnic minority, roughly 22.8 percent.101 This

97. The qualitative data from our larger project reveals that plaintiffs lose faith in the legal authorities, specifically, and the legal system, more broadly. Interviews with plaintiffs show that they become disillusioned during the litigation process because they believe that judges make value judgments about the litigants, and in turn, about the merits of the case, particularly in instances in which the plaintiff is pro se. See Ellen Berrey, Stephen Hoffman & Laura Beth Nielsen, Situated Justice: Plaintiffs’ and Defendants’ Perceptions of Fairness in Employment Discrimination Cases, in LAW & SOC’Y REV. (forthcoming) (manuscript at 1, 18–25, 31–33) (on file with the authors); Chen, supra note 34, at 1117 (“A diverse judiciary . . . enhances courts' credibility among affected communities who would otherwise feel they have no voice within the institution.”); Ifill, supra note 34, at 410 (“Because they can bring important and traditionally excluded perspectives to the bench, minority judges can play a key role in giving legitimacy to the narratives and values of racial minorities.”). 98. See Taylor v. Louisiana, 419 U.S. 522, 526–27 (1975) (noting that the American concept of a jury trial contemplates a jury drawn from a cross-section of the community); Hernandez v. Texas, 347 U.S. 475, 482 (1954) (reversing defendant’s conviction because members of a particular ethnic group were systematically excluded from serving on juries). 99. See, e.g., Nielsen, Nelson & Lancaster, supra note 28, at 187 (finding that only 100 of the 1,672 federal employment discrimination cases in their random sample reached trial). 100. Federal Judicial Center, supra note 68. 101. Id. This figure is higher than the percentage of racial/ethnic minorities in the legal profession

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

348 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

figure is slightly larger for female judges; 174 federal district court judges are women—roughly 29 percent.102 These figures do not account for other characteristics that further limit the diversity of the bench, such as females who are racial/ethnic minorities.103 More critically, variables such as socioeconomic status, geographic region of sitting judges, and life experiences may play a critical role in how individuals understand cases that involve more than simply applying the law to the facts. Increased diversity does not mean appointing judges who have predetermined positions, but instead those who have different ways of looking at the world.104 In the context of discrimination claims, anecdotal and empirical accounts show that those who experienced prejudice or bias are more likely to develop empathy resulting from that experience.105 We believe this evidence can be beneficial for judges who preside over civil rights matters. As seen by the comments of Justices Thomas and Sotomayor, there are instances in which a simple holding fails to describe and capture the meaning and significance behind words and actions.106 In the context of civil rights claims, the manner in which discrimination manifests is difficult to pinpoint with a one or two sentence discussion. overall. In 2010, only 11.1 percent of lawyers were racial/ethnic minorities. Bureau of Labor Statistics, Employed Persons by Detailed Occupation, Sex, Race, and Hispanic or Latino Ethnicity, ftp://ftp.bls.gov/pub/special.requests/lf/aat11.txt (last visited Jan. 8, 2012). However, it is lower than the percentage of racial/ethnic minorities in general, which is over 25 percent. Karen R. Humes, Nicholas A. Jones & Roberto A. Ramirez, Overview of Race and Hispanic Origin: 2010, 2010 CENSUS BRIEFS, (Mar. 2011), http://www.census.gov/prod/cen2010/briefs/c2010br-02.pdf (C2010-BR02) (reporting that 72.4 percent of the general population identifies as white). 102. Federal Judicial Center, supra note 68. This figure is slightly lower than the legal profession overall, where women make up 31.5 percent of all lawyers. Bureau of Labor Statistics, supra note 101. 103. For example, only twenty-six federal district court judges are black females, or 4.4 percent. Federal Judicial Center, supra note 68. 104. See, e.g., Joy Milligan, Note, Pluralism in America: Why Judicial Diversity Improves Legal Decisions About Political Morality, 81 N.Y.U. L. REV. 1206, 1235 (2006) (“Diversity of viewpoint, then, might be a first-best goal for the judiciary. If we included judges from a range of ideological backgrounds, and ensured that they deliberated together on appellate panels, we might expect a high level of openness to alternative conceptions of political morality.”); Theresa M. Beiner, What Will Diversity on the Bench Mean for Justice?, 6 MICH. J. GENDER & L. 113, 150 (1999) (“The best judge . . . should be able to see and assess the differing perspectives of the many parties and persons involved in the litigation. This is where diversity becomes important.”). 105. See supra text accompanying notes 1–7. 106. While this article sought to show the importance of a diverse judiciary through legal narrative and empirical evidence, the call for a more diverse judiciary has been previously made by some of the Justices themselves. Justice Ginsburg has commented that “[a] system of justice is the richer for the diversity of background and experience of its participants.” Ruth Bader Ginsburg, The Supreme Court: A Place for Women, 32 SW. U. L. REV. 189, 190 (2003).

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 349

B. EMPATHY AND THE EFFICACY OF THE ADVERSARIAL SYSTEM The adversarial system is characterized as uniform and predictable. In the ideal circumstance, if one judge adjudicates a claim and believes the plaintiff suffered illegal discrimination, a similarly situated judge should arrive at a similar decision. The point is that there should be a degree of predictability to the civil litigation process. If the outcomes of a case can vary if the plaintiff is from a different minority group than the judge, the ideals of an adversarial system governed by the rule of law are diminished. Our data show that case viability in discrimination claims depends greatly on two variables. First, the minority status of a judge influences whether a case will survive summary judgment. Our logistic regression model and the predicted probabilities not only show significant differences in how judges decide discrimination cases, but they also show that such differences depend on the minority status of the plaintiff and judge. Second, the representational status of the plaintiff affects whether a case will survive summary judgment. One obvious explanation for this finding is that a pro se plaintiff does not have the legal expertise or financial resources to respond to a motion for summary judgment. As a result, the evidence a plaintiff proffers is usually a personal account with little or no physical evidence such as an email exchange or testimony from coworkers. However, when we look closer at our data, black male plaintiffs are significantly more likely to be pro se than other groups, and in particular, white plaintiffs.107 Findings from the quantitative analysis alone might suggest a few interpretations. Regarding race, one possible theory is that black plaintiffs, and by extension the majority of Title VII race claimants, are less likely to be able to afford attorneys or earn attorneys’ cooperation in a contingency fee arrangement.108 Another possible theory is that these plaintiffs have less social capital, meaning fewer personal and social connections of the type that would match them with attorneys; in this scenario, plaintiffs might never attempt to seek out attorneys or gain initial access, regardless of ability to pay.109

C. ACCOUNTING FOR EMPATHY AND THE RULE OF LAW Given the disparities in summary judgment outcomes, how can we promote an agenda of judicial diversity while maintaining the consistency

107. Nelson et al., supra note 95, at 23–24. 108. Id. at 27. 109. Nelson et al., supra note 95.

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

350 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

and predictability of the adversarial system and the rule of law? We believe that social science research can help judges learn to be more empathetic, and, in turn, better positioned to have a better understanding of their cases. Specifically, psychology research on empathetic induction consistently shows improved attitudes toward members of marginalized groups.110 This research not only shows a growing awareness for victims,111 but also helps individuals to understand members of stigmatized groups.112 However, teaching individuals to be more empathetic varies based on intragroup and intergroup characteristics.113 Research reports that group membership can moderate the impact of empathetic induction; individuals with different group membership exhibit less empathy than those who are similarly situated.114 Notwithstanding these findings, the research in this area consistently shows that empathetic induction helps individuals recognize more nuanced, situational narratives that are distinct from their own.115 Along the same lines, social science can expose inequality and the harmful effects of discrimination, phenomena that have become more

110. See, e.g., Batson et al., supra note 40, at 1665 (finding that participants induced to feel empathy for a heroin user had more positive attitudes toward people addicted to drugs than participants not induced to feel empathy); id. at 1656 (finding that empathy induction proved to improve attitudes concerning people with AIDS, who are homeless, and convicted murderers); Krystina A. Finlay & Walter G. Stephan, Improving Intergroup Relations: The Effects of Empathy on Racial Attitudes, 30 J. APPLIED SOC. PSYCHOL. 1720, 1731–32 (2000) (finding that participants who read about discrimination against blacks or were instructed to be empathetic toward victims of discrimination improved their attitudes concerning blacks); James D. Johnson et al., Rodney King and O.J. Revisited: The Impact of Race and Defendant Empathy Induction on Judicial Decisions, 32 J. APPLIED SOC. PSYCHOL. 1208, 1216 (2002) (finding that empathy-induced participants had altered and more lenient judicial determinations about a defendant); Theresa K. Vescio, Gretchen B. Sechrist & Matthew P. Paolucci, Perspective Taking and Prejudice Reduction: The Meditational Role of Empathy Arousal and Situational Attributions, 33 EUR. J. SOC. PSYCHOL. 455, 467 (2003) (finding that participants instructed to “take perspective” of a black male as he described his difficulties experienced more empathy for him and more favorable attitudes about blacks). 111. C. Batson et al., Empathy and Attitudes: Can Feeling for a Member of a Stigmatized Group Improve Feelings Toward the Group?, 72 J. PERSONALITY & SOC. PSYCHOL. 105, 107 (1997). This study found that people who were responsible for their own victimhood—for example, a person who contracted AIDS because of unprotected sex—invoked a less empathetic response. However, participants could still be empathetic if they learned of a victim’s responsibility after empathy induction. Id. at 107–11. 112. See supra note 110. 113. See generally C. Daniel Batson & Nadia Y. Ahmad, Using Empathy to Improve Intergroup Attitudes and Relations, 3 SOC. ISSUES & POL’Y REV. 141 (2009) (discussing four empathy states, empathy at the intergroup level, and the role of empathy in existing programs meant to improve intergroup relations). 114. Johnson et al., supra note 110 at 1209. 115. See supra note 110.

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

2012] EXAMINING EMPATHY 351 implicit, embedded, and institutionalized. Social science evidence has undeniably been a successful tool for policy and legal change. Brown v. Board of Education is often cited as the paradigmatic example of this.116 The Supreme Court turned to social science to justify that de jure racial segregation of schools not only is inherently unequal, but also that the impact of segregation practices is harmful.117 For judges who may not have a genuine understanding of matters concerning discrimination, social science has the power to expose embedded inequality in society and how it impacts individuals emotionally, professionally, and economically. For skeptics who believe that legal claims are frivolous or are simply meant to further political objectives (for example, the need for and efficacy of affirmative action policies), scientific research provides a level of objectivity and reliability to the discourse that goes beyond one judge’s opinion—for example, showing objective data that organizational practices have the power to substantially exacerbate or mitigate bias in pay and promotion practices.

VI. CONCLUSION The most recent Supreme Court Justice confirmations reignited an important debate worthy of legal and empirical examination. Specifically, the call for empathetic judges sparked discussions concerning the nature of decisionmaking and whether a judge’s background should enter the courtroom. Critics of the empathetic judge view cognitive and affective awareness of a litigant’s state or condition not only as an indicator of partiality, but also as a mere window dressing for political motivations. For advocates of the empathetic judge, a diverse vision and understanding of cases involving more contextualized disputes such as discrimination can lead to outcomes that account for embedded issues such as implicit bias. While legal scholars have long explored the impact of a diverse judiciary, few have examined it from an empirical standpoint. We sought to fill this gap by focusing our analysis on individual-level variables including the race, gender, and representational status of the plaintiffs in tandem with judicial characteristics. Even after controlling for potential confounding variables such as the judge’s political party affiliation, years on the bench, and other case characteristics, we found that minority judges tend to assess discrimination claims differently than white judges. Although this finding is consistent with previous work on judicial decisionmaking, the

116. Brown v. Bd. of Educ., 347 U.S. 483, 493–94 (1954). 117. Id.

WEIN14.DOC (DO NOT DELETE) 2/27/2012 2:58 PM

352 SOUTHERN CALIFORNIA LAW REVIEW [Vol. 85:313

insignificance of political party in our data suggest that some apolitical factor was motivating judges.

2009 DENOVO 133 Page 1 2009 Cardozo L. Rev. de novo 133

Cardozo Law Review de novo Empathy and Justice

2009

*133 EMPATHETIC JUDGING AND THE RULE OF LAW

Susan A. Bandes [FNa1]

Copyright © 2009 Yeshiva University; Susan A. Bandes

POP QUIZ [FN1]

In which of the following excerpts from U. S. Supreme Court opinions is empathy employed? Circle all that apply.

a) We consider the underlying fallacy of the plaintiff's argument to consist in the assumption that the en- forced separation of the two races stamps the colored race with a badge of inferiority. If this be so, it is not by reason of anything found in the act, but solely because the colored race chooses to put that construction upon it. (Plessy v. Ferguson). [FN2]

b) Poor Joshua! Victim of repeated attacks by an irresponsible, bullying, cowardly, and intemperate father, and abandoned by respondents who placed him in a dangerous predicament and who knew or learned what was going on, and yet did essentially nothing. (Justice Blackmun's dissent in DeShaney v. Winnebago County Dept of Social Services). [FN3]

c) It is self-evident that a mother who comes to regret her choice to abort must struggle with grief more an- guished and sorrow more profound when she learns, only after the event, what she once did not know: that she allowed a doctor to *134 pierce the skull and vacuum the fast-developing brain of her unborn child, a child as- suming the human form. (Gonzales v. Carhart). [FN4]

d) I've got suspicion that some drug is on this kid's person. My thought process is I would rather have the kid embarrassed by a strip search, if we can't find anything short of that, than to have some other kids dead because the stuff is distributed at lunchtime and things go awry. (Justice Souter, at the oral argument in Redding v. Saf- ford United School District). [FN5]

e) Private suits against unconsenting states ... present the indignity of subjecting a State to the coercive pro- cess of judicial tribunals at the instance of private parties ... Not only must a State defend or default but also it must face the prospect of being thrust, by federal fiat and against its will, into the disfavored status of a debtor. ( Alden v. Maine). [FN6]

THE EMPATHY DEBATE

From the moment presidential candidate Obama announced his desire to appoint empathetic judges, a con-

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 2009 DENOVO 133 Page 2 2009 Cardozo L. Rev. de novo 133

tentious and often frustrating debate was inevitable. The debate was bound to be frustrating because empathy is a term with no fixed meaning. [FN7] It was bound to be contentious because the notion of empathy in judging appears to conflict with the ideal of the rule of law.

Recent Supreme Court confirmation hearings have been dominated by a radically oversimplified view of the rule of law. [FN8] The view is that judges--even Supreme Court justices--are “umpires who don't make the rules, they apply them.” [FN9] If one accepts this view that law can be discovered and applied *135 without in- terpretation, it follows that judges should not allow their prior experiences, perceptions or beliefs to influence their decisions, and that instead a Supreme Court justice should be “transformed” into “a different person” upon taking the judicial oath, a person who simply does “what the law requires in every single case.” [FN10] Judicial candidates who acknowledge the influence of prior experience or beliefs are, in this view, acknowledging im- proper ideological bias. [FN11]

Obama's statement on the campaign trail that “we need somebody who's got the heart, the empathy, to recog- nize what it's like to be a young teenage mom ... [and] to understand what it's like to be poor, or African- American, or gay, or disabled, or old,” [FN12] was an explicit rejection of this view of the Court's role. Obama made the statement in a speech to Planned Parenthood. He prefaced it with critical remarks about Justice Kennedy's opinion in the recent partial birth abortion case, and about the shallowness of the Roberts confirma- tion hearing. It amounted to a declaration that Supreme Court justices do have interpretive philosophies, that it is legitimate to take them into account, and that Obama intends to appoint justices with a different set of values.

These remarks, followed by others about the value of empathy, created a firestorm, fanned by Judge Soto- mayor's acknowledgment that her background and worldview shape her jurisprudence. Because the concepts of empathy and the rule of law are both moving targets, many scholars and pundits seem to be talking past one an- other. Nevertheless, a remarkable and long-overdue conversation is unfolding about judicial character and how it ought to affect decision-making. This essay addresses some of the definitional ambiguities that make construct- ive debate so difficult.

In addition, the essay argues that it is misleading to discuss whether judges should exercise empathy. They should, and they inevitably do. The questions are for whom they exercise it, how accurately they exercise it, how aware they are of their own limitations and blind spots, and what they do to correct*136 for those blind spots.

EMPATHY VERSUS SYMPATHY

Empathy, as the term is most commonly used, [FN13] is the ability to take the perspective of another, [FN14] or “an imaginative leap into the mind of others.” [FN15] Empathy is a capacity, not an emotion. It dif- fers from sympathy or compassion, both of which are emotions. Empathy entails understanding another person's perspective. Sympathy is a feeling for or with the object of the emotion. [FN16] Empathy does not require, or necessarily lead to, sympathy. [FN17] Empathy, unlike sympathy, does not necessarily lead to action on behalf of its object, or the desire to take action on his behalf. [FN18] Justice Blackmun's famous “Poor Joshua” lament in DeShaney is a good example of an expression of sympathy. It reflects not just empathetic understanding of Joshua's perspective, but a visceral sense that Joshua had been dealt with unjustly and deserved a different out- come.

As Dr. Michael Franz Basch, a psychotherapist and prominent scholar on the topic of empathy, observes:

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 2009 DENOVO 133 Page 3 2009 Cardozo L. Rev. de novo 133

Empathy is first and foremost a capacity. Strictly thinking, it is value-free. Empathic thinking ... is a function that the human brain at a certain level of development is potentially capable of performing, no more and no less. This is often not understood, and empathy becomes confused with altruism and other- directedness, though it need not be employed in the service of either goal ... What one does with the in- sight provided by empathic understanding remains to be determined by the nature of the relationship between the people involved and the purpose for which the empathic capacity *137 was engaged by its user in the first place. [FN19] A psychotherapist uses empathy to promote the goals of therapy--trust, healing, self-knowledge. A judge uses empathy as a tool toward understanding conflicting claims. Empathy assists the judge in understanding the litigants' perspectives. It does not help resolve the legal issue of which litigant ought to prevail.

JUST ONE TOOL IN THE JUDICIAL TOOLBOX

Judge Posner describes empathy as one of several important tools in the judge's toolbox. Posner argues that when faced with legal questions lacking determinative answers, judges need to consult good judgment, which he defines as “an elusive faculty best understood as a compound of empathy, modesty, maturity, a sense of propor- tion, balance, a recognition of human limitations, sanity, prudence, a sense of reality and common sense.” [FN20] In the context of judicial decision-making, empathy is an essential capacity for understanding what's at stake for the litigants. Ideally, a judge will have the capacity to put herself in the shoes of all those with a stake in her ruling.

Those who are concerned by Obama's call for empathetic judges often assume that he means for judges to decide cases entirely by means of empathy. They assume that empathy will drive or override, rather than inform, judicial judgment.

The notion of empathy--and of the importance of background--has often been raised before without evoking such criticism. Judicial nominees and their supporters have long assured Congress that they both intend to up- hold the rule of law and are capable of empathy for those less fortunate. Clarence Thomas's controversial nom- ination to the Supreme Court got a crucial boost when liberal judge Guido Calabresi wrote that Thomas under- stands “what discrimination really means” and knows “the deep needs of the poor and especially poor blacks.” [FN21] Senator Danforth, after reading Judge Calabresi' s remarks into the Congressional Record, added his own assurances that Thomas's heart would be with “the ordinary *138 folk” if he were on the Supreme Court. [FN22] In his confirmation hearings, Samuel Alito sought to reassure those concerned about his capacity to em- pathize with workers and the poor by describing his Italian-immigrant father and his own upbringing in “an un- pretentious, down to earth community.” Senator Dewine added his assurances that “Judge Alito ... understands that judicial opinions affect real people and have real consequences.” He went on to quote Alito's own words, in a passage that has been widely circulated:

When I have cases involving children, I can't help but think of my own children. When I get a case about discrimination, I have to think about people in my own family who suffered discrimination because of their ethnic background or because of religion or because of gender, and I do take that into account. [FN23] One can only speculate about why President Obama's evocation of empathy has been treated so differently. There was no suggestion that Justice Thomas or Justice Alito would decide cases based solely on empathy, without reference to governing law or other constraints. That distinction is frequently elided in the current de-

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 2009 DENOVO 133 Page 4 2009 Cardozo L. Rev. de novo 133

bate.

BLIND JUSTICE AND MINDBLINDNESS

One pundit expressed the widely held sentiment that “Lady Justice doesn't have empathy for anyone. She rules strictly based upon the law and that's really the only way that our system can function properly under the Constitution.” [FN24] But in fact it should be uncontroversial that judges need empathy. Empathy builds on the understanding (known as theory of mind) [FN25] that others are separate from us, with separate mental states, desires, beliefs and perceptions. [FN26] The inability to “apprehend what seems to be going through someone else's mind” is called mindblindness. [FN27] Empathy helps us to understand*139 what others are thinking, feeling and perceiving, and to predict how others will react. It is an essential capacity for living in the social world, and a basic component of moral reasoning. [FN28] A lack of empathy is one of the deficits associated with autism in children. [FN29] A total lack of empathy, coupled with an equally total lack of remorse, is the main defect of psychopaths. [FN30]

The law aims to channel and influence human behavior. To apply the law, judges must constantly seek to un- derstand and predict motivations, intentions, perceptions, and other aspects of human conduct. Empathy makes that understanding possible.

MISTAKING SELECTIVE EMPATHY FOR UNBIASED JUDGING

Judges often face litigants from backgrounds with which they are familiar and comfortable. Their perspect- ive-taking on behalf of such litigants is so natural it is unlikely to be coded as empathy at all. We tend to reserve the term for the more difficult feat of understanding the perspectives of those from very different backgrounds. Those who spend their days surrounded by people with shared backgrounds, assumptions and perspectives may mistake their own perspective for the universal. This mistake is an occupational hazard for judges, who are en- couraged by the trappings of their role to speak in a universal voice [FN31] and to regard themselves as taking “the view from nowhere.” [FN32]

Consider, for example, Jeffrey Toobin's recent articles about the background and jurisprudence of Chief Justice Roberts. Roberts sees himself as an umpire, calling them as he sees them, unhampered by a preconceived world view. Yet Toobin observes that:

After four years on the Court ... Roberts's record is not that of a humble moderate but, rather, that of a doctrinaire conservative ... In every major case since he became the nation's seventeenth Chief Justice, Roberts has sided with the *140 prosecution over the defendant, the state over the condemned, the execut- ive branch over the legislative, and the corporate defendant over the individual plaintiff. Even more than Scalia, who has embodied judicial conservatism during a generation of service on the Supreme Court, Roberts has served the interests, and reflected the values, of the contemporary Republican Party. [FN33] An article published at the time of Roberts' confirmation hearing concluded that Roberts' world view seemed set at an early stage. It recounts an upbringing in a conservative community and schooling at an elite boarding school, both of them wealthy, nearly all white, and insulated from the political and social turmoil of the sixties and seventies. It describes a father who was an executive for a steel company that was hit with sex and race dis- crimination claims. It describes Roberts himself as consistent and steadfast in his conservative views from an early age. [FN34] Toobin reports that in private practice and in the first Bush Administration, a substantial por-

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 2009 DENOVO 133 Page 5 2009 Cardozo L. Rev. de novo 133

tion of Roberts' work consisted of representing the interests of corporate defendants who were sued by individu- als. [FN35]

Toobin recounts the following anecdote: As Chief Justice, Roberts was charged with preparing an annual re- port to Congress. In 2006,

he devoted his entire report to arguing for raises for federal judges, and he even went so far as to call the status quo on salaries a “constitutional crisis ... This request to Congress was universally popular among Roberts' colleagues ... Congress, however snubbed the Chief Justice. Six-figure salaries, lifetime tenure, and the opportunity to retire at full pay did not look inadequate to the elected officials, who make the same account as judges, and must face ordinary voters. Roberts's blindness on the issue may owe something to his having inhabited a rarefied corner of Washington for the past three decades. [FN36] The point of the anecdote, as Toobin makes clear, is not that the Chief Justice was wrong in some substant- ive sense about the need for raises, but that he failed to realize the narrowness of his own perspective. The larger point is not that Roberts has the wrong judicial philosophy, but that he denies having any legal philosophy. It is not that his world view has a *141 deleterious effect on his jurisprudence, but that he denies it has any effect on his jurisprudence. This sort of selective empathy is the most dangerous. Because it mistakes its own perspective for “the view from nowhere,” it fails to seek out other perspectives.

When commentators critique the call for judicial empathy, they often assume that run-of-the-mill judging is empathy-free. It is this assumption that leads to the charge that “empathy is simply a code word for judicial act- ivism.” [FN37] If empathy always runs toward the poor and the disenfranchised, and if decisions in favor of the privileged and the powerful are simply unmarked judging, then empathy is activist. But in fact the justices often exercise empathy on behalf of governmental officials, including other judges. Justice Rehnquist, discussing why judges have afforded themselves absolute immunity from civil rights suits while denying it to so many other government officials, suggested that:

If one were to hazard an informed guess as to why such a distinction in treatment between judges and prosecutors, on the one hand, and other public officials on the other, obtains, mine would be that those who decide the common law know through personal experience the sort of pressures that might exist for such decisionmakers in the absence of absolute immunity, but may not know or may have forgotten that similar pressures exist in the case of nonjudicial public officials to whom difficult decisions are commit- ted. [FN38] The justices frequently exhibit empathy for corporate defendants and other powerful litigants. As I've dis- cussed elsewhere, in University of Alabama v Garret the Court showed far more empathy for the state university dragged into court against its will than it did for Patricia Garrett, who was illegally fired by the University for taking a medical leave while she battled breast cancer. [FN39]

In a widely quoted opinion piece in the Wall Street Journal, John Hasnas argued that:

[I]n general, one can feel compassion for and empathize with individual plaintiffs in a lawsuit who are facing hardship. They are visible. One cannot feel compassion for or empathize with impersonal cor- porate defendants, who, should they incur liability, will pass the costs on to consumers, reduce their out- put, or cut employment. Those who must pay more for products, or are unable to obtain *142 needed goods or services, or cannot find a job are invisible. [FN40] It may indeed be easier to have sympathy or compassion for the poor tenant than for the wealthy landlord, or

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 2009 DENOVO 133 Page 6 2009 Cardozo L. Rev. de novo 133

for the alleged victim of police brutality than for the accused perpetrator. This is an important point. Proximity and sympathy may steer judges wrong. [FN41] But empathy--when defined as the ability to take the perspective of another-- is a different story. A judge may have a far easier time taking the perspective of the landlord or the police officer from a background similar to his own, than the perspective of the working-class tenant or the po- lice-abuse victim with an arrest record. [FN42] Empathy for the affluent and powerful, like empathy for the poor and the disenfranchised, needs to be called by its rightful name.

EMPATHY: HAVING IT AND GETTING IT

Sol Wachtler, during his tenure as Chief Judge of the New York Court of Appeals, supported the Rockefeller drug laws with their draconian prison terms. After eight months in prison, [FN43] he saw things differently. Cri- ticizing an article about prisons by Abe Rosenthal that he found superficial, Wachtler wrote “I had written and said much the same thing during the twenty-five years I served on the bench. How could I have been so superfi- cial?” [FN44] His experience had shown him that “long prison sentences are not the answer.” [FN45] He mused, “I would like to take Abe for a walk with me through my unit.” [FN46] Actually walking in someone else's shoes is one way to gain perspective about the “world out there” [FN47]--in this case, the hard way.

*143 AN EXAMPLE: THE MIDDLE SCHOOL STRIP SEARCH CASE

There are other ways. Judges also learn from one another, from litigants, from amicus briefs and other sources. Consider, for example, Safford v. Redding, [FN48] the recently decided case about whether the strip search of middle-school student Savana Redding violated the Fourth Amendment. To resolve the Fourth Amend- ment issue, the Court needed to determine how intrusive the search was, how important the government interest was, and whether the government adopted a reasonable means of addressing its concern. To assess how intrusive such a search was, it needed to focus on how it was experienced by the litigant and on how it would be experi- enced by others in her place. To understand the nature of the governmental interest, it needed to put itself in the place of school administrators. Unless the Court could understand the perspectives of all the litigants, it risked making its ultimate determination based on skewed and incomplete information.

In the Safford argument, the justices spent substantial time examining the viewpoint of the school adminis- trators faced with keeping students safe from dangerous drugs. Justices Roberts and Alito asked numerous ques- tions indicating their appreciation of the difficult choice facing an administrator confronted with a tip that a stu- dent is carrying contraband. Here is Justice Souter, for example, imagining what would go through the head of the principal charged with keeping the students safe: “My thought process is I would rather have the kid embar- rassed by a strip search, if we can't find anything short of that, than to have some other kids dead because the stuff is distributed at lunchtime and things go awry.” [FN49]

The Court spent far less time during the argument trying to understand the viewpoint of the victim of the strip search. Apart from Justice Ginsburg, only Justice Breyer raised questions about her perspective. He struggled with the issue, asking for guidance, [FN50] but also consulting his own memory of *144 changing in the locker room to try to gauge the nature of the intrusion. [FN51] Justice Ginsburg pointed out in response that this was no locker room suit-up, but the search of a thirteen-year-old girl forced to strip to her underwear and shake out her bra and underpants in front of school officials who suspected her of concealing prescription ibuprofen.

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 2009 DENOVO 133 Page 7 2009 Cardozo L. Rev. de novo 133

Judges have various means at their disposal for examining their assumptions about how the world works. Justice Breyer, commendably, knew he needed more information about how a thirteen year old girl would exper- ience such a search. The issue was addressed in at least one amicus brief, citing studies on the effects of strip- searching children, [FN52] and it was addressed by counsel. Justice Breyer also consulted his own experience. Justice Ginsburg suggested that this experience did not shed much light on the intrusiveness of the strip search at issue. Empathy is not always accurate, but it can be improved, often with the help of others.

The resulting opinion reflects the Court's effort to educate itself on Redding's perspective. It acknowledges her subjective experience of the search as “embarrassing, frightening, and humiliating.” [FN53] In determining the reasonableness of her expectation, it cites the empirical study mentioned above, which offered evidence of “the consistent experiences of other young people similarly searched, whose adolescent vulnerability intensifies the patent intrusiveness of the exposure.” [FN54] It acknowledges the significant difference between undressing for gym and being subjected to a strip search. [FN55] Finally, the Court rightly notes that “the indignity of the search does not, of course, outlaw it.” [FN56] Its fuller understanding of Redding's perspective allowed a more accurate balancing of interests, *145 but did not resolve the issue of how the balance should be struck.

GETTING IT WRONG

Selective empathy is inevitable. More dangerous is lack of awareness of the limits of individual perspective. Justice Kennedy' s infamous language in Gonzales v Carhart [FN57] illustrates the consequences of this lack of awareness. The majority of the Court “held that a woman's decision to follow her physician's advice can be over- ridden by the government, based on a new principle never advanced or documented by either side in the case: protecting “the bond of love the mother has for her child.” [FN58] Justice Kennedy asserted:

While we find no reliable data to measure the phenomenon, it seems unexceptionable to conclude some women come to regret their choice to abort the infant life they once created and sustained. Severe depression and loss of esteem can follow ... It is self-evident that a mother who comes to regret her choice to abort must struggle with grief more anguished and sorrow more profound when she learns, only after the event, what she once did not know: that she allowed a doctor to pierce the skull and vacuum the fast- developing brain of her unborn child, a child assuming the human form. [FN59] Here Justice Kennedy assumes rather than inquires. Courts are faced with difficult questions about how to properly assess empirical studies, [FN60] and about when they ought to rely on such studies. [FN61] But in this case Justice Kennedy simply failed to seek out accurate information. [FN62] An opinion like this, *146 larded with phrases like “we find no reliable data” and “it seems unexceptionable to conclude” and “it is self-evident,” reflects not just a failure of empathy, but a failure to comprehend the need for it. Or as Jeffrey Rosen tartly ob- served, “Kennedy ... prefers romantic generalizations about ‘real people’ to listening to them.” [FN63]

We all use empathy, and despite our best intentions, it is always selective and riddled with blind spots. We can try to correct for this partiality if we are self-aware. But those who study cognitive psychology and decision- making find that we aren't all that good at identifying and critiquing our own background assumptions. [FN64] A better way to encourage this sort of correction is through debate with others who hold differing viewpoints. Judges, like the rest of us, make better decisions when forced to examine and articulate their assumptions. A range of backgrounds and life experiences on the Court increases the odds that those assumptions are challenged when they are off-base, or at least that no judge assumes his or her own perspective is universal. Supreme Court justices, like the rest of us, make better decisions in an atmosphere of lively debate than in an echo chamber.

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 2009 DENOVO 133 Page 8 2009 Cardozo L. Rev. de novo 133

“THE EMPATHY TO RECOGNIZE WHAT IT'S LIKE TO BE A YOUNG TEENAGE MOM”

Obama's statements about the sorts of judges he hopes to appoint elicited criticism from those who believe that empathy should play no role in judging. But many commentators, including several sophisticated legal scholars, were particularly troubled by Obama's stated preference for judges who would exercise empathy, not simply for all litigants, but for certain groups. Steven Calabresi, for example, expressed concern that *147 “Mr. Obama's emphasis on empathy in essence requires the appointment of judges committed in advance to violating [their oath to administer justice without respect to persons, and do equal right to the poor and the rich].” [FN65]

Assuming that empathy is an important capacity for a judge, it is still reasonable to ask why then-candidate Obama didn't come out in favor of judicial empathy toward all litigants. [FN66] Yet, in fact, the notion of em- pathy for stakes out important ground in the national debate about the rule of law and the role of the Supreme Court.

First, as I have discussed, the context of Obama's initial statement clarifies his message. His point was not that judges should no longer simply be umpires who approach each case without prior conceptions about how the world works or how the law should be interpreted. His point was that no Supreme Court justice actually be- haves this way. Judges have philosophies. Judges are influenced by their background assumptions. A Supreme Court justice should have “a sharp and independent mind and a record of excellence and integrity,” [FN67] but that isn't the only question. Philosophy matters as well. As I've argued elsewhere:

If a reference to judges in the mold of Justices Scalia and Thomas is used and understood as code for “judges who will simply apply the law as written and not impose their own preference” then it becomes a non-ideological act to appoint people sharing this philosophy. Under this understanding, appointing judges like Roberts and Alito is not a political act because their jurisprudence is not a philosophy at all; it is simply proper, unmarked, unbiased judging. [FN68] Thus the very idea of a jurisprudential philosophy is positioned as an activist idea, and the effect is to insu- late candidates like (now) Justices Roberts and Alito from any serious inquiry into their philosophies. Obama's statement rejected *148 this framework. It communicated his beliefs that jurisprudential philosophy is not just the province of the activist judge; that he wanted a different sort of judge; and that discussion on these matters is not only legitimate but essential.

In short, Obama was asserting that the current Supreme Court has been too one-sided in its empathy. At the same time he was articulating his own vision of the Constitution, and of the Court's role in interpreting it. He has said this explicitly. For example:

When you look at what makes a great Supreme Court justice, it's not just the particular issue and how they rule, but it's their conception of the Court. And part of the role of the Court is that it is going to pro- tect people who may be vulnerable in the political process, the outsider, the minority, those who are vul- nerable, those who don't have a lot of clout ... If we can find people who have life experience and they un- derstand what it means to be on the outside, whit it means to have the system not work for them, that's the kind of person I want on the Supreme Court. [FN69] We can and should debate whether we want our next Supreme Court justice to share the President's values. The important thing is not to get sidetracked, as we have in the past, by the notion that Supreme Court justices are mere technocrats who can simply apply the law without making value choices. That notion leads only to con- firmation hearings in which neither Congress nor the American people are given the information they need to

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 2009 DENOVO 133 Page 9 2009 Cardozo L. Rev. de novo 133

evaluate candidates for the highest court in the land.

[FNa1]. Susan A. Bandes is a Professor of Law at the DePaul University College of Law. I first discussed some of the points made in this article in two blog posts: Empathetic Judges and the Rule of Law, AM. CONST. SOC., May 20, 2009, available at http://www.acslaw.org/node/13450, and Why is Empathy Controversial? Or Liberal?, BALKINIZATION, May 25, 2009, available at http:// balkin.blogspot.com/2009/05/why-is-empathy-controversial-or-liberal.html.

[FN1]. Answers will be discussed throughout this essay. See also The “Empathy” Exam, ht- tp://althouse.blogspot.com/2009/06/empathy-exam.html (June 13, 2009, 10:29AM) (Ann Althouse's Constitu- tional Law exam question on identifying empathy).

[FN2]. 163 U.S. 537, 551 (1896).

[FN3]. 489 U.S. 189, 213 (Blackmun, J., dissenting).

[FN4]. 550 U.S. 124, 159-160 (2007).

[FN5]. Transcript of Oral Argument at 43-44, Safford Unified Sch. Dist. #1 v. Redding, No. 08-479, 2009 WL 1789472 (Apr. 21, 2009) available at http:// www.supremecourtus.gov/oral_arguments/argument_transcripts/08-479.pdf.

[FN6]. 527 U.S. 706, 749 (1999).

[FN7]. And to compound the definitional problem, President Obama's use of the term shifted subtly in later ad- dresses on the subject. See infra note 66.

[FN8]. Judge Posner observes that “[n]either [Roberts] nor any other knowledgeable person actually believed or believes that the rules that judges in our system apply, particularly appellate judges and most particularly the justices the U.S. Supreme Court, are given to them the way the rules of baseball are given to umpires.” RICHARD A. POSNER, HOW JUDGES THINK 78 (2008).

[FN9]. Susan Bandes, We Lost it at the Movies: The Rule of Law goes from Washington to Hollywood and Back Again, 40 LOY. L.A. L. REV. 621, 621, 622 n.6 (citing Roberts confirmation hearings).

[FN10]. Id. at 621, 622 n.2 (citing Alito confirmation hearings).

[FN11]. At the same time, the popular understanding of the rule of law is oddly schizophrenic. This so called “ideal” conception “works in tandem with the conception of law as ‘a pragmatic, perhaps vulgar, account of the routine practices of biased, differentially endowed, and fallible actors.”’ Id. at 648 (citing PATRICIA EWICK & SUSAN SILBEY, THE COMMON PLACE OF LAW: STORIES FROM EVERYDAY LIFE 226-27 (1998)). See also Keith J. Bybee, The Rule of Law is Dead! Long Live the Rule of Law (Mar. 27, 2009), available at ht- tp:// papers.ssrn.com/sol3/papers.cfm?abstract_id=1404600.

[FN12]. Senator Barack Obama, Campaign speech to Planned Parenthood (July 2007) (cited in William Safire, Zombie Banks, N.Y. TIMES, May 17, 2009 at 26).

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 2009 DENOVO 133 Page 10 2009 Cardozo L. Rev. de novo 133

[FN13]. See Susan Bandes, Empathy, Narrative and Victim Impact Statements, 63 U. CHI. L. REV. 361, 373 & n.52 (1996) (discussing the term empathy as a moving target).

[FN14]. PAUL BLOOM, DESCARTES' BABY 113 (2004); CANDACE CLARK, MISERY AND COMPANY: SYMPATHY IN EVERYDAY LIFE 34 (1997).

[FN15]. CLARK, supra note 14. Clark distinguishes among several types of empathy: cognitive, emotional and physical. Id. at 38. Cognitive empathy is perspective-taking. The rarer emotional empathy, a kind of emotional contagion, “makes us experience others' suffering as our own.” Id. Others make different distinctions. The safest course of action with empathy is to clarify one's use of the term at the outset.

[FN16]. Id. at 44-45.

[FN17]. BLOOM, supra note 14, at 118.

[FN18]. Conversely, it is possible to feel compassion or pity for someone without really understanding his per- spective. DANIEL GOLEMAN, SOCIAL INTELLIGENCE: THE NEW SCIENCE OF HUMAN RELATION- SHIPS 62 (2006).

[FN19]. Michael Franz Basch, Empathic Understanding: A Review of the Concept and Some Theoretical Con- siderations, 31 J. AM. PSYCHOANALYTIC ASS'N 101, 119, 123 (1983) (emphasis added).

[FN20]. POSNER, supra note 8, at 117.

[FN21]. 137 Cong. Rec. S14283-03 (1991).

[FN22]. Id.

[FN23]. Meeting of the Senate Judiciary Committee Subject: The Nomination of Samuel Alito to the Supreme Court, FEDERAL NEW SERVICE, Jan. 24, 2006.

[FN24]. Jessica Weisner, Define Empathy: The Next SCOTUS Pick, TALKING POINTS MEMO, May 22, 2009 (quoting Wendy Long of the Judicial Confirmation Network), available at http:// tpm- cafe.talkingpointsmemo.com/talk/blogs/rac/2009/05/define-empathy-the-next-scotus.php.

[FN25]. JAAK PANKSEPP, AFFECTIVE NEUROSCIENCE: THE FOUNDATIONS OF HUMAN AND AN- IMAL EMOTIONS 276 (1998).

[FN26]. William D. Casebeer, Moral Cognition and its Neural Constituents, 4 NATURE REVS. NEUROS- CIENCE 840, 844 (2003).

[FN27]. GOLEMAN, supra note 18, at 135.

[FN28]. Casebeer, supra note 26, at 844.

[FN29]. BLOOM, supra note 14, at 28-29.

[FN30]. John Seabrook, Suffering Souls: The Search for the Roots of Psychopathy, THE NEW YORKER, Nov. 10, 2008, available at http://

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 2009 DENOVO 133 Page 11 2009 Cardozo L. Rev. de novo 133

www.newyorker.com/reporting/2008/11/10/081110fa_fact_seabrook?currentPage=1.

[FN31]. Bandes, supra note 13, at 377; Robert A. Ferguson, The Judicial Opinion as Literary Genre, 2 YALE J.L. & HUMAN. 201, 204-208 (1990).

[FN32]. THOMAS NAGEL, THE VIEW FROM NOWHERE (1989).

[FN33]. Jeffrey Toobin, No More Mr. Nice Guy: The Supreme Court's Stealth Hard-Liner, THE NEW YORK- ER, May 25, 2009, at 42.

[FN34]. Amy Goldstein and R. Jeffrey Smith, Midwestern Scholar with a Steady Conservative Bent, WASH. POST, Sept. 4, 2005, at A06. I thank Mary Ann Case for bringing this article to my attention.

[FN35]. Toobin, supra note 33.

[FN36]. Id.

[FN37]. Weisner, supra note 24 (quoting Wendy Long).

[FN38]. Butz v. Economou, 438 U.S. 478, 529 (1978) (Rehnquist, J., dissenting).

[FN39]. Susan Bandes, Fear and Degradation in Alabama: The Emotional Subtext of University of Alabama v. Garrett, 5 U. PA. J. CONST. L. 520 (2003).

[FN40]. John Hasnas, The Unseen Deserve Empathy Too, WALL ST. J., May 29, 2009, at A17, available at ht- tp:// online.wsj.com/article/SB124355502499664627.html.

[FN41]. On this point see also Susan Bandes, Emotions, Values, and the Construction of Risk, 156 U. PA. L. REV. PENNUMBRA 421, 426 (2008), http:// www.pennumbra.com/responses/03-2008/Bandes.pdf. Hasnas' in- teresting central argument about proximity deserves more space than I can give it here.

[FN42]. Susan Bandes, Patterns of Injustice: Police Brutality in the Courts, 47 BUFF. L. REV. 1275, 1321 (1999); John Conroy, Town Without Pity, THE CHICAGO READER, Jan. 12, 1996, at 22 (noting that a judge may have an easier time identifying with an erect and courageous torturer than with an unpopular victim).

[FN43]. SOL WACHTLER, AFTER THE MADNESS: A JUDGE'S OWN PRISON MEMOIR 263 (1997) (telling the story of Wachtler's eleven month imprisonment after pleading guilty to harassing Joy Silverman).

[FN44]. Id. at 263.

[FN45]. Id. at 266.

[FN46]. Id. at 264.

[FN47]. Beal v. Doe, 432 U.S. 438, 463 (1977).

[FN48]. Safford Unified School District #1 v. Redding, No. 08-479, 2009 WL 1789472 (April 21, 2009).

[FN49]. Transcript of Oral Argument, Safford Unified Sch. Dist. #1 v. Redding, No. 08-479, 2009 WL 1789472 (Apr. 21, 2009), available at http:// www.supremecourtus.gov/oral_arguments/argument_transcripts/08-479.pdf.

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 2009 DENOVO 133 Page 12 2009 Cardozo L. Rev. de novo 133

[FN50]. Justice Breyer: I'm trying to work out why is this a major thing to say strip down to your underclothes, which children do when they change for gym, they do fairly frequently, not to--you know, and there are only two women there. Is--how bad is this, underclothes? That's what I'm trying to get at. I'm asking because I don't know. Id.

[FN51]. Justice Breyer: In my experience when I was 8 or 10 or 12 years old, you know, we did take our clothes off once a day, we changed for gym, okay? And in my experience, too, people did sometimes stick things in my underwear - (laughter) Or not my underwear. Whatever. Whatever. Id.

[FN52]. Brief for National Association of Social Workers et al. as Amici Curiae, Safford Unified Sch. Dist. #1 v. Redding, No. 08-479, 2009 WL 1789472 (Apr. 21, 2009).

[FN53]. Safford Unified Sch. Dist. #1 v. Redding, No. 08-479, 2009 WL 1789472 (Apr. 21, 2009).

[FN54]. Brief for National Association of Social Workers et al., supra note 52.

[FN55]. The common reaction of these adolescents simply registers the obviously different meaning of a search exposing the body from the experience of nakedness or near undress in other school circumstances. Changing for gym is getting ready for play; exposing for a search is responding to an accusation reserved for suspected wrongdoers and fairly understood as so degrading that a number of communities have decided that strip searches in schools are never reasonable. Safford, 2009 WL 1789472, at *7.

[FN56]. Id.

[FN57]. Gonzales v. Carhart, 550 U.S. 124 (2007).

[FN58]. National Women's Law Center, Gonzales v. Carhart: The Supreme Court Turns Its Back on Women's Health and on Three Decades of Constitutional Law, May 2007, available at ht- tp://www.nwlc.org/pdf/GonzalesvCarhart2.pdf.

[FN59]. Carhart, 550 U.S. at 159-160.

[FN60]. See e.g., John Donohue, Have “Woman-Protective” Studies Resolved the Abortion Debate? Don't Bet on It, BALKINIZATION, Sept. 23, 2008, http:// balkin.blogspot.com/2008/09/have-woman-protective-studies-resolved.html (arguing that “judges and legislators must insist on greater methodological sophistication before empirical studies can illuminate legal and policy choices”). A discussion of these difficult issues is beyond the scope of this short essay.

[FN61]. Moreover, the Court may disagree on what empirical question is at issue. See for example the opinions in Kennedy v. Louisiana, the case involving whether the death penalty may be imposed for child rape. Both Justice Kennedy's majority opinion and Justice Alito's dissent cited empirical studies, but the former focused on the effects of a death penalty prosecution on the child victim, 128 S. Ct. 2641, 2662 (2008), and the latter fo- cused on the effects of the rape itself, 128 S. Ct. 2641, 2677 (2008) (Alito, J., dissenting).

[FN62]. See e.g., Nada L Stotland, MD, MPH, The Woman-Protective Strategy as a Campaign of Misinforma- tion, BALKINIZATION, Sept. 17, 2008, available at ht- tp://balkin.blogspot.com/2008/09/woman-protective-strategy-as-campaign.html; Terry A. Maroney, Emotional Common Sense as Constitutional Law, 62 VAND. L. REV. 851, 888-901 (2009); see generally Chris Guthrie,

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 2009 DENOVO 133 Page 13 2009 Cardozo L. Rev. de novo 133

Carhart, Constitutional Rights, and the Psychology of Regret, 81 S. CAL. L. REV. 877 (2008).

[FN63]. Jeffrey Rosen, Supreme Leader: The Arrogance of Justice Anthony Kennedy, THE NEW REPUBLIC, June 18, 2007, at 16, 17.

[FN64]. See generally Jonathan Haidt & Fredrik Bjorklund, Social Intuitionists Answer Six Questions About Moral Psychology, in 2 MORAL PSYCHOLOGY 181 (W. Sinnott-Armstrong ed., 2008), available at http:// ss- rn.com/abstract=855164 (arguing that moral judgment is a social process); Kevin M. Carlsmith & John M. Dar- ley, Psychological Aspects of Retributive Justice, 40 ADVANCES IN EXPERIMENTAL SOCIAL PSYCHO- LOGY 193 (2008), available at http://ssrn.com/abstract=1031193 (arguing that discussion with others helps us identify the flaws in our own assumptions).

[FN65]. Steven G. Calabresi, Obama's Redistribution. Constitution, WALL ST. J., Oct. 28, 2008, at A17.

[FN66]. His later statements were less definitive about the need to have empathy for certain groups For example, just before announcing the Sotomayor nomination, he said: “I will seek somebody who understands that justice isn't about some abstract legal theory or footnote in a case book; it is about how our laws affect the daily realit- ies of people's lives, whether they can make a living and care for their families, whether they feel safe in their homes and welcome in their own nation. I view that quality of empathy, of understanding and identifying with people's hopes and struggles, as an essential ingredient for arriving at just decisions and outcomes.” Robert Gibbs, White House Press Secretary, Press Briefing (May 1, 2009), quoted in Joseph Williams, Obama may break with tradition for high court pick; Retirement of Souter opens opportunities, BOSTON GLOBE, May 2, 2009, at 8.

[FN67]. Id.

[FN68]. Bandes, We Lost it at the Movies, supra note 9, at 648.

[FN69]. Senator Barack Obama, Democratic Debate in Las Vegas (Nov. 15, 2007), available at ht- tp://www.nytimes.com/2007/11/15/us/politics/15debate-transcript.html. 2009 Cardozo L. Rev. de novo 133

END OF DOCUMENT

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works.

160 UPALR 1561 Page 1 160 U. Pa. L. Rev. 1561

University of Pennsylvania Law Review May, 2012

Symposium: Sentencing Law: Rhetoric and Reality

Essay

*1561 SENTENCING: A ROLE FOR EMPATHY

Denny Chin [FNd1]

Copyright (c) 2012 University of Pennsylvania Law Review; Denny Chin

Introduction 1562

I. The Process 1566

II. Substantive Considerations 1572

III. Emotion and Empathy in Senten- 1576 cing

Conclusion 1582

*1562 Introduction

When Supreme Court Justice David H. Souter announced his retirement in 2009, President Barack Obama observed that “empathy”--which he defined as the quality of “understanding and identifying with people's hopes and struggles”--was an “essential ingredient” for a judge. [FN1]

The reaction from some was swift and caustic. Empathy, some said, was a code word for activism--for activist judges. [FN2] Indeed, Michael Steele, Chairman of the Republican National Committee at the time, was quoted as say- ing, “I don't need some judge sitting up there feeling bad for my opponent because of their life circumstances or their condition. And short changing me and my opportunity to get fair treatment under the law. Crazy nonsense empathetic. I'll give you empathy. *1563 Empathize right on your behind. Craziness.” [FN3] After the President's remarks, some mem- bers of the Senate Judiciary Committee began asking judicial candidates about their views on the role empathy should play in a judge's consideration of a case, apparently to identify candidates who might let empathy creep into their de- cisionmaking. [FN4]

Is empathy an important trait for a judge? Is there a role for empathy in the law? What about the related concept of emotion? Is it correct that “[a] good judge should feel no emotions” and that “the ideal judge is divested ‘of all fear[], an- ger, hatred, love, and compassion?”’ [FN5]

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 2 160 U. Pa. L. Rev. 1561

Empathy, of course, should play no role in a judge's determination of what the law is. [FN6] Judges determine the law based on statutes, case law, and legal principles, guided by the rule of stare decisis. With statutory questions, we look to the plain wording of the governing statute (or sentencing guideline), and, where there is ambiguity, we apply the rules of statutory construction and seek to ascertain legislative intent. [FN7] *1564 We do not determine the law or decide cases based on “feelings” or emotions or whether we empathize with one side or the other. We instruct our juries that they “are not to be swayed by sympathy,” [FN8] and we tell them that “once you let fear or prejudice[] or bias or sym- pathy interfere with your thinking there is a risk that you will not arrive at a true and just verdict.” [FN9] These concepts apply to judges as well.

Nonetheless, there is a place within the law for empathy and emotion. In my view, empathy is an essential character- istic for a judge. [FN10] Despite the rhetoric, the reality is that empathy and emotion play an essential role in the real- world, day-to-day administration of justice--particularly in sentencing. [FN11] And we should be clear: by “empathy” I do not mean “sympathy.” I do not mean feeling “bad” or “sorry” for someone and letting that emotion influence the de- cisionmaking. Rather, by empathy I mean the capacity to understand and appreciate the perspective of others, [FN12] whether that perspective is of individuals *1565 trying to “make a living or care for their families” [FN13] or corpora- tions required to defend themselves against frivolous claims brought by vexatious litigants. [FN14] We were selected to be judges because of our experiences in life, and because of the wisdom, good judgment, and sense of justice that hope- fully we have developed as a result of those experiences. It would make no sense for us to set aside these attributes once we reach the bench.

In this Essay, I will consider the role of empathy and emotion in sentencing. I do so from the perspective of someone who has sentenced hundreds of individuals. I was a trial judge for almost sixteen years, during which time I was assigned 699 criminal cases with 1256 defendants. [FN15] The vast majority were convicted--after a guilty plea or trial--and I was required to sentence them. [FN16] In doing so, I came to understand and appreciate the importance of empathy and emo- tion in sentencing. [FN17]

Sentencing involves both process and substance. First, a sentencing court must comply with all procedural require- ments. [FN18] Second, the *1566 sentencing court must also impose a sentence that is substantively reasonable and falls “within the range of permissible decisions.” [FN19] The sentencing judge must take certain substantive factors into ac- count. A failure to do so will constitute procedural error [FN20] and may also lead to a sentence that is substantively un- reasonable. Consideration of both the procedural and the substantive aspects of sentencing is important to determine the proper role, if any, of empathy and emotion in sentencing. Accordingly, first, I discuss the sentencing process; second, I discuss the substantive considerations that bear on the sentencing decision; and, third, I discuss the role of empathy and emotion in sentencing.

I. The Process

A defendant is entitled to a sentencing process that meets the procedural requirements of law--including those set forth in section 3553 of the Sentencing Reform Act of 1984, [FN21] the Sentencing Guidelines, [FN22] and the Federal Rules of Criminal Procedure. [FN23] Compliance with these requirements helps to ensure a sentence that is both proced- urally fair and substantively reasonable.

The court is required to impose sentence “without unnecessary delay.” [FN24] At the guilty plea or following a guilty verdict, the court will typically schedule sentencing approximately ninety days later. In most cases, the Probation Depart- ment will conduct a presentence investigation, [FN25] which will include an interview of the defendant. [FN26] The Pro- bation*1567 Department will issue a presentence report (PSR) that will contain extensive information about the defend-

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 3 160 U. Pa. L. Rev. 1561

ant, including his or her personal background (e.g., education, employment, financial information, and family history) and the details of the crime of conviction. It will contain a Guidelines analysis that calculates the offense level, criminal history category, and resulting sentencing range. [FN27]

The Probation Department must disclose the draft PSR to the defendant, his or her attorney, and the government at least thirty-five days before sentencing. [FN28] The defendant has an opportunity to object to the Probation Department, [FN29] which must then consider the objections and submit a final PSR to the court at least seven days before senten- cing. [FN30] After consulting with both defense counsel and the government, the Probation Department will typically re- commend a specific sentence in the final PSR.

The lawyers will also make written submissions, including sentencing memoranda--briefs addressing legal issues or advocating for a certain sentence, such as time served, a below-Guidelines sentence, or a within-Guidelines sentence. De- fense counsel may also submit exhibits that provide information about the defendant, including medical records, psychi- atric reports, and photographs of family members. Additionally, defense counsel will submit letters from the defendant, family members, [FN31] employers, and others in support of the defendant. These letters are important and can be effect- ive. [FN32] The government will also, on occasion, submit letters from victims and victim impact statements, or docu- ments that provide details of the crime. [FN33]

*1568 At times, there may be factual disputes that can have a bearing on sentencing, such as the amount of loss in a fraud case [FN34] or whether a victim is a “vulnerable victim.” [FN35] In these situations, the district court will conduct an evidentiary hearing--a Fatico hearing [FN36]--at which witnesses testify and counsel presents evidence. [FN37]

Finally, there is the sentencing hearing, the culmination of the sentencing process. The defendant, defense counsel, and the government will be present. Often, the defendant's family and supporters will be there, and their presence is im- portant as it is some indication that the defendant will return to a supportive environment. [FN38] On occasion, victims will attend as well. Of course, the proceedings are public, and members of the public--including representatives of the media--may also observe.

The sequence may vary from courtroom to courtroom, but victims, defense counsel, the defendant, and the govern- ment all have a right to be heard. [FN39] In my experience, victims rarely exercised their right to address*1569 the court at the sentencing hearing. [FN40] Even though there were thousands of victims in the Madoff case, for example, only nine spoke at the sentencing. [FN41] When victims do speak, however, there is much emotion. [FN42] I remember vividly a murder case where the daughter of the victim spoke. She explained that her father had not been present for her graduation from high school, he was going to miss her attending college, and he would not be around when she got mar- ried and had children. Yet, she said to the defendant, “[E]ven after all that pain and anger that I have inside, even after being daddy's little girl, I forgive you. . . . I'm going to pray for you, because that's the way I was raised.” [FN43]

Defense counsel always speaks. In fact, after the Supreme Court held in United States v. Booker that the Sentencing Guidelines were advisory*1570 only, [FN44] I found that defense counsel talked longer at sentencings. This is a good thing in my view. The lawyers are trying harder because they know they now have a better chance of obtaining a below- Guidelines sentence for their client. In fact, the role of a defense lawyer has changed somewhat since Booker. Before, when involved in plea-bargaining and sentencing, defense counsel was more of a tactician, trying to take advantage of the intricacies and technical aspects of the Guidelines. Although these matters are still important after Booker, defense counsel now has a greater opportunity to persuade a judge to impose a below-Guidelines sentence with eloquent oration- -telling a compelling story that persuades the sentencing judge that the defendant is deserving of a lower sentence.

Most defendants will take advantage of their opportunity to speak. Some defendants remain silent. Although it is

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 4 160 U. Pa. L. Rev. 1561

their prerogative to decline to speak, in my view it is a mistake. The defendant's statement to the court is important as it helps the judge try to answer a number of questions: What was the defendant thinking when he committed his crime? Is he remorseful? Is he sincere? What are the chances he will turn his life around? Should I give him another chance? The way a defendant speaks, what he says, and how he says it may help point me to these answers. [FN45]

The government usually will address the court, [FN46] although often, in my experience, it will only urge that the court impose a sentence within the applicable Guidelines range. On occasion, the government will advocate for a sen- tence at the top of the range. [FN47]

*1571 The court has responsibilities at the sentencing hearing. It must make certain findings before imposing sen- tence, calculate the applicable Guidelines range, [FN48] and rule on any objections or disputed matters that may affect sentencing. [FN49] And then comes the moment when the defendant is asked to rise, and the court imposes a sentence. When I sentenced Patrick Regan, a highly decorated police officer convicted of perjury, the courtroom was filled with police officers in support. As Regan stood to be sentenced, in the back of the courtroom, first one police officer, then an- other, and then virtually the entire courtroom rose in support. [FN50]

The court is required to “state in open court the reasons for its imposition of the particular sentence” [FN51] as well as provide an explanation sufficient “to allow for meaningful appellate review.” [FN52] The court also has an opportun- ity to address the defendant directly. It was not my style to preach or scold the defendant, but if I felt a harsh sentence was warranted, I did not hesitate to impose one and to explain why. Usually, but not always, I tried to say something pos- itive to the defendant after I imposed sentence so as to encourage him going forward. If I thought a defendant was bright and articulate and had potential to turn his life around, I would say so. [FN53] I would suggest, for example, that he not let his family down again.

*1572 As a result of the process--the presentence investigation and the preparation of the PSR, the submission of sen- tencing memoranda, and other documentary support, and the presentation of arguments and statements at the sentencing hearing--we usually have a great deal of information about the defendant. We are able to make an informed decision, and we are able to sentence a defendant, not just for the crime he committed, but also for who he is, who he has been, and who he may be in the future. [FN54]

II. Substantive Considerations

In sentencing a defendant, a court must take a number of substantive considerations into account. These criteria are found in the Sentencing Guidelines, the Sentencing Reform Act, and the case law.

The sentencing court must first consider the Guidelines. The court must begin the sentencing proceeding “by cor- rectly calculating the applicable Guidelines range.” [FN55] The court must ascertain the base offense level, make any ap- propriate adjustments to arrive at the total offense level, determine the criminal history category, consider whether any mandatory minimums apply, and, finally, calculate the advisory Guidelines range.

Although the Guidelines are now advisory only, sentencing judges are not free to ignore them or treat them “merely as a ‘body of casual advice.”’ [FN56] As sentencing law has evolved since I first became a judge, [FN57] *1573 district judges have gained greater discretion and flexibility, and they are now free even to reject a particular Guideline based on personal policy disagreements. [FN58] Although one could argue that the Guidelines have lost their significance under these circumstances, in my view they still play a critical role as they provide an enormously helpful starting point. It is useful to begin with an empirically based “heartland” range that is drawn from the collective wisdom and experiences of

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 5 160 U. Pa. L. Rev. 1561

colleagues from around the country, [FN59] but as a consequence of Booker, we now have much greater ability to fash- ion a just sentence. The Guidelines are now as they should be-- true guidelines, advisory and not mandatory in nature. [FN60] They are something to which we should give “respectful consideration.” [FN61]

Once the applicable Guidelines range is determined, the sentencing court “must give serious consideration” to wheth- er the circumstances warrant an above- or below-Guidelines sentence. [FN62] The court “shall consider” what are known as the “statutory factors”:

*1574 (1) the nature and circumstances of the offense and the history and characteristics of the defendant; (2) the need for the sentence imposed-- (A) to reflect the seriousness of the offense, to promote respect for the law, and to provide just punishment for the offense;

(B) to afford adequate deterrence to criminal conduct;

(C) to protect the public from further crimes of the defendant; and

(D) to provide the defendant with needed educational or vocational training, medical care, or other correc- tional treatment in the most effective manner . . . . [FN63]

The statute provides that the sentencing court “shall impose a sentence sufficient, but not greater than necessary, to comply with the purposes set forth in paragraph (2)” of § 3553(a). [FN64] The “sufficient, but not greater than neces- sary” language has become a common refrain of defense counsel.

The statutory factors reflect the traditional goals of punishment: retribution, deterrence, incapacitation, and rehabilit- ation. [FN65] In sentencing Bernard L. Madoff for securities fraud, I relied heavily on two of these traditional goals, de- terrence and retribution. Although Mr. Madoff was seventy-one years old, [FN66] I imposed a sentence of 150 years nonetheless because “the symbolism [was] important.” [FN67] I stated:

One of the traditional notions of punishment is that an offender should be punished in proportion to his blame- worthiness. Here, the message must be sent that Mr. Madoff's crimes were extraordinarily evil, and that this kind of irresponsible manipulation of the system is not merely a bloodless financial crime that takes place just on paper, but that it is instead,*1575 as we have heard, one that takes a staggering human toll. The symbolism is important because the message must be sent that in a society governed by the rule of law, Mr. Madoff will get what he de- serves, and that he will be punished according to his moral culpability. [FN68] I also had in mind an objective not included among the traditional goals of punishment--helping victims heal. [FN69] I said:

[T]he symbolism is also important for the victims. The victims include individuals from all walks of life. The victims include charities, both large and small, as well as academic institutions, pension funds, and other entities. Mr. Madoff's very personal betrayal struck at the rich and the not-so-rich, the elderly living on retirement funds and social security, middle class folks trying to put their kids through college, and ordinary people who worked hard to save their money and who thought they were investing it safely, for themselves and their families. .... A substantial sentence will not give the victims back their retirement funds or the moneys they saved to send their children or grandchildren to college. It will not give them back their financial security or the freedom from financial worry. But more is at stake than money, as we have heard. . . .

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 6 160 U. Pa. L. Rev. 1561

. . . A substantial sentence, the knowledge that Mr. Madoff has been punished to the fullest extent of the law, may, in some small measure, help these victims in their healing process. [FN70] The abstract traditional goals of punishment and the statutory factors are also reflected in more concrete terms in a host of questions that we ask ourselves in virtually every sentencing: How serious was the criminal conduct? What was the loss to the victims and society, both monetary and otherwise? How culpable was the defendant, and what was his role? What motivated him to break the law? Were there any mitigating factors? What is his state of mind now? Is he re- morseful? Did he break the law before? Is he likely to do it again? Does he deserve another chance? Will he redeem him- self? Did he otherwise lead a productive life, and was this an aberrational mistake? What is the impact on the defendant's family?

*1576 Of course, some of these factors conflict. A sentencing judge, for example, should want to give the defendant and his or her family some hope that he or she will return in time for them to continue as a family; [FN71] yet, the goals of retribution and deterrence may call for stronger punishment. Notions of retribution and deterrence often tug in a differ- ent direction from the goal of rehabilitation. To some extent, a sentencing judge is also trying to predict the future, as she must determine whether the defendant really learned his lesson, whether he is likely to break the law again, and whether he will finally turn his life around. [FN72]

III. Emotion and Empathy in Sentencing

The reality is, of course, that sentencings are almost always emotional, and often highly so. I have three examples of unusually emotional sentencings.

First, while sentencings are usually somber, I recall one sentencing that was a happy occasion. The defendant was a crack addict. She was a cooperator and had testified against the leaders of a narcotics trafficking ring. At trial, she looked like the junkie she was, sitting on the witness stand, withered and frail, anxiously testifying in her prison jumpsuit. Some months later, after the trial, she appeared for sentencing. She had undergone drug treatment and was clean of drugs. She wore a *1577 pretty, bright yellow dress, and she looked great. And her lawyer brought a bouquet of flowers to give to her after the sentencing. He brought flowers because the sentencing had become a celebration--the defendant had turned her life around, and everyone expected that I would give her time served. Indeed, that is what I did. She would not have to return to jail, and she could continue her new life.

A year later, she was back in my courtroom. The drugs had reclaimed her, and she was brought back for violating the terms of her supervised release. [FN73]

Second, in another case, I was accused by defense counsel of being too emotional. The defendant had initially pled guilty to passport fraud. Before he was sentenced, the terrorist attacks of September 11, 2001, took place. The defendant engaged in an “elaborate scheme” to avoid further prosecution by pretending that he had been killed in the World Trade Center attack. [FN74] His scheme was discovered, and he was charged with and pled guilty to additional crimes- -bail-jumping and obstruction of justice. [FN75] At sentencing, I upwardly departed and imposed an above-Guidelines sentence of forty-eight months imprisonment--in part because I found his conduct “despicable” and “com-plete[ly] selfish[]” at a time of enormous tragedy and stress. [FN76] Defense counsel objected and asked me to reconsider, sug- gesting that I had let emotion unduly affect my sentence. I responded:

I don't believe that emotion is unduly affecting my judgment here. There is no doubt emotion comes into play to some extent. Emotion comes into play in every sentencing decision. . . . Emotion comes into play when the Court downwardly departs or the Court shows compassion and imposes a sentence that is lower than one would

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 7 160 U. Pa. L. Rev. 1561

expect, and that happens. When you were a prosecutor [addressing defense counsel] in a case before me and I sen- tenced the defendants to what I thought was a relatively high sentence, emotion came into play. Emotion always does. Obviously, however, you can't let emotion cloud your judgment, and I don't believe I have done that here. [FN77] *1578 The Second Circuit affirmed this aspect of the sentence, finding “no fault with the District Court's decision to depart.” [FN78]

Third, in what was perhaps my most difficult sentencing, I had to sentence a woman who was raising six children: five of her own, ages five to thirteen (by three different fathers, none of whom was available to take custody of the chil- dren), and her fourteen-year-old sister (their mother was a crack addict). [FN79] But the defendant had pled guilty to bank robbery--she was a lookout for a violent bank robbery that led to a high-speed chase, a shoot out between the two bank robbers and the police, and the death of one of the assailants. [FN80] The Guidelines range was fifty-seven to sev- enty-one months, [FN81] and because this was before Booker, I had much less flexibility. Defense counsel asked for a noncustodial sentence, arguing that any significant prison sentence would mean that the six children would be placed in foster care. [FN82] The decision was particularly difficult because the defendant was a good mother--she was working and managing to raise six children on her own, with the children doing reasonably well in school (maintaining B or B+ averages). [FN83]

The defendant's family circumstances surely were extraordinary, [FN84] but I still had to decide whether to depart from the Guidelines, and, if so, to what extent. [FN85] I was deeply concerned about the children and their future, and I understood the struggles of the family--in other words, I empathized with them. [FN86] As I wrote in my decision grant- ing the downward departure motion, “[A] sentence within the Guidelines range of 57 to 71 months would almost cer- tainly and irreparably destroy the family as a unit.” [FN87] Still, the defendant participated in a bank robbery, and I felt that a sentence of probation or home confinement *1579 would send “the wrong message” leaving the defendant “utterly unpunished.” [FN88] I decided some imprisonment was necessary. I wrote: “The Court will endeavor to impose a sen- tence that not only furthers the goals of retribution and deterrence, but that also will give [the defendant] and her children some hope that they will be able to continue as a family unit upon her release.” [FN89] In the end, I sentenced her to thirty months in prison. [FN90] I wanted to give the defendant and her family some hope, because without hope a de- fendant has little reason to want or try to do better. [FN91]

Some years later, the New York Times managed to track down the defendant, after she had completed her sentence. She reported that when she went to prison, friends and relatives--including her mother, who rose to the occasion and dealt with her own drug problem--“stepped in” to help take care of the children. [FN92] After her release, the family was able to stay together. She found employment and was also studying nursing. She told the Times, “I feel like I got that second chance that everybody's talking about. . . . And I'm taking full advantage of that.” [FN93]

Sentencing is perhaps the most important responsibility of a trial judge, and surely the most difficult. [FN94] Emo- tion is one reason it is so difficult. The competing considerations evoke strong sentiments--anger, indignation, shame, sorrow, grief, despondency, and hope. The sentencing judge is not immune from these emotions.

The law is not emotionless, as some would suggest. [FN95] Rather, it recognizes that some emotion and passion are appropriate. [FN96] As Judge *1580 Learned Hand observed, “It is impossible to expect that a criminal trial shall be con- ducted without some show of feeling; the stakes are high, and the participants are inevitably charged with emotion.” [FN97] The same is true, of course, for sentencing.

There is no “right” answer as to what a particular sentence should be; rather, there usually is a range of acceptable

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 8 160 U. Pa. L. Rev. 1561

sentences, and often that range is quite wide. [FN98] Some measure of emotion helps judges reach the right answer--or at least the more correct answer. [FN99] In short, a judge is more likely to reach a just answer if he or she cares. [FN100]

Emotion and empathy cut both ways. They are important both to the prosecution and to the defense. It is not simply a matter of making the judge get angry at or feel sorry for the defendant for what he or she has done. Rather, emotion- -some emotion, emotion both *1581 ways, emotion not alone but in combination with the law, logic, and reason--helps judges get it right. [FN101]

Empathy is particularly important when it comes to sentencing, when a judge is called upon to pass judgment on an- other human being. As Judge Jack B. Weinstein has observed, “Sentencing . . . turns on the judge's heart and life experi- ence. It reveals the human face of the law. Without empathy between judge and defendant, sentencing lacks humanity. It becomes a form of robotism.” [FN102] When confronted, for example, with the question of whether to send a mother to jail and take her away from her family, we will not find the answer in a book or statute or case. We must call upon our life experiences and the wisdom and judgment that hopefully we have gained as we weigh competing considerations to arrive at a just and fair sentence. [FN103] The ability to have some understanding of the defendant's motivations and “hopes and struggles” can only help in that endeavor. [FN104]

Edward Devitt, a highly regarded federal judge, wrote a guide for new federal judges half a century ago, entitled Ten Commandments for the New Judge. [FN105] He published a slightly revised version in 1979. [FN106] The first of the Commandments was: “Be kind.” [FN107] Judge Devitt explained,

If we judges could possess but one attribute, it should be a kind and understanding heart. The bench is no place for cruel or callous people regardless*1582 of their other qualities and abilities. There is no burden more onerous than imposing sentence in criminal cases. Would then that the judge had the wisdom of Solomon. But ab- sent that, and possessing plenary and awesome power, the judge can thank God for a kindly heart. An understand- ing heart was the gift of God asked by the ancient king, and it is that gift above all others for which a judge should pray. [FN108]

Conclusion

I conclude by discussing one more case. In 2006, Vernon Lawson applied to be naturalized as a U.S. citizen. [FN109] The government denied the application, and he sought review in the Southern District of New York. [FN110] I heard the case in the district court. The sole question was whether Mr. Lawson was a person of “good moral character,” [FN111] as the government agreed he met all the other requirements to become an American citizen. [FN112] This was a military naturalization case, [FN113] and thus, the relevant time period spanned from one year before Mr. Lawson applied (i.e., from 2005) to when the case was decided. [FN114] Earlier conduct, however, could be considered to the extent it bore on Mr. Lawson's character during the relevant time period. [FN115]

In that respect, Mr. Lawson had a substantial strike against him: in 1985, twenty years earlier, he had killed his wife. [FN116]

There were mitigating circumstances. He had enlisted in the Marines at age eighteen. [FN117] He served thirteen months of combat duty in Vietnam, as a consequence of which he developed drug and alcohol addictions and post- traumatic stress disorder. [FN118] When he returned from the war, he had little support; indeed, post-traumatic stress disorder had not yet even been recognized as a diagnosis. [FN119] It was against *1583 this background that he lost con- trol in a quarrel with his wife and killed her. [FN120]

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 9 160 U. Pa. L. Rev. 1561

Mr. Lawson was convicted of manslaughter, although, significantly, he was acquitted of murder. [FN121] He served more than thirteen years in prison. [FN122] There, he overcame his drug and alcohol problems, earned three degrees (two with honors), completed several training programs, and counseled and taught other inmates, drawing on his own painful experiences. [FN123] Upon his release, he obtained gainful employment and worked for eight years as a drug ab- use counselor, helping countless individuals who were trying to deal with their own addictions. [FN124] He moved back home with his mother and took care of her as her health failed. [FN125] He went to church every Sunday and regularly volunteered for church activities, cooking curry goat and curry chicken for fundraisers and events. [FN126] He brought leftover food to homeless veterans in the park. [FN127] He played chess in a neighborhood chess club. [FN128] His mother had tended a garden in front of the apartment complex, and when she died, he took over the garden. [FN129]

Yes, Mr. Lawson took a life, but twenty-five years had gone by. In the meantime, he was prosecuted and convicted- -and sentenced. He paid his debt to society, while making the most of his time in prison. After his release, he returned to his community and led a positive and constructive life. [FN130]

I ruled in his favor. I held that Vernon Lawson was a person of “good moral character.” [FN131] I did not sentence Mr. Lawson, nor was his criminal prosecution before me. Nonetheless, I have included his story here because it implic- ates many of the concepts relevant to sentencing, including the goals of punishment, the role of empathy, and the opera- tion of the criminal justice system.

*1584 Mr. Lawson committed an unspeakable crime. The criminal justice system could not forgive him for what he did, but it could--and did--treat him with some understanding and appreciation for the difficulties he had encountered. The jury surely had some empathy for him in 1986 when, despite the circumstances of the brutal stabbing, it acquitted him of murder and convicted him only of manslaughter, finding that he had acted under “extreme emotional disturb- ance.” [FN132] Likewise, the sentencing court must have had some understanding of the struggles that led Mr. Lawson to commit his crime, as it imposed a sentence of only ten to twenty years for the intentional taking of a life. [FN133]

Mr. Lawson was punished. At the same time, he was given some hope and the opportunity for redemption. He made the most of that opportunity, becoming, in the end, a productive member of society.

The government elected not to appeal, and a few weeks after my decision, Mr. Lawson was naturalized, in a cere- mony at our Courthouse. Afterwards, he stopped by my Chambers. I was able to shake his hand, and he was able to shake mine--as an American citizen.

[FNd1]. United States Circuit Judge, United States Court of Appeals for the Second Circuit. This Essay is drawn from re- marks given on October 29, 2011, at the Symposium on “Sentencing Law: Rhetoric and Reality,” at the University of Pennsylvania Law School. Prior to being appointed to the Second Circuit, I served as a United States District Judge in the Southern District of New York from September 1994 until April 2010. My observations herein are drawn largely from that experience. I acknowledge, with gratitude, the assistance of my law clerks, Emily Aldridge, Jane Kim, Kevin Sullivan, and Samuel Wasserman, and my judicial assistant, Sharon Volckhausen.

[FN1]. The President issued a written statement, which included the following: Now, the process of selecting someone to replace Justice Souter is among my most serious responsibilities as Presid- ent, so I will seek somebody with a sharp and independent mind and a record of excellence and integrity. I will seek someone who understands that justice isn't about some abstract legal theory or footnote in a case book; it is also about how our laws affect the daily realities of people's lives, whether they can make a living and care for their families,

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 10 160 U. Pa. L. Rev. 1561

whether they feel safe in their homes and welcome in their own nation. I view that quality of empathy, of understanding and identifying with people's hopes and struggles, as an essential ingredient for arriving a[t] just decisions and outcomes. I will seek somebody who is dedicated to the rule of law, who honors our constitutional traditions, who respects the integrity of the judicial process and the appropriate limits of the ju- dicial role. I will seek somebody who shares my respect for constitutional values on which this Nation was founded and who brings a thoughtful understanding of how to apply them in our time. Remarks on the Retirement of Supreme Court Justice David H. Souter, 1 Pub. Papers 604, 604 (May 1, 2009).

[FN2]. See, e.g., Kim McLane Wardlaw, Essay, Umpires, Empathy, and Activism: Lessons from Judge Cardozo, 85 Notre Dame L. Rev. 1629, 1631 (2010) ( “No sooner had President Barack Obama uttered the word ‘empathy’ in connec- tion with judicial appointments than the word took on a life of its own. It became a code word for judicial overreaching, and it served as the blank slate onto which politicians painted doomsday scenarios of a judiciary run amok.” (footnote omitted)); Josh Gerstein, Obama's Search for ‘Empathy’ Shapes Supreme Court Replacement Debate, Politico (May 4, 2009, 4:15 AM), http:// www.politico.com/news/stories/0509/22058.html (quoting Senator Orrin Hatch as saying, “What does [empathy] mean? Usually that's a code word for an activist judge.”); see also Major Garrett, Obama Pushes for ‘Empathetic’ Supreme Court Justices, FoxNews.com (May 1, 2009), http:// www.foxnews.com/politics/2009/05/01/obama-pushes-empathetic-supreme-court-justices (“How does President Obama spell ‘empathy’? S-C-O-T-U-S.”).

[FN3]. Matt Corley, Steele on Judges with ‘Empathy’: ‘I'll Give You Empathy. Empathize Right on Your Behind!,’ Think Progress (May 8, 2009, 2:37 PM), http://thinkprogress.org/politics/2009/05/08/39363/steele-empathize-behin.

[FN4]. In November 2009, following my confirmation hearing, written questions about empathy were collected and posed to me by Senator Jeff Sessions. For example, I was asked, “What role do you believe that empathy should play in a judge's consideration of a case?” and, “Do you think that it's ever proper for judges to indulge their own subjective sense of empathy in determining what the law means?” Confirmation Hearings on Federal Appointments: Hearings Before the S. Comm. on the Judiciary (pt. 4), 111th Cong. 835-36 (2011) (written questions of Sen. Jeff Sessions, Member, S. Comm. on the Judiciary). Indeed, a website was created to track the responses of judicial candidates to the Senate ques- tionnaire about empathy. See Senate Debate on Empathy, Center for Building Culture of Empathy, http:// cultureofem- pathy.com/references/senate-debate (last visited Mar. 15, 2012); see also Wardlaw, supra note 2, at 1648 (“‘[E]mpathy’ quickly became a three-syllable call to arms, inciting opposition to the President's judicial nominees.”).

[FN5]. Terry A. Maroney, The Persistent Cultural Script of Judicial Dispassion, 99 Calif. L. Rev. 629, 630-31 (2011) (quoting Thomas Hobbes, Leviathan 203 (A.R. Walker ed., Cambridge Univ. Press 1904) (1651)). Professor Terry Ma- roney has noted that “[i]nsistence on emotionless judging--that is, on judicial dispassion--is a cultural script of unusual longevity and potency.” Id. at 630. He disagrees, however, with the view that judges should be emotionless and con- cludes that the “cultural script of judicial dispassion” should be put “aside.” Id. at 681.

[FN6]. See Diane S. Sykes, Gender and Judging, 94 Marq. L. Rev. 1381, 1387-88 (2011) (expressing disagreement with President Obama “to the extent that the President's standard for deciding hard cases is meant to suggest that a judge's em- pathy should determine the substantive content of the law”).

[FN7]. See, e.g., United States v. Mullings, 330 F.3d 123, 124-25 (2d Cir. 2003) (“When interpreting the Guidelines, we begin with the basic rules of statutory construction, and we give all terms in the Guidelines their ordinary meanings un- less there are persuasive reasons not to do so.”). See generally United States v. Verkhoglyad, 516 F.3d 122, 127-37 (2d Cir. 2008) (considering relevant statutory language, applicable guidelines, and Sentencing Commission policy statements

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 11 160 U. Pa. L. Rev. 1561

in reviewing the sentence for reasonableness as to both length of sentence and process by which it was reached).

[FN8]. Leonard B. Sand et al., Modern Federal Jury Instructions § 2.01 (Instruction 2-12) (2011).

[FN9]. Id.

[FN10]. See Guido Calabresi, Dedication, What Makes a Judge Great: To A. Leon Higgin-botham, Jr., 142 U. Pa. L. Rev. 513, 513 (1993) (“To be truly great a judge needs wisdom, that sense of balance which allows one to weigh what cannot be measured, generosity of spirit, that compassion which causes one to know what it is like to be in trouble and in pain, and to desire instinctively to reach out and help, and above all courage, that fire which compels one to do what is right though the heavens--and one's own career--may fall.”); Erwin Chemerinsky, Korematsu v. United States: A Tragedy Hopefully Never to Be Repeated, 39 Pepp. L. Rev. 163, 164-65 (2011) (“I would think the opposite of some- body with empathy is a sociopath. Surely we don't want sociopaths on the United States Supreme Court, but we do want the Justices to consider the social impact of their ruling.”); Carlton F.W. Larson, Tribute, Judge Michael Daly Hawkins, The Jury System, and American Democracy, 43 Ariz. St. L.J. 49, 55 (2011) (“When President Obama famously de- scribed empathy as a significant judicial virtue, he was widely mocked by people who mistook empathy for sympathy. Empathy, properly understood, is the ability to put oneself in another's shoes and understand how the world appears to him or her. Surely that is what good judges do every day.” (footnote omitted)); Sykes, supra note 6, at 1388 (“Empathy is a virtue, and it is also a desirable quality in a judge, who of course must interpret and apply the law in the context of real- life cases. We cannot properly decide our cases without acquiring some insight into the contextual realities of each party's situation, and a judge's knowledge of the human condition and capacity to identify with others is important to that endeavor.”).

[FN11]. See United States v. Ballard, 322 U.S. 78, 94 (1944) (Jackson, J., dissenting) (comparing “dispassionate judges” to “Santa Claus,” “Uncle Sam,” and “Easter bunnies”).

[FN12]. See Wardlaw, supra note 2, at 1646-47 (“Empathy allows the judge to appreciate more fully the problem before her; it does not solve it for her; it does not dictate a result.”).

[FN13]. Remarks on the Retirement of Supreme Court Justice David H. Souter, supra note 1, at 604.

[FN14]. See In re Martin-Trigona, 9 F.3d 226, 227 (2d Cir. 1993) (“The unfortunate tendency of some individuals to ab- use the litigation process has prompted courts to adopt a variety of techniques to protect both themselves and the public from the harassing tactics of vexatious litigants.”).

[FN15]. Even after I was confirmed as a circuit judge in April 2010, I continued to sentence defendants, as I kept my en- tire criminal docket.

[FN16]. In fact, I sentenced many defendants more than once, as some would violate their terms of supervised release and would be brought back before me to be sentenced for the violations. For example, I sentenced a defendant in a secur- ities case three times--first for the original crime and twice more for violations of supervised release. See Benjamin Weiser, A Judge's Education, One Sentence at a Time, N.Y. Times, Oct. 9, 2011, at MB1.

[FN17]. The New York Times published two comprehensive articles on sentencing that focused on a number of my cases. The first considered my sentencing of financier Bernard Madoff to 150 years in prison. See Benjamin Weiser, Madoff Judge Recalls Rationale for Imposing 150-Year Sentence, N.Y. Times, June 29, 2011, at A1. The second ex- amined several more of my sentencings and addressed the difficulties and challenges judges encounter when passing

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 12 160 U. Pa. L. Rev. 1561

judgment on individuals convicted of crimes. See Weiser, supra note 16. These articles were also instructive because the reporter found defendants who were willing to talk publicly about their experiences in the criminal justice system. I also agreed to be interviewed and to talk about cases that were no longer pending. Although it is not part of our culture as judges to talk to the press on the record about specific cases, I agreed to do so because I felt it would benefit the public to learn more about the sentencing process.

[FN18]. See United States v. Tutty, 612 F.3d 128, 130-31 (2d Cir. 2010) (“We ‘must first ensure that the district court committed no significant procedural error....”’ (quoting Gall v. United States, 552 U.S. 38, 51 (2007))); United States v. Cavera, 550 F.3d 180, 189-90 (2d Cir. 2008) (en banc) (noting that judicial deference is warranted only if the appellate court is “satisfied that the district court complied with the Sentencing Reform Act's procedural requirements”).

[FN19]. Cavera, 550 F.3d at 189 (quoting United States v. Rigas, 490 F.3d 208, 238 (2d Cir. 2007)); accord United States v. Dorvee, 616 F.3d 174, 179 (2d Cir. 2010) (citing Cavera in observing that reasonableness is determined by ex- amining whether a sentence is “within the range of permissible decisions”).

[FN20]. See Cavera, 550 F.3d at 190 (discussing how failure to consider sentencing factors set forth in 18 U.S.C. §3553(a) (2006) is procedural error).

[FN21]. 18 U.S.C. §3553 (“Imposition of a sentence”).

[FN22]. U.S. Sentencing Guidelines Manual chs. 1, 5-6 (2011) ( “Introduction,” “Determining the Sentence,” and “Sentencing Procedures”).

[FN23]. Fed. R. Crim. P. 32 (“Sentencing and Judgment”).

[FN24]. Id. 32(b)(1).

[FN25]. Id. 32(c)(1)(A). In some limited situations, the presentence investigation and a presentence report are waived. Id.

[FN26]. Id. 32(c)(2).

[FN27]. Id. 32(d).

[FN28]. Id. 32(e)(2).

[FN29]. Id. 32(f).

[FN30]. Id. 32(g).

[FN31]. The Honorable Gerald E. Rosen, an experienced District Judge for the Eastern District of Michigan, has written about the emotional challenges of being a jurist as “[v]irtually every week[ he] receive[s] letters from the families of de- fendants who are facing sentence,...relating heart-rending stories of serious illness in the family, or financial hardship and deep emotional loss for the children, parents, spouses and other family members of the defendant.” Gerald E. Rosen, The Hard Part of Judging, 34 Suffolk U. L. Rev. 1, 6 (2000).

[FN32]. For example, when I sentenced Oscar Wyatt, Jr., who pleaded guilty to crimes involving the United Nations oil- for-food program, to a below-Guidelines sentence, I was influenced by the many letters submitted to me in support of Wyatt. Alan Feuer, One-Year Term for Oilman Convicted in Iraq Kickbacks, N.Y. Times, Nov. 28, 2007, at A12.

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 13 160 U. Pa. L. Rev. 1561

[FN33]. I received many such letters in the Madoff case. See United States v. Madoff, 626 F. Supp. 2d 420, 425-27 (S.D.N.Y. 2009) (addressing request by media to unseal emails submitted by victims); Leslie Wayne, Madoff's Victims Speak in Court Letters, N.Y. Times, Mar. 21, 2009, at B2 (“They write of sleepless nights, broken dreams, and retire- ments postponed.”). Pursuant to the Crime Victims' Rights Act, 18 U.S.C. §3771(a)(4) (2006), victims have the right to be heard at various stages of a criminal case, including at sentencing. See Jayne W. Barnard, Tribute, Listening to Vic- tims, 79 Fordham L. Rev. 1479, 1488-89 (2011) (discussing the acknowledgment of victims and their stories of loss in the Madoff case).

[FN34]. See, e.g., United States v. Abiodun, 442 F. Supp. 2d 88, 98-101 (S.D.N.Y. 2006) (calculating the amount of loss in a stolen credit report case), aff'd in part, vacated in part, 536 F.3d 162 (2d Cir. 2008).

[FN35]. See, e.g., United States v. Sangemino, 136 F. Supp. 2d 293, 298-301 (S.D.N.Y. 2001) (holding that an elderly widow in a securities fraud case was a “vulnerable victim” pursuant to section 3A1.1(b) of the U.S. Sentencing Guidelines Manual and imposing a two-level increase in the offense level).

[FN36]. See United States v. Fatico, 603 F.2d 1053, 1057 & n.9 (2d Cir. 1979) (addressing issues relating to evidentiary sentencing hearings and concluding that the government should not be held to a beyond a reasonable doubt standard of proof with respect to sentencing issues).

[FN37]. When a defendant raises a factual dispute and forces a Fatico hearing, there is a risk that the court will impose a more severe sentence because the court will see the evidence. Indeed, one of the benefits of pleading guilty is that the court most likely will not see the evidence. In the Sangemino case, because the defendant contested whether the victim was a “vulnerable victim,” I held an evidentiary hearing and listened to the recordings of the defendant trying to take ad- vantage of a lonely, elderly widow. I was deeply troubled by what I heard. See Sangemino, 136 F. Supp. 2d at 296-98 (describing a series of phone calls in which the defendant unflinchingly lured and siphoned money from a widow despite her fall into financial troubles).

[FN38]. See Rosen, supra note 31, at 6 (“[A]t the sentencing hearing itself, the defendant's obviously distraught family is often in the courtroom, reminding the judge just by their presence of the defendant's human side.”).

[FN39]. Fed. R. Crim. P. 32(i)(4).

[FN40]. Rule 32(i)(4)(B) of the Federal Rules of Criminal Procedure and 18 U.S.C. §3771(a)(4) provide victims with the right to be reasonably heard before a sentence is imposed.

[FN41]. Barnard, supra note 33, at 1483; see also id. (“There were... thousands of documented direct investors in Madoff's Ponzi scheme. Of that number only 113 consented to have their statements submitted to Judge Chin... [a]nd only a handful of them--nine in total--actually stood up to provide victim allocution.” (footnotes omitted)).

[FN42]. Such was the case at the Madoff sentencing. See id. at 1484-87 (summarizing victims' statements, discussing “the problem with emotional allocution,” and observing that “the kind of naked emotion often seen in victim allocution- -the finger pointing, the name-calling, the raining down of curses-- can lead, as one federal judge suggested, to ‘some kind of lynching”’).

[FN43]. The victim's daughter said the following: [A] couple of years ago, I was about 14 years old, and I woke up in the morning, I woke up in the morning and I had to go to day camp, and my father used to take me to day camp every morning, and I looked all over the house and he

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 14 160 U. Pa. L. Rev. 1561

was not there. I didn't know what happened. I was only 14 years old. After that day I never saw my dad again since then. He has missed two graduations, my first graduation from eighth grade. I was valedictorian and he was not there to see it. Then I went to high school. He didn't see that graduation, either. Now I'm going to college, I got a scholarship, I'm going to college for free, and my father missed that, too. Soon I'm going to get married and have children and my father is not going to see that, either. I just want to say that even after all that pain and anger that I have inside, even after being daddy's little girl, I for- give you. And some day in the future, when I'm very successful and I have a good job, I'm going to send you food, I'm going to send you clothes and I'm going to pray for you, because that's the way I was raised. I could stand here and tell you I hate you, but I don't. I can stand here and tell you I hope you rot in hell, but I don't. I hope you live forever. Transcript of Sentencing Hearing at 6-7, United States v. Padilla, No. 97-0809 (S.D.N.Y. Aug. 17, 2000).

[FN44]. 543 U.S. 220, 244-46 (2005).

[FN45]. The defendant's intent, of course, and the extent of his culpability are factors to consider in determining the length of his sentence. See, e.g., Enmund v. Florida, 458 U.S. 782, 798 (1982) (“Unless the death penalty when applied to those in [the defendant]'s position measurably contributes to one or both of these goals [retribution and deterrence], it ‘is nothing more than the purposeless and needless imposition of pain and suffering,’ and hence an unconstitutional pun- ishment.” (quoting Coker v. Georgia, 433 U.S. 584, 592 (1977))).

[FN46]. See Fed. R. Crim. P. 32(i)(4)(A)(iii) (requiring the court to provide the government an opportunity to speak at the sentencing hearing).

[FN47]. Cf., e.g., Payne v. Tennessee, 501 U.S. 808, 825 (1991) (“[T]he State has a legitimate interest in counteracting the mitigating evidence which the defendant is entitled to put in, by reminding the sentencer that just as the murderer should be considered as an individual, so too the victim is an individual whose death represents a unique loss to society and in particular to his family.” (alteration in original) (quoting Booth v. Maryland, 482 U.S. 496, 517 (1987) (White, J., dissenting)) (internal quotation marks omitted)).

[FN48]. Gall v. United States, 552 U.S. 38, 49 (2007).

[FN49]. Fed. R. Crim. P. 32(i)(3)(B). In most cases, these rulings are made from the bench. Occasionally, however, judges will issue a written opinion addressing significant sentencing disputes, as I did in a case in which I granted a mo- tion for a downward departure for extraordinary family circumstances. See United States v. White, 301 F. Supp. 2d 289, 293, 296 (S.D.N.Y. 2004) (granting downward departure where a mother of five children, who also took care of her four- teen-year-old sister, pled guilty to bank robbery).

[FN50]. See Weiser, supra note 16.

[FN51]. 18 U.S.C. §3553(c) (2006); see also United States v. Buissereth, 638 F.3d 114, 117 (2d Cir. 2011) (holding that the court has a statutory obligation to state in open court the reasons for a given sentence).

[FN52]. Gall, 552 U.S. at 50; accord United States v. Dorvee, 616 F.3d 174, 180 (2d Cir. 2010) (quoting Gall's lan- guage).

[FN53]. The New York Times interviewed several defendants I sentenced, including Daniel Sangemino. He was twenty- five years old at the time and had pled guilty to securities fraud. I sentenced him first to thirty-seven months, then to

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 15 160 U. Pa. L. Rev. 1561

eight more months after he violated his supervised release, and then to an additional sixteen months after yet another vi- olation of supervised release. At the third sentencing, I said to him: “I don't know what you are doing with your- self....This is really your last chance.” Weiser, supra note 16. He completed his sentence and did manage to stay drug free. He told the Times that I had treated him fairly. He said about me: “I'll never forget his expression. He wasn't angry. He was, like, ‘C'mon.”’ Id. For a description of Sangemino's original criminal conduct, see United States v. Sangemino, 136 F. Supp. 2d 293, 294-98 (S.D.N.Y. 2001).

[FN54]. As the Supreme Court has held in the death penalty context, A process that accords no significance to relevant facets of the character and record of the individual offender or the circumstances of the particular offense excludes from consideration in fixing the ultimate punishment of death the possibility of compassionate or mitigating factors stemming from the diverse frailties of humankind. It treats all persons convicted of a designated offense not as uniquely individual human beings, but as members of a faceless, undifferenti- ated mass to be subjected to the blind infliction of the penalty of death. Woodson v. North Carolina, 428 U.S. 280, 303-04 (1976) (plurality opinion).

[FN55]. Gall, 552 U.S. at 49.

[FN56]. United States v. Cavera, 550 F.3d 180, 189 (2d Cir. 2008) (en banc) (quoting United States v. Crosby, 397 F.3d 103, 113 (2d Cir. 2005)).

[FN57]. Prior to the Sentencing Reform Act of 1984 and the adoption of the Guidelines in 1987, district judges had “broad discretion” in sentencing. Koon v. United States, 518 U.S. 81, 92 (1996); see also Williams v. United States, 503 U.S. 193, 219 n.17 (1992) (White, J., dissenting) (noting the “near-absolute discretion vested in the district courts prior to sentencing reform”). The Guidelines, of course, limited that discretion by requiring district judges to sentence within the applicable Guidelines range “if the case is an ordinary one.” Koon, 518 U.S. at 92. Much of the sentencing court's discretion has been restored with the line of cases culminating in the Supreme Court's holding that the Guidelines were advisory only. See United States v. Booker, 543 U.S. 220, 226-27, 245 (2005) (holding that (1) the Sixth Amendment as construed by Blakely applies to the federal Sentencing Guidelines and juries must find facts relevant to sentencing; (2) provisions of the federal sentencing statute making the Guidelines mandatory are unconstitutional and therefore must be “severed and excised”; and (3) as modified, the federal sentencing statute made the Guidelines “effectively advisory”); Blakely v. Washington, 542 U.S. 296, 298-300, 313-14 (2004) (holding that the defendant was entitled to a jury trial with respect to the disputed factual issue of whether he acted with “deliberate cruelty,” which would permit the trial court to sentence him to an “exceptional sentence” above the “standard range”); Apprendi v. New Jersey, 530 U.S. 466, 490 (2000) (“[A]ny fact that increases the penalty for a crime beyond the prescribed statutory maximum must be submitted to a jury.”).

[FN58]. See Kimbrough v. United States, 552 U.S. 85, 101 (2007) (concluding that the sentencing judge may consider policy disagreement with crack/cocaine disparity in the Guidelines when imposing sentence); see also Spears v. United States, 555 U.S. 261, 264 (2009) (per curiam) (observing that Kimbrough recognized that district courts have authority to vary from crack cocaine Guidelines based on policy disagreement with them).

[FN59]. See Gall, 552 U.S. at 46 (“For even though the Guidelines are advisory rather than mandatory, they are...the product of careful study based on extensive empirical evidence derived from the review of thousands of individual sen- tencing decisions.”).

[FN60]. See Cavera, 550 F.3d at 189 (“It is now...emphatically clear that the Guidelines are guidelines--that is, they are truly advisory.”). In Rita v. United States, the Supreme Court held that a sentence within the applicable Guidelines range

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 16 160 U. Pa. L. Rev. 1561

is not presumptively reasonable. 551 U.S. 338, 347 (2007).

[FN61]. United States v. Jones, 531 F.3d 163, 174 (2d Cir. 2008).

[FN62]. Gall, 552 U.S. at 46.

[FN63]. 18 U.S.C. §3553(a)(1)-(2) (2006). Other statutory factors include the kinds of sentences available, policy state- ments issued by the Sentencing Commission, the need to avoid disparity in sentencing, and the need to provide restitu- tion to victims. Id. §3553(a)(3)-(7).

[FN64]. Id. §3553(a).

[FN65]. See, e.g., Graham v. Florida, 130 S. Ct. 2011, 2028 (2010) ( “[T]he goals of penal sanctions that have been re- cognized as legitimate [are] retribution, deterrence, incapacitation, and rehabilitation.” (citing Ewing v. California, 538 U.S. 11, 25 (2003) (plurality opinion))); Kennedy v. Louisiana, 554 U.S. 407, 420 (2008) (“[P]unishment is justified un- der one or more of three principal rationales: rehabilitation, deterrence, and retribution.”); Hudson v. United States, 522 U.S. 93, 101 (1997) (recognizing that a punishment should serve “the traditional goals of punishment, namely, retribution and deterrence” (internal quotation marks omitted)).

[FN66]. See Weiser, supra note 17.

[FN67]. Transcript of Sentencing Hearing at 47, United States v. Madoff, No. 09-0213 (S.D.N.Y. June 29, 2009).

[FN68]. Id.

[FN69]. The cases discussing the traditional goals of punishment do not include helping victims heal, see supra note 65, and this was not a consideration that I recall relying on explicitly in prior sentences. Given the magnitude of the harm in the Madoff case, however, and what I perceived then to be the slim chance of meaningful recovery for the victims, I felt this was an important consideration.

[FN70]. Transcript of Sentencing Hearing, supra note 67, at 47-49.

[FN71]. In Graham v. Florida, when addressing the constitutionality of life without parole for a juvenile, the Supreme Court wrote: The State does not execute the offender sentenced to life without parole, but the sentence alters the offender's life by a forfeiture that is irrevocable. It deprives the convict of the most basic liberties without giving hope of restora- tion, except perhaps by executive clemency--the remote possibility of which does not mitigate the harshness of the sen- tence. As one court observed in overturning a life without parole sentence for a juvenile defendant, this sentence “means denial of hope; it means that good behavior and character improvement are immaterial; it means that whatever the future might hold in store for the mind and spirit of [the convict], he will remain in prison for the rest of his days.” 130 S. Ct. 2011, 2027 (2010) (alteration in original) (citation omitted) (quoting Naovarath v. State, 779 P.2d 944, 944 (Nev. 1989)).

[FN72]. I sentenced someone recently for a murder he had committed in 1992, when he was eighteen years old. If I had sentenced him then, I would have had to try to predict the future, taking into account the factors relevant to the senten- cing of minors. Younger people, for example, are more capable of redemption. See Graham, 130 S. Ct. at 2026-30. The defendant was not apprehended, however, for nearly twenty years, and I could see that by then he had been able to change his ways and lead a productive life. I sentenced him to twenty-four years of imprisonment.

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 17 160 U. Pa. L. Rev. 1561

[FN73]. Weiser, supra note 16.

[FN74]. United States v. Leung, 360 F.3d 62, 64-65 (2d Cir. 2004); see also Weiser, supra note 16 (discussing the senten- cing).

[FN75]. Leung, 360 F.3d at 66.

[FN76]. Id. at 70 (alteration in original) (quoting sentencing transcript).

[FN77]. Id. at 71 (quoting sentencing transcript).

[FN78]. Id. at 72. The Second Circuit, though, concluded that I had erred in my grouping analysis and that I had failed to make a required finding with respect to acceptance of responsibility. Id. at 69-70.

[FN79]. United States v. White, 301 F. Supp. 2d 289, 291 (S.D.N.Y. 2004); see also Weiser, supra note 16 (interviewing defendant in United States v. White).

[FN80]. White, 301 F. Supp. 2d at 290-91.

[FN81]. Id. at 293.

[FN82]. Id. at 296.

[FN83]. Id. at 291-92.

[FN84]. Id. at 295.

[FN85]. Id. at 295-96.

[FN86]. My grandfather was a waiter in Chinese restaurants, and I was one of five children raised by a Chinese cook and a seamstress who worked in Chinatown garment factories. Denny Chin, Representation for Immigrants: A Judge's Per- sonal Perspective, 78 Fordham L. Rev. 633, 635-38 (2009).

[FN87]. White, 301 F. Supp. 2d at 296.

[FN88]. Id.

[FN89]. Id. at 297.

[FN90]. Weiser, supra note 16.

[FN91]. See Graham v. Florida, 130 S. Ct. 2011, 2027 (2010) (observing that a sentence that denies hope renders good behavior and character development meaningless (citing Naovarath v. State, 779 P.2d 944, 944 (Nev. 1989))). Of course, in many situations the absence of hope is debilitating not only for the defendant, but also for her family.

[FN92]. Weiser, supra note 16.

[FN93]. Id.

[FN94]. See Mark W. Bennett, Hard Time: Reflections on Visiting Federal Inmates, 94 Judicature 304, 304 (2011) (“It is

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 18 160 U. Pa. L. Rev. 1561

an awesome responsibility to take one's liberty away.”); Jack B. Weinstein, Does Religion Have a Role in Criminal Sen- tencing?, 23 Touro L. Rev. 539, 539 (2007) (“Sentencing, that is to say punishment, is perhaps the most difficult task of a trial court judge.”).

[FN95]. See supra note 3 and accompanying text.

[FN96]. See, e.g., United States v. Farhane, 634 F.3d 127, 167 (2d Cir. 2011) (“[S]um-mations...are not ‘detached expos- ition[s].”’ (second alteration in original) (quoting United States v. Wexler, 79 F.2d 526, 530 (2d Cir. 1935))). As Profess- or Terry Maroney has noted, [L]aw is infused with emotion and ideas about emotion. Examples range from the excited utterance exception to the hearsay rule (reflecting the idea that statements made while in an intense emotional state are likely to be truthful), to heightened protection of homes (because of presumed emotional attachment to them), to awards of damages for emotion- al suffering (which assumes pain can be monetized), to victim impact statements (thought to promote emotional “closure”). Maroney, supra note 5, at 642.

[FN97]. Wexler, 79 F.2d at 529-30.

[FN98]. In contrast, there is a “right” process. Specifically, a defendant is entitled to a fair sentencing process, in which he is vigorously represented, he and his lawyer have a full opportunity to be heard, the sentencing court gives due consid- eration to the statutory and other relevant factors, and the court makes necessary rulings and findings and explains its de- cision.

[FN99]. See Irving R. Kaufman, The Anatomy of Decisionmaking, 53 Fordham L. Rev. 1, 16 (1984) (“[O]ur intuition, emotion and conscience are appropriate factors in the jurisprudential calculus.”).

[FN100]. “[M]uch of the scholarship [on the role of emotion in the law] posits that it is not only impossible but also un- desirable to factor emotion out of the reasoning process: by this account, emotion leads to truer perception and, ulti- mately, to better (more accurate, more moral, more just) decisions.” Susan Bandes, Empathy, Narrative, and Victim Im- pact Statements, 63 U. Chi. L. Rev. 361, 368 (1996); see also Stephen P. Garvey, “As the Gentle Rain Falls from Heav- en”: Mercy in Capital Sentencing, 81 Cornell L. Rev. 989, 1043 (1996) (“Emotions can...overwhelm or ‘unhinge’ our faculty of reason. Yet our emotions are not irrational....When we experience an emotion we generally do so for a reas- on....When, among other things, they are ‘intelligently proportionate in intensity to their objects,’ they act as ‘trustworthy guides to moral insight.”’ (footnotes omitted) (quoting Michael S. Moore, The Moral Worth of Retribution, in Respons- ibility, Character, and the Emotions 179, 190 (Ferdinand Schoeman ed., 1987))).

[FN101]. As Judge Kim McLane Wardlaw has written, Nobody would suggest that judges throw the law to the wind and decide cases based exclusively on individual sentiments of justice. But it is irresponsible to pretend that one's notions of justice do not play, or may not play, a role in the cases for which precedent fails to command one outcome or another. They constitute one--and only one--of the judge's points of reference. Empathy allows the judge to appreciate more fully the problem before her; it does not solve it for her; it does not dictate a result. Wardlaw, supra note 2, at 1646-47 (footnote omitted).

[FN102]. Weinstein, supra note 94, at 539.

[FN103]. See Wardlaw, supra note 2, at 1644 (“Life experiences provide each of us with sentiments of right and wrong,

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 19 160 U. Pa. L. Rev. 1561

fair and unfair, rational and irrational, just and unjust. And that is true for judges as it is for anyone else.”).

[FN104]. Remarks on the Retirement of Supreme Court Justice David H. Souter, supra note 1, at 604. As Judge Wardlaw has written, “It is those judges who are unable to understand the views and problems of others--who are unable to assess problems from any vantage point other than their own--who may not be up to the task of administering justice equally and impartially.” Wardlaw, supra note 2, at 1649.

[FN105]. See Edward J. Devitt, Ten Commandments for the New Judge, 47 A.B.A. J. 1175 (1961).

[FN106]. See Edward J. Devitt, Ten Commandments for the New Judge, 65 A.B.A. J. 574 (1979), reprinted in 82 F.R.D. 209 (1979).

[FN107]. Id. at 209.

[FN108]. Id. at 209-10.

[FN109]. Lawson v. U.S. Citizenship & Immigration Servs., 795 F. Supp. 2d 283, 285 (S.D.N.Y. 2011).

[FN110]. Id. at 292-93.

[FN111]. Id. at 295-96.

[FN112]. Id.

[FN113]. Thus, the relevant statutory provision was 8 U.S.C. §1440(a) (2006).

[FN114]. Lawson, 795 F. Supp. 2d at 294.

[FN115]. Id. at 293-95.

[FN116]. Id. at 289.

[FN117]. Id. at 286.

[FN118]. Id. at 286-89.

[FN119]. Id. at 288-89 & n.10.

[FN120]. Id. at 289, 298.

[FN121]. Id. at 289, 296; see also 8 C.F.R. §316.10(b)(1)(i) (2011) (providing that a person convicted of murder at any time “shall” be found to lack good moral character).

[FN122]. Lawson, 795 F. Supp. 2d at 289-90.

[FN123]. Id. at 289-90, 298.

[FN124]. Id. at 290.

[FN125]. Id. at 291.

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. 160 UPALR 1561 Page 20 160 U. Pa. L. Rev. 1561

[FN126]. Id.

[FN127]. Id.

[FN128]. Id.

[FN129]. Id.

[FN130]. Id. at 298.

[FN131]. Id. at 297-300.

[FN132]. Id. at 289.

[FN133]. Id. 160 U. Pa. L. Rev. 1561

END OF DOCUMENT

© 2013 Thomson Reuters. No Claim to Orig. US Gov. Works. Identifying Judicial Empathy: Does Having Daughters Cause Judges to Rule for Women’s Issues?∗

Adam Glynn† Maya Sen‡ [email protected] [email protected] July 28, 2012

Abstract In this paper, we leverage the natural experiment of a child’s gender to identify the effect of having daughters on the votes of judges. Using new data on the family lives of U.S. Courts of Appeals judges, we find that, conditional on the number of children a judge has, judges with daughters consistently vote in a more pro-woman fashion on gender issues than judges who have only sons. This result survives a number of robustness tests and appears to be driven primarily by Republican judges. More broadly, this result demonstrates that personal experiences influence how judges make decisions, and it is the first paper to show that empathy may indeed be a component in how judges decide cases.

∗Comments and suggestions welcome. This draft has benefited from helpful feedback from Tom Clark, Ryan Enos, Joshua Fischman, Sean Gailmard, Sandy Gordon, Jennifer Hochschild, Nahomi Ichino, Gary King, Jeffrey Lax, Kevin Quinn, Arthur Spirling, Matthew Stephenson, and participants at the Harvard Applied Statistics Workshop, the Harvard Kennedy School Women and Public Policy Workshop, the 2011 Midwest and European Political Science Association annual conferences, and the 2012 Political Economy and Public Law Conference at the University of Virginia. In addition, we are grateful to Melissa Niedrich, Alex Crabill, and Michelle Pearse for their research assistance. This research was supported by the Institute for Quantitative Social Science. †Department of Government and Institute for Quantitative Social Science, Harvard University, 1737 Cambridge Street, Cambridge, MA 02138 (http://scholar.harvard.edu/aglynn). ‡Department of Political Science, University of Rochester, Harkness Hall 307, Rochester, NY 14627 (http://scholar.harvard.edu/msen).

1 “We need somebody who’s got the heart, the empathy, to recognize what it’s like to be a young teenage mom. The empathy to understand what it’s like to be poor, or African-American, or gay, or disabled, or old. And that’s the criteria by which I’m going to be selecting my judges.”

- Barack Obama, on a 2007 campaign stop

1 Introduction

Upon Justice David Souter’s retirement in May of 2009, President Obama made it clear that one of the criteria he would use in selecting Souter’s replacement would be “empathy” – that is, a potential nominee’s ability to identify “with people’s hopes and struggles as an essential ingredient for arriving at just decisions and outcomes” (Obama, 2009). The statement echoed much of what Obama emphasized throughout his campaign: that the ability to empathize with others is, and ought to be, a key criterion for nomination to the nation’s federal courts. Obama’s press statement was among the first by a U.S. president acknowledging the possible effect of empathy on the way judges decide cases, and it ignited a fierce and ongoing debate. Critics have strongly questioned Obama’s assertion that judges do (or should) have any kind of empathetic feelings based on their relationships with others. As one commen- tator complained, having “empathy” would translate into judges “being partial instead of being impartial,” when, in fact, “a judge is supposed to have empathy for no one but simply to follow the law” (Garrett, 2009). Supporters, on the other hand, have applauded Obama’s sense that empathy and identifying with others was and ought to be an important part of judicial decision making, with some commentators going as far as arguing that the “single value we should demand in a justice [has] nothing to do with race or gender,” but “every- thing to do with empathy for others” (Lithwick and West, 2010). So divisive has been this debate that discussions about “empathy” largely dominated coverage of Sonia Sotomayor’s and Elena Kagan’s Supreme Court nominations, leading one news organization to label the

2 attendant vitriol as the “Empathy Wars” (Just, 2009). Despite the public shouting match, there has been little to no theoretical or empirical evidence within the scholarly literature proving (or disproving) the theory that personal relationships could affect judicial decision making. We do, however, have plenty of anecdotal evidence. For example, Sally Blackmun, the daughter of future Justice Harry Blackmun, became pregnant at 19, dropped out of college, and married her 20-year old boyfriend. The pregnancy eventually ended in miscarriage, but Sally Blackmun would later say about this period that “I was not at all pleased with myself about [it]. It was a big disappointment to my parents [and it] was a decision that I might have made differently[] had Roe v. Wade been around” (Cooper, 2004). Indeed, seven years later, Justice Blackmun, her father, would go on to write that opinion, leading many to cite his daughter’s experience as a catalyst. Similarly, it is widely believed that William Rehnquist’s close relationship with his struggling divorced daughter Janet affected his views on issues like the Family and Medical Leave Act (Bazelon, 2009; Novak, 2003). However, while the Blackmun and Rehnquist examples provide some observational evidence, one can usually provide an alternative explanation for such opinions on the basis of jurisprudence. This paper is the first to provide empirical support for the idea that empathy and personal relationships, as distinct from race and gender, may affect how judges decide cases, and this evidence cannot be explained by jurisprudence alone. We do so by by focusing on the particular personal transformation that Justices Blackmun and Rehnquist experienced: having daughters. Not only could parenting daughters cause a judge to change his or her substantive position (for example, by becoming more progressive on gender issues), but it also provides a clear methodological advantage: once a couple decides to have a child, the sex of that child is outside of that couple’s control, resulting in a natural quasi-experiment (Washington, 2008). Employing a new dataset on federal judges’ families in tandem with a new dataset on nearly one thousand gender-related cases, we show that judges with at least

3 one daughter vote in a more liberal fashion on gender issues than judges with sons, conditional on the number of children. The effect is robust and appears driven largely by Republican appointees. Taken together, these results are the first to demonstrate that empathy and personal relationships – as distinct from partisanship or ascriptive characteristics – may indeed influence how judges reach decisions. This paper proceeds as follows. Section2 discusses the existing literature on personal attributes and judicial decision making, focusing specifically on the literature linking par- enting and political beliefs. Section3 links these issues to the judicial context and develops our core hypotheses. We discuss the data and methods in Section4 and present the key results in Section5. We consider possible mechanisms and alternate explanations in Section 6 and conclude in Section7 by emphasizing the implications of this research on the existing debate about empathy and judicial actors.

2 Whether Empathy and Daughters Affect Beliefs

Relatively little attention scholarly attention has been paid to the possible influence that judges’ relationships or personal connections could exert on their decision making, no doubt due in part to the inherent difficulties involved in researching judges’ personal lives. Most closely on point is the growing body of literature that examines how judges’ ascriptive back- grounds influence decision making.1 The studies most relevant to our inquiry have examined differences in decision making between men and women judges. For example, Boyd, Epstein and Martin(2010) conclude that judge gender does indeed make a difference when it comes to case outcomes – but only in instances involving sex discrimination. Other literature has

1An even larger body of literature considers the strong influence on decision making of judges’ political beliefs – i.e., the “attitudinal” model of judicial decision making and related scholarship (Rohde and Spaeth, 1976; Schubert, 1974; Segal and Spaeth, 2002; Howard and Segal, 2002; Sunstein et al., 2006). Mindful of this literature, we note the possibility that the effect of having girls might vary according to the political attitudes of individual judges.

4 found that being a woman judge has an effect on cases involving sexual harassment, sex discrimination, or the sentences of criminal defendants (Baldez, Epstein and Martin, 2006; Peresie, 2005; Massie, Johnson and Gubala, 2002; Segal and Spaeth, 2002; Crowe, 1999; Davis, Haire and Songer, 1993). Others have found little or no effects associated with a judge’s gender (Manning, Carroll and Carp, 2004; Kulik, Perry and Pepper, 2003; Ashen- felter, Eisenberg and Schwab, 1995). A companion body of literature has found differences between and among judges of different races (Kastellec, 2012; Cox and Miles, 2008; Scherer, 2004; Gottschall, 1983). Another subset of studies have looked at the role that individual experiences, as opposed to individual ascriptive traits (e.g., gender), might have on judicial decision making. Most of these studies have, however, looked at purely professional experiences (Epstein, Knight and Martin, 2003), and most of these have done so within the criminal sentencing context (Sisk, Heise and Morriss, 1998). Very few studies have looked at personal experiences, and, to our knowledge, no empirical inquiry has ever been done on the potential effect that familial relationships (e.g., being a parent) might have on judicial decision making. This fact is surprising given that judges are routinely asked to make difficult decisions regarding guardianships, custody disputes, and juvenile sentencing. Although we can point to no literature within judicial politics, public opinion scholarship provides some evidence on parenting – and in particular the transformative experience of parenting girls versus boys.2 Most of this literature suggests that having daughters leads

2A robust literature within political science has, however, explored how individuals form and hold onto political attitudes. One school is that partisanship is an affective characteristic that is passed on from parent to child (Campbell et al., 1966; Zuckerman, Dasovi´cand Fitzgerald, 2007) through social education and also perhaps through genetic links (Dawes and Fowler, 2009). Dovetailing with this, studies such as Green, Palmquist and Schickler(2004) have suggested that political ideology is more of a social identification (rather than a rationally grounded one) and, as such, relatively stable: once an individual has identified as either “Democrat” or “Republican” very little can change the identification. In this sense, people are about as likely to switch political parties as they are to convert religions or begin identifying with a different racial or ethnic group (Green, Palmquist and Schickler, 2004). Somewhat less attention has been paid to those factors that have the potential to change partisanship – e.g., moving to different city, marrying a person with a different ideology or religion, etc. We show in this paper that having daughters may be one of these

5 individuals to have more liberal political and social positions than those who have sons. For example, Warner(1991) finds that both men and women who have daughters are more likely to support feminist positions than those who do not. Warner and Steel(1999) similarly demonstrate that support for policies promoting gender equity increases for both men and women when they have daughters. More recently, Oswald and Powdthavee(2010) finds evidence that having daughters makes British parents more “left-wing” when it comes to their politics, while having sons makes them more “right-wing.” Shafer and Malhotra(2011) likewise provide evidence that parenting daughters as opposed to sons leads men to have less traditional views of gender roles, but has no effect on women. The exception to this growing body of literature is Conley and Rauscher(2010), who find that having female children leads to more conservative identification. In other fields, having daughters has been shown to increase the probability that parents will divorce (Dahl and Moretti, 2004), that male investors will shift their money into stocks as opposed to bonds (Bogan, 2009), and that fathers will work longer and earn more wages but not as much as they would have if they had had sons (Lundberg and Rose, 2002). This literature thins significantly when it comes to elite decision making. To our knowl- edge, no study has looked at this issue in the context of executive decision making. Within the legislative context, the study most similar to our own is Washington(2008), which found that U.S. Congress members who have daughters are more likely to vote in favor of women’s issues (as defined by the legislative scores assigned to them by the National Organization for Women and by the American Association of University Women) than those who have sons. Washington(2008) also takes a similar approach to the one taken here, leveraging the randomized nature of gender assignment pre-birth. Her results have also been replicated and strengthened by subsequent matching analyses in Iacus, King and Porro(2011). There are, however, strong reasons to think that Congressional legislators differ substantially from factors.

6 federal judges, who are not only unelected and appointed for life, but also older on average when they begin their appointments. Judges also serve under the sometimes binding con- straint of legal precedent – which has led some commentators to assert that judges should not place any import on personal relationships.

3 Identification of Daughters’ Impact

We note at the outset that the basic decision to have a child is undoubtedly deeply connected with – and motivated by – factors such as race, gender, religion, political ideology, age, etc. These are in turn the same factors that we might expect to also affect judicial decision mak- ing. To this extent, we do not have a clear prediction as to the association between daughters and progressive votes because it is possible that a conservative judge would not only have (a) many children and, as a result, (b) many daughters. We should also not be surprised that, being conservative, such a judge may also have (c) viewpoints that run contrary to those espoused by progressive women’s groups. Thus, the total number of children a judge has not only affects the number of girls but also reflects his or her personal beliefs. We take this into account by examining the effect of having daughters only after con- ditioning on the number of children a judge has, following the methodology of Washington (2008). Conditional on having a child, the sex of the child is analogous to a natural experi- ment; that is, once a couple decides to have a child, the sex of that child is effectively outside of that couple’s control and will therefore be unrelated to other individual characteristics, including partisanship or ideology. If judges are affected by their daughters to vote in the liberal direction, then this natural quasi-experiment implies our first and most important hypothesis: conditional on the number of children a judge has, an increase in the number of daughters should be associated with an increase in the likelihood of voting in the liberal

7 direction on gender-related cases.3 A possible complication to this identification strategy is the use by couples of “fertility stopping rules,” or the practice by which couples continue or stop having children in part based on the gender of the children they already have.4 It could be the case, for example, that judges (1) keep having children until gender parity is achieved, (2) are conservative and so are content with having had only boys, or (3) are liberal so are content with having had all girls. To address concern (1) and to partially address concerns (2) and (3), we more closely examine couples who only have one child – for whom the fertility stopping rules are, as we note later on, significantly less of a concern. This leads us to a related supposition: the daughters effect should also manifest among judges with only one child.5 We provide a detailed discussion of fertility stopping rules and other threats to identification in Section6. In addition, the extant literature on parenting and political attitudes, and the more lim- ited literature on parenting by elite political actors, suggests that the number of daughters a judge has will increase the likelihood of a vote in a more progressive direction – but only on issues with a substantial gender dimension. This is a notion in line with the literature on gender and judging (Boyd, Epstein and Martin, 2010), which has established that female judges vote differently than male judges, but only with regard to cases involving sex dis-

3Phrasing this in terms of the proportion of children that are daughters, although tempting, would conflate two processes: the proportion would increase as the number of daughters increases but alternatively as the number of children decreases. Hence, we would not be able to isolate the daughters effect without further assumptions. 4These sorts of fertility stopping rules are referred to in the demography literature as “differential stopping behavior” (Clark, 2000). In the United States, the preference is in gender balance, and many families prefer to continue having children until they have at least one boy and one girl (Clare and Kiser, 1951; Freedman, Freedman and Whelpton, 1960; Sloane and Lee, 1983; Yamaguchi and Ferguson, 1995). Evidence that this preference has diminished over time is explored in Pollard and Morgan(2002). 5This research design more generally fails with regard to other kinds of child-gender manipulations, such as the adoption of children. Parents often have strong preferences about a child’s gender, and, if given a choice, may opt for adopting a girl over a boy, a boy over a girl, or one child of each gender. (Chief Justice John Roberts, for example, has two adopted children, one boy and one girl.) Because parental values might may play a large role in the gender of non-biological children, we drop from this analysis any adopted children or step-children. In addition, we assume that judges are not having sex-selective procedures, including sex-selective abortions.

8 crimination. We therefore expect that a judge who has many daughters will be more likely to support expansive abortion rights, vote in favor of women plaintiffs in sex discrimination cases, and advocate for the treatment of gender as a suspect classification under the U.S. Constitution. We would not, however, expect the influence of having daughters to extend to other issue areas – business litigation, criminal procedural, taxation, bankruptcy, etc. Our second hypothesis is therefore that, all other things being equal, judges with daughters will have more liberal voting records on women’s rights cases, and on women’s rights cases only. Lastly, although we posit that having daughters will on average increase the probability of judges voting in a feminist direction in gender-related cases, we also hypothesize that the effect will vary across different population subsets. For example, because judges appointed by Democrats may already hold progressive opinions on gender-related issues, they might be less affected by the experience of having girls as opposed to sons. Thus, our third hypothesis is that the greatest impact of having daughters will be among judges appointed by Republicans. Relatedly, because women judges will already have been exposed to female socialization, they may also be less affected by having girls as opposed to boys. Our fourth hypothesis is therefore that the daughters effect will be driven by male, not female, judges. Both of these hypotheses are in line with the results found for legislators by Washington(2008).

4 Data and Methods

Data6 come from the U.S. Courts of Appeals, the middle tier in the federal court system. We choose appellate courts over lower-level district courts because appellate courts hear cases that involve purely legal (or mixed law-factual) questions, which provide the sort of context within which personal experiences may exert some influence and which stand in contrast to the factual questions heard by lower courts. In addition, with approximately two hundred

6All data and accompanying statistical code will be posted to a public online data archive at the conclusion of this project.

9 Number of Children 0 1 2 3 4 5 6 7 8 9 N Democrat 12 13 33 24 15 4 - 1 - 1 103 Republican 13 8 44 30 15 7 3 - 1 - 121 Number of Girls 0 1 2 3 4 5 6 7 8 9 Democrats 26 35 29 10 1 2 - - - - Republicans 36 43 31 9 2 0 - - - -

Table 1: Number of children and girls for U.S. Courts of Appeals judges participating in gender-related cases (1996-2002)

judges at any given time, we have more data on appellate courts than if we just focused on the nine justices on the U.S. Supreme Court. Lastly, the Courts of Appeals follow a long-standing practice of randomly assigning cases to judges; thus, judges with daughters may not request to hear gender-related cases.

Family Data. While several databases on appellate judges’ personal characteristics exist, we know of no extant dataset that contains information on judges’ family lives. We therefore created a new dataset that includes the number of children for each appeals judge and how many of them are girls. Table1 provides a summary. This information was gathered from publicly available sources, including biographical blurbs published in compendiums like Who’s Who in American Law, college alumni or church newsletters, newspaper articles, and public announcements such as obituaries or weddings notices. In some instances, we looked to Senate confirmation hearings, in which references to judges’ families are often made. We combined this data with existing data amassed by Zuk, Barrow and Gryski(2009) on judges’ partisanship (e.g., the party affiliation of the appointing president), the year they were born, and the judge’s race, gender, and religious identification. Note that data on the birth order of children – i.e., which child was born first, second, third, etc. – was either completely unavailable or impossible (or risky) to infer, as was information on the approximate age of each child.

10 Case-level Data. Our supposition is that having daughters will affect judicial decision making in those cases having a specific gender frame or angle. Because extant databases have few of these sorts of cases, we collected new data to capture the universe of gender- related cases decided by the U.S. Courts of Appeals. To collect this data, we started with all of the cases, published and unpublished,7 decided from 1996 and 20028 that had “gen- der,” “pregnancy,” or “sex” in the LexisNexis case classification headings. This initial search yielded approximately 1,450 cases, but included 134 cases having nothing to do with women’s rights (e.g., race-based discrimination cases where gender was mentioned only in passing); we therefore subset this population into cases that explicitly involved (1) employment dis- crimination on the basis of gender by private actors, (2) employment discrimination on the basis of pregnancy by private actors, (3) reproductive rights or abortion, and (4) claims made under Title IX. This left us with 1,325 cases that had a gender component, of which the greatest share (92%) were employment discrimination cases. Of these 1,325 cases, 310 were brought by male or transsexual plaintiffs or by plaintiffs representing men’s or LGBT interests. Because these cases did not go to the heart of our inquiry (and because many included male plaintiffs suing for adverse employment decisions due to sexual harassment by the plaintiff), we excluded them. After discarding cases brought by men, we were left 990 gender-related cases involving discrimination against women or women’s rights. For each of these, we collected the names of the appeals court judges on the three-judge panel (excluding non-appeals court visiting judges). We also coded the directionality of each of the judges’ votes as being (1) anti-

7We chose to include unpublished cases as well because having girls could affect the decision to publish an opinion; including both published and unpublished cases avoids such potential bias. 8For the sake of legal and historical consistency, we focus on cases from June 27, 1996 moving forward. This has the effect of including only those cases that were decided after the last major Supreme Court case involving women’s rights issues, United States v. Virginia. We choose 2002 as our end point to make this data analysis parallel to the collection of cases compiled by Songer, Kuersten and Haire(2007). We do this because we use the data from Songer, Kuersten and Haire(2007) in order to address the question of whether daughters have an effect on non-gender related cases.

11 All Democrats Republicans Women Men Mean No. Children 2.47 2.40 2.54 1.58 2.66 Mean No. Girls 1.24 1.33 1.16 0.71 1.34 Proportion who have 0 children 0.11 0.12 0.11 0.29 0.08 1 Child 0.09 0.13 0.07 0.21 0.07 2 Children 0.34 0.32 0.36 0.26 0.36 3 Children 0.24 0.23 0.25 0.13 0.26 4 Children 0.13 0.15 0.12 0.08 0.15 5 Children 0.05 0.04 0.06 0.03 0.05 6 Children or More 0.03 0.02 0.03 - 0.03 Proportion Female 0.17 0.26 0.09 - - Proportion Republican 0.54 - - 0.29 0.59 Proportion White 0.91 0.78 0.99 0.93 0.91 Mean Year Born 1932.55 1931.23 1933.43 1938.57 1931.49 N 224 103 121 38 186

Table 2: Demographics of U.S. Court of Appeal Judges who voted on gender-related cases (1996-2002) feminist or (2) partially or entirely feminist. The feminist decision would be one in which the judge voted in favor of the female plaintiff or in favor of the plaintiff representing women’s interest (for example, if the Equal Employment Opportunity Commission was bringing a suit on behalf of women employees). The anti-feminist decision would be one in which the judge voted in favor of defendant employers, or in favor of sustaining a restrictive anti-abortion statute. The end result was 2,674 unique votes cast by 244 unique appeals court judges. We were able to track down fertility data for 224 of these judges (Tables1 and2). In general, it was more difficult to confirm the absence of children than the presence and sex of children. It is therefore quite likely that many of judges for whom we are missing fertility data did not have any children, and are not relevant for the results discussed in the next section. In addition, we could not find the gender breakdown for the handful of individuals with extremely large numbers of children (e.g., Republican Wallace Clifford, who had 15 children, or Democrat Henry Politz, who had 11); for these individuals, we simply had no counterpoints to provide comparison, meaning that they would be excluded in any event.

12 Democrats Republicans 0 Girls 1 Girl 2 Girls 3 Girls 4 Girls 5 Girls 0 Girls 1 Girl 2 Girls 3 Girls 4 Girls 5 Girls 1 Child 0.46 0.54 - - - - 0.38 0.62 - - - 2 Children 0.15 0.48 0.36 - - - 0.32 0.50 0.18 - - - 3 Children 0.08 0.46 0.33 0.12 - - 0.13 0.37 0.37 0.13 - - 4 Children 0.07 0.07 0.53 0.33 0 - 0.07 0.27 0.60 0.07 0 - 5 Children 0 0 0.25 0.50 0 0.25 0.14 0.00 0.43 0.29 0.14 0 7 Children 0 0 0 0 1 0 0 0.33 0 0.67 0 0 9 Children 0 0 0 0 0 1 0 0 0 0 1 0

Table 3: Proportion of girls (conditional on number of children) for U.S. Courts of Appeals judges participating in gender-related cases (1996-2002)

Lastly, there is no ex ante reason to believe that any missingness or measurement error would be differential for daughters as opposed to sons. Summary statistics on the 224 judges for whom we could find fertility data are described in Table2. These judges have on average 2.47 children per judge, with Republicans having more children on average than Democrats (2.54 versus 2.40 – both higher than the population average of 2.1), but the difference in proportion of female children between Democrats and Republicans is small enough to be attributable to chance. Female judges tend to have fewer children than male judges (1.58 versus 2.66), with nearly 30% of female judges having zero children. Given the extraordinarily small fraction of judges who more than five children, we at times drop these individuals from the analysis (which we motivate in our results discussion below). We also repeat the analyses for judges who only have one child, for whom fertility stopping rules appear to be less of a motivation. Other concerns pertaining to fertility stopping rules are addressed in Section6.

Methodology. Our primary explanatory variable is the number of natural-born daughters that each judge has, conditioned on the total number of children, per the methodology of Washington(2008). We treat the number of daughters as categorical variables in order to explore whether the effect of having girls proceeds in a non-linear fashion. As we discuss below, the effect of having daughters may be the strongest for having one girl and may dimin- ish for subsequent daughters non-linearly. We therefore at times include a dummy variable

13 Less Feminist More Feminist

Republicans

Democrats

All

0.0 0.5 1.0

Proportion of Cases Decided in a Feminist Direction

Figure 1: Distribution of the proportion of cases decided in a feminist direction out of all gender cases decided 1996-2002. representing whether the judge has at least one girl or not. Furthermore, we condition on the total number of children by including fixed effects, which estimates the effects of having girls among judges with the same number of children. (In all analyses, 0 is the omitted category for girls and 1 is the omitted category for children.) A potential problem with looking at appellate judges is that these judges nearly always hear cases in groups of three. Thus, a judge who has many daughters might lobby his panel colleagues on issues pertaining to women’s rights more strongly than if he had sons, thereby creating a certain panel effect. Such a scenario may result not only in lessened independence across observations, but may also result in an inaccurate estimate of the daughters effect. (From a causal perspective, this issue represents a possible violation of the Stable Unit

14 Min. 1st Qu. Median Mean 3rd Qu. Max. All Judges 1 5 8 11.10 14 46 Democrats 1 5 7 10.12 13 39 Republicans 1 5 9 11.94 14 46

Table 4: Distribution of the number of gender-related cases heard per judge, 1996-2002.

Treatment Variance Assumption, or SUTVA (Rubin, 1980).) As in the analysis of Boyd, Epstein and Martin(2010), however, we would expect that if judges on a panel influence each other’s decisions, this should tend to produce underestimates for the effects of daughters: i.e., if anything, judges without daughters will, due to the influence of their colleagues, act more like judges with daughters. To address some of these issues empirically, however, we examine as our key outcome variable a judge’s overall track record of voting on gender-related cases by examining the proportion of votes made in a feminist direction. This measure provides a straightforward analogy to Washington(2008)’s Congressional scores and is less sensitive to potential panel effects. (In fact, the composition of Congress changes very little or not at all from vote to vote, while judges are likely to be members of very different panels for each vote.) The distribution of these votes is graphically displayed in Figure1. Another consideration is that, unlike votes taken in Congress, the number of gender- related cases heard by each judge varies (Table4). This variation is natural, and is largely due to (1) different lengths of service, (2) different jurisdictions hearing different numbers of cases, and (3) the random fluctuations associated with (random) case assignment. Thus, we cannot use an ordinary least squares specification with the proportion of cases decided in a feminist direction as the outcome variable; doing so would “weight” as equal judges who cast only one vote with judges who cast dozens, which would violate one of the key OLS assumptions: the variance of the outcome would clearly vary according to whether the judge heard 1 case or 46. We therefore take a weighted least squares (WLS) approach by weighing each judge by the number of cases he or she heard, summary statistics for which are reported

15 in Table4. We note that the substantive results are unaffected when running an unweighted OLS model. Lastly, to assess how other factors could play a role, we at times include in our analysis control variables that are standard in the judicial politics literature (and also used in other “daughters effects” studies – Washington(2008) and Iacus, King and Porro(2011)), including (1) partisanship (as measured by party of the appointing President), (2) gender, (3) age (at time of commission), (4) race, and (5) whether the judge is Catholic. Some of these characteristics may be realized after a judge has decided to have children, which means that their inclusion may introduce post-treatment bias. (For example, a judge who has at least one daughter might be less likely to self-identify with a conservative religious group later on in life, perhaps directly due to the influence of having a girl.) Although mindful of this post-treatment issue, we nonetheless find that our substantive conclusions do not change with the inclusion (or removal) of these covariates.

5 Results

We begin by presenting our core results in Table5, which displays the coefficients of WLS analyses for all 199 judges who had at least one child (Models 1 and 2). Our outcome variable here, and in all subsequent analyses, is the judges’ track record on voting on gender rights cases, which we take to be the proportion of cases that each judge decided in a feminist- leaning direction. Model 1 shows the results when we include the number of girls as a categorical variable – that is, we compare how the effect varies between having no daughters and having up to five daughters, compared among judges who have the same number of children. Substantively, what the first row of Model 1 tells us is that an increase in the number of girls from none to 1, conditional on the total number of children, translates on average into a 9% increase in the proportion of gender-related cases in which a judge will vote

16 All Judges 4 Children or Fewer Non-Gender Cases Model 1 Model 2 Model 3 Model 4 Model 5 Intercept 0.31∗ 0.32∗ 0.31∗ 0.32∗ 0.42∗ (0.04) (0.04) (0.04) (0.04) (0.03) 1 Girl 0.09∗ 0.09∗ (0.04) (0.04) 2 Girls 0.05 0.05 (0.04) (0.04) 3 Girls 0.06 0.08 (0.06) (0.07) 4 Girls -0.35 (0.45) 5 Girls 0.27 (0.17) At Least 1 Girl 0.07 ∗ 0.07 ∗ 0.04 (0.03) (0.04) (0.03) Child Dummy Variables XX XX X N 199 199 182 182 189 R2 0.06 0.04 0.04 0.03 0.04 adj. R2 -0.00 -0.00 0.01 0.01 0.02 Resid. sd 0.63 0.63 0.62 0.62 0.67 Standard errors in parentheses ∗ indicates significance at p < 0.05

Table 5: Weighted least squares results. Outcome is judges’ proportion of feminist (or liberal) votes on gender-related cases (Columns 1-4) and non-gender related cases (Column 5), 1996-2002. All models include fixed effects for total number of children and use weights based on number of cases heard. Columns 3, 4, and 5 include judges with 1-4 children. Column 5 uses a random subset of all published U.S. Court of Appeal cases, 1996-2002 (n = 6,652), collected by songer. in a progressive or feminist direction. Comparable results are given by looking at increases from 0 to 2 girls, 0 to 3 girls, and 0 to 5 girls, although the small number of judges with so many daughters means that these estimates are statistically insignificant. Surprisingly, we see a negative effect when we examine increases from 0 to 4 girls, although the effect is not significant. Nearly the same substantive result is yielded by Model 2, in which we include a dummy variable for whether a judge has at least one daughter. Here, having at least one daughter

17 corresponds to a 7% increase in the proportion of cases in which a judge will vote in a feminist direction. The findings in both Model 1 and Model 2 are significant at the 5% level, meaning that we can be fairly confident of a positive, non-zero relationship between having at least one girl and being more likely to have a higher share of votes be in a progressive or feminist direction. Taken together, the results from Model 1 and Model 2 demonstrate that the greatest effect from having daughters comes simply from having at least one girl; there is no added impact on having additional girls and the effect does not increase linearly. Thus, we move forward by collapsing explanatory variables to look at the effect of having at least one daughter, conditioning on the total number of children (Model 2), and we take this as our “treatment” of interest. We are also mindful of the extremely small number of judges with five or more children. The small sample size in these groups leads to extremely poor balance in terms of fraction of girls (Table3) and results in an inability to precisely estimate the effects among this group, as seen in Model 1. (In fact, of the group of judges with 5 or more children, only one judge has zero girls.) Mindful that these 17 judges could be driving the results, we also present results limiting the sample space to judges with four children or fewer – who collectively represent 92% of all judges. These results are presented in Table5’s Models 3 and 4 and demonstrate that the effect of having daughters is strong and significant: having at least one girl results in an approximate 7% increase in the share of cases in which a judge will vote in a feminist direction. (The results are even stronger if we limit the sample space to the 80% of judges with fewer than four children; because the effect for judges with four children appears mixed, we include the four-child judges in the interest of transparency and presenting conservative results.) Because these results are comparable in all respects, and because they reduce reliance on a small sample of unusually highly fertile judges, we move forward comparing judges with four children or fewer in subsequent analyses. As a point of comparison, we also report a WLS regression of the outcome on having at

18 Model 1 Model 2 Intercept 0.27∗ 0.26∗ (0.05) (0.06) Child is Girl 0.16∗ 0.16∗ (0.07) (0.07) Democrat 0.04 (0.07) N 21 21 R2 0.23 0.24 adj. R2 0.19 0.16 Resid. sd 0.55 0.56 Standard errors in parentheses ∗ indicates significance at p < 0.05

Table 6: Weighted least squares results for judges with 1 child. Outcome is judges’ proportion of feminist votes on gender-related cases (1996-2002). All models include weights for number of cases heard.

least one daughter for non-gender-related cases using a random subset of 6,652 published cases from the U.S. Courts of Appeals, 1996-2002 collected and coded by Songer, Kuersten and Haire(2007). The outcome variable here is the judges’ track records of liberal (not feminist) voting on these non-gender related cases. These results, presented in Table5, Column 5, show that there is no statistically significant relationship between having at least one daughter and voting in a more liberal direction. This is true regardless of model specification, and regardless of whether we include all judges or only judges with four children or fewer. Thus, we have evidence supporting our second hypothesis: to the extent that we see a “daughters effect,” it is only for cases having a gender dimension.

Daughters Effect For Single-Child Judges. A possible concern is that judges are using fertility stopping rules in deciding whether to have children. To provide some evidence on this point, we also display results in Table6 from models that include only the 21 judges with one child. (These do not include child fixed effects, as all the judges have one child.) As noted earlier, fertility stopping rules are less of an issue for this subset. Within this

19 Republicans Only Democrats Only Men Only Women Only Intercept 0.31∗ 0.32∗ 0.29∗ 0.33∗ (0.05) (0.05) (0.05) (0.06) At Least 1 Girl 0.07• 0.03 0.07∗ 0.08 (0.04) (0.05) (0.04) (0.08) 2 Children −0.06 0.14• 0.02 0.09 (0.05) (0.05) (0.05) (0.09) 3 Children −0.02 0.10 0.05 −0.04 (0.06) (0.06) (0.05) (0.10) 4 Children −0.06 0.19∗ 0.05 0.03 (0.07) (0.07) (0.06) (0.13) N 99 85 158 26 R2 0.07 0.12 0.05 0.16 adj. R2 0.03 0.08 0.02 −0.00 Resid. sd 0.60 0.65 0.67 0.75 Standard errors in parentheses ∗ indicates significance at p < 0.05 • indicates significance at p < 0.10

Table 7: Weighted least squares results for judges with four children or fewer. Outcome is judges’ proportion of feminist votes on gender-related cases (1996-2002). All models include fixed effects for total number of children and weights for number of cases heard. subset, having one daughter as opposed to one son is linked to an even higher 16% increase in the proportion of gender-related cases decided in a feminist direction. Despite the smaller sample size, the effect is significant at the 5% level with a p-value of 0.03. The effect persists after including controls for party of appointing President (Table6, Column 2).

Daughters Effect By Party. If the assumption of the random nature of sex assignment (conditional on the number of children) holds, then the models Table5 should identify the causal effect of having at least one daughter, with the caveats discussed below. As we have noted throughout, however, we are interested in further assessing how this effect might vary over population subsets, especially by party. There is some concern about conditioning on party because this variable is measured post- treatment. That is, party identification could change early on in the judge’s career, and,

20 given the comparative elderly age at which judges are nominated, the act of having daughters could sway how a judge will eventually identify and which party’s President would nominate him or her. In other words, because most judges have completed their reproductive years prior to our measurement of their party, we might worry that their partisanship is affected by having daughters. Mindful of this concern, we note that the following analysis assumes that a binary coding of partisanship has not been affected by having daughters. The first two columns of Table7 present results that are disaggregated by the party of the appointing president. Column 1 presents the results for judges that were appointed by a Republican president, and Column 2 presents the results for judges that were appointed by a Democratic president. As before, the outcome variable is the proportion of cases a judge decided in a feminist direction on gender-related cases. Comparing the two columns makes clear that the effect is being driven by Republican judges – the effect for Republican judges is an average 7% increase in the proportion of cases decided in a feminist direction (and significant at the 10% level, with a p-value of 0.09, for judges with fewer than five children, and at the 5% level, with p-value of 0.03, for judges with fewer than four children for judges with three children or fewer) while the effect for Democrat judges is on average 3% (and is not signifiant under any model specification or population subset). Thus, we see some evidence in favor of the idea that the daughters effect works primarily for individuals whose political affiliations suggest they would otherwise be conservative. However, the difference between Democrat and Republican legislators is not statistically significant at traditional levels. This is consistent with the findings by (Washington, 2008). As a robustness check, we note that the inclusion of controls for (1) age, (2) Catholicism, and (3) gender in addition to party of appointing president does not change the overall daughters effect estimate (Table 8, below), which is still significant for all judges at the 5% level.

Daughters Effect By Gender. Washington(2008) found that the daughters effect for

21 All Judges Republicans Only Intercept 0.14 0.12 (0.16) (0.19) At Least 1 Girl 0.09∗ 0.08• (0.04) (0.04) 2 Children −0.02 −0.03 (0.06) (0.07) 3 Children −0.01 −0.01 (0.06) (0.07) 4 Children −0.05 −0.08 (0.07) (0.08) Age 0.00 0.00 (0.00) (0.00) Catholic −0.06 −0.03 (0.04) (0.04) Woman −0.04 −0.02 (0.05) (0.07) Democrat 0.14∗ (0.04) African American −0.03 (0.08) N 130 85 R2 0.18 0.06 adj. R2 0.12 −0.02 Resid. sd 0.59 0.58 Standard errors in parentheses ∗ indicates significance at p < 0.05 • indicates significance at p < 0.10

Table 8: Weighted least squares results for judges with four children or fewer. Outcome is judges’ proportion of feminist votes on gender-related cases (1996-2002). All models include fixed effects for total number of children and weights for number of cases heard.

Congressional representatives was particularly strong for male legislators while insignificant for female legislators. For judges, we would also expect to see a daughters effect primarily for men and not for women. One difficulty with this is, however, that the pool of female judges differs markedly from the pool of male judges. As Table2 shows, there are fewer women than men on the appellate courts during this period. Additionally, women judges

22 are much more likely than male judges to have no children (29% versus 8%) and on average have fewer numbers of children than do male judges (1.58 versus 2.66); consequently, there are also fewer women (n = 38) than men (n = 186) in the sample. Despite the low sample size, we show some results from models including the gender of the judge (Columns 3 and 4 of Table7). The results are only partially consistent with the theoretical predictions. We do find a strong and significant effect for the male judges (Column 3), which confirms the prediction that an effect will exist for men. However, the effect for female judges is similar in magnitude (although it is insignificant at this sample size). Given the incredibly small number of women, and the even smaller of women who have children, we cannot rule out that there is no relationship between parenting girls (as opposed to sons) and voting in a more feminist direction on gender-related cases. We therefore abstain from making firm conclusions about this hypothesis.

Daughters Effect After Controlling for Additional Covariates. We conclude this discussion by noting that the overall effect of having daughters persists with the inclusion of other covariates, including (1) age, (2) Catholic religion, (3) judge gender, (4) party of appointing President, and (4) race. There are very few judges of minority backgrounds, with 91% of all judges in the sample being white; here we use the distinction between white and black judges, who comprise 7% of the population (and we cannot even make this comparison for Republicans, among whom there are no black judges). Results are presented in Table8, and show that the overall daughters effect is not perturbed by the inclusion of additional covariates. For Republicans, for whom we previously saw the greatest effect, the effect is significant at the 10% level for judges with fewer than five children, but the magnitude of the effect is otherwise comparable.

23 6 Alternative Explanations of the Results

To summarize, the results presented suggest that there exists a positive effect of having girls on whether a judge votes in a liberal direction on gender-related cases. The effect is most striking when we look at whether a judge has at least one girl and appears driven primarily by Republican appointments. Taken together, the results suggest that the shift in voting comes from the act of parenting daughters, which possibly cause judges to learn about women’s issues (Reingold and Foust, 1998; Bolzendahl and Myers, 2004), or causes them to change or re-align their preferences (Gerson, 1993; Kane and Sanchez, 1994; Coltrane, 1997; Bolzendahl and Myers, 2004). It is also likely that judges feel a strong emotional attachment to their daughters and so, when they make a decision, they do so in a way that is favorable to women, or that they rely on motivated reasoning through process of emotion. Other theories may, however, also explain the daughters effect, bringing with them dra- matically different substantive interpretations. A key aspect of our analysis – and also of Washington(2008) and Iacus, King and Porro(2011) – has been that the likelihood of a child being born a boy or a girl is out of the parents’ control and, once a couple decides to have a child, the child’s gender is unrelated to any factors that could affect decision making – e.g., gender, political ideology, religion, race, etc.9 However, because we could not obtain information on the birth order of children, we could only condition on the number of children, and we are therefore concerned about the effects of fertility stopping rules.

9The possibility does exist, however, that in fact the likelihood of having either a son or daughter varies from one person to another according to specific traits. For example, it may be the case that more con- servative people have a higher chance of have more boys, or that more progressive or liberal people have a higher chance of having girls. Although no study to our knowledge has made these ideologically based claims, studies conducted by Kanazawa have put forth the possibility that more physically attractive peo- ple have more daughters (Kanazawa, 2007) and that taller and heavier people (Kanazawa, 2005), battered women (Kanazawa, 2008), and violent men (Kanazawa, 2006) tend to have a disproportionate share of sons. Despite Kanazawa’s research, we note that this literature is hotly contested (Gelman and Weakliem(2009) and Denny(2008) provide rebuttals) and that, with the possible exception of violent tendencies, it is unlikely that these other traits would influence judicial decision making.

24 Fertility Stopping Rules. We cannot ignore the possibility that people with certain beliefs may be more inclined to have sons over daughters (or vice versa) – and that this could mislead us into thinking that a daughters effect exists. Of particular concern is that judges in our sample may be using fertility stopping rules to control the gender balance of one’s offspring. Consider Table3, which provides some limited suggestion of fertility stopping rules: Democrats are more likely to have one boy and one girl, while Republicans more likely to have two boys. This might suggest that Republicans are content with having two boys and do not choose to have more children to achieve gender parity. Although we do not rule out that judges might be using fertility stopping rules, they do not appear to be driving the results here for three reasons. First, as we have discussed in Section5, the daughters effect remains intact (and even increases in magnitude) when we examine judges with only one child. These are judges for whom concerns about fertility stopping rules are minimal. Second, if true that more liberal judges continue having children until they have at least one girl (or that more conservative judges are content to “stop” with having all boys), then this would mean that having girls is correlated with across-the-board liberal beliefs – which would translate into liberal voting across all cases, not just ones with a substantial gender dimension. In this analysis, however, we see an effect of having daughters, but only for gender-related cases. Lastly, although we do see in Table3 some evidence of the possible use of fertility stopping rules across party, we have presented evidence that the daughters effect persists conditioning on party; in other words, we can see some effect of having girls even when we subset the data to examine just Republicans (and not at all when we subset to Democrats). For fertility stopping rules to call this into question would mean that there would have to be differentially applied stopping rules within all Republicans, something for which we have no evidence.

25 Selection Bias. Finally, as in Washington(2008) and Iacus, King and Porro(2011), we must consider the potential for selection bias in how appeals court judges are appointed. As we have emphasized elsewhere, we assume that having child is a fundamentally life-altering experience, and that anything that happens following the birth of a child will be deeply affected by that experience. This raises the question of whether the decision to become a judge, or the decision by the President to appoint a judge to the federal appellate courts, is affected by the number of daughters an individual has (conditional on the number of children one has), and whether this effect results in conservative judges having fewer daughters. We are not aware of any evidence that Presidents take into account the number or gender of potential nominees’ children when making judicial appointments, and the initial career decisions of top-tier law students are often made prior to having children. If a selection effect is responsible for a daughters effect, then it is likely occurring once individuals have already started their legal careers. Even if such selection is occurring, different theories of selection would imply different observed effects. For example, if having sons increases an individual’s desire for high paying jobs (as suggested by Lundberg and Rose(2002)), then having sons might make a career in the judiciary less attractive than a career in the private sector. If this primarily affects conservative judges in the lower-tier courts, then they may choose to exit the judiciary. Alternatively, having daughters may have a generally liberalizing effect, and if appellate court judges are usually selected from among highly conservative lower court judges (by Republican Presidents) and highly liberal lower court judges (by Democrat Presidents), then some conservative judges with daughters might select (or be selected) out of the judiciary. While concerns about selection could complicate the interpretation of the results, it is not clear whether either explanation would indicate an estimated effect for all cases or just for gender-related cases. For example, if a selection effect is due to selection by the president, we would likely see this effect for all cases, as presidents are unlikely to select judges only on the

26 basis of one type of case. Similarly, if having daughters tends to drive generally conservative judges out of the judiciary, then we would expect to see a daughters effect for all types of cases, not just gender-related cases. We do not, however: we see only an effect for having daughters for cases having to do with gender-related issues. Thus, it seems unlikely that the effects seen here derive exclusively from a selection bias story.

7 Why Empathy Matters in Judicial Selection

Political science scholarship on the courts has mostly focused on the impact on judicial decision making of two salient attributes: preferences (e.g., ideology, partisanship) and as- criptive characteristics (e.g., race, gender). Less well-developed is the notion that personal relationships and experiences – including contacts with close family members – could in- fluence decision making. In this paper, we presented evidence that some types of personal relationships do in fact affect judges’ actions. Indeed, across cases involving gender issues, judges who parent daughters as opposed to sons are more likely to reach liberal decisions. This might be the case because having daughters causes judges to learn about women’s is- sues (Reingold and Foust, 1998; Bolzendahl and Myers, 2004) or because their preferences simply change over time (Gerson, 1993; Kane and Sanchez, 1994; Coltrane, 1997). Having daughters might also forge strong emotional attachments, leading judges to cast votes more generally in favor of women. Whatever the reason, these results are squarely in line with similar studies in a Congressional context (Washington, 2008; Iacus, King and Porro, 2011) as well as those in the public opinion context (Warner, 1991; Warner and Steel, 1999; Shafer and Malhotra, 2011). Two points are further worth noting. First, this effect exists when decision making con- cerns gender-related issues – discrimination, sex discrimination, pregnancy discrimination, etc. The effect does not extend to other areas, including racial or ethnic discrimination,

27 or to a liberal or progressive voting inclination more generally. These results are, to this extent, in accord with those of Washington(2008), which found the effects of having daugh- ters for Congressmembers’ NOW and AAUW scores, and with Boyd, Epstein and Martin (2010), which found effects associated with being a female (as opposed to male judge) in cases involving sex or gender discrimination issues. These results are also inconsistent with a general selection bias – that conservative lawyers with daughters select (or are selected) out of the judiciary and/or that liberal lawyers with sons select (or are selected) out of the judiciary. Second, the effect is robust to different model specifications and it persists even after controlling for those personal characteristics that judicial politics scholars routinely take into account (e.g., age, gender, race, Catholicism). Eliminating judges with extremely large numbers of children only strengthens the effect. Although more research is needed to identify the exact pathway of the effect (learning versus motivated reasoning), having daughters does cause judges to vote differently. Taken together, our empirical findings have several implications for judicial processes. The first is that personal relationships do indeed hold the possibility of shaping views. We show here that decision making appears to be shaped not just by ideology or, in the legal context, applicable legal precedent, but also by a host of other considerations, including intimate experiences and personal connections. In this immediate paper, we have shown the effect via having daughters, but we might think the effect exists via other pathways – for example, having a gay or lesbian child or sibling (e.g., Dick Cheney or Newt Gingrich), or having a disabled or mentally ill son or daughter (e.g., Sarah Palin). We encourage schol- arship to explore these pathways actively; empathy does shape judicial decisions, and the natural quasi-experiment of having daughters provides only one kind of empirical example. Second, this fact has broader implications on descriptive representation on the courts. Scholarship has demonstrated that women judges decide cases differently from men (Boyd, Epstein and Martin, 2010), and that African Americans also decide cases differently from

28 whites (Kastellec, 2012). However, what we see here is that white male judges who have daughters are more likely to vote in a liberal direction – despite not having those ascriptive characteristics that would otherwise be linked to progressive views on women’s rights. To this extent, despite Sonia Sotomayor’s comment that “a wise Latina woman with the richness of her experiences would more often than not reach a better conclusion than a white male who hasn’t lived that life,” we find that empathy is a cross-cutting effect. Non-minority judges can change their views to reflect personal relationships and experiences. Those experiences might vary according to background; what the quasi-experiment of daughters shows is that empathy is not, and need not, limited to one group or one party.

29 References

Ashenfelter, Orley, Theodore Eisenberg and Stewart J. Schwab. 1995. “Politics and the judiciary: The influence of judicial background on case outcomes.” The Journal of Legal Studies 24(2):257–281.5 Baldez, L., L. Epstein and A.D. Martin. 2006. “Does the US Constitution Need an ERA?” Journal of Legal Studies 35(1):243–283.5 Bazelon, Emily. 2009. “The Place of Women on the Court.”. URL: http: // www. nytimes. com/ 2009/ 07/ 12/ magazine/ 12ginsburg-t. html 3 Bogan, Vicki. 2009. “Investment Decisions and Offspring Gender.” Working Papers .6 Bolzendahl, Catherine I. and Daniel J. Myers. 2004. “Feminist attitudes and support for gender equality: Opinion change in women and men, 1974-1998.” Social Forces 83(2):759– 789. 24, 27 Boyd, Christina L., Lee Epstein and Andrew D. Martin. 2010. “Untangling the Causal Effects of Sex on Judging.” American Journal of Political Science 54:389–411.4,8, 14, 28 Campbell, Angus, Philip E. Converse, Warren E. Miller and Donald E. Stokes. 1966. “The American Voter.”.5 Clare, Jeanna E. and Clyde V. Kiser. 1951. “Social and psychological factors affecting fertility. XIV. Preference for children of given sex in relation to fertility.” The Milbank Memorial Fund Quarterly 29(4):440–492.8 Clark, Shelley. 2000. “Son preference and sex composition of children: Evidence from India.” Demography 37(1):95–108.8 Coltrane, Scott. 1997. Family man: Fatherhood, housework, and gender equity. Oxford University Press, USA. 24, 27 Conley, Dalton and Emily Rauscher. 2010. “The Effect of Daughters on Partisanship.” NBER Working Paper .6 Cooper, Cynthia L. 2004. “Daughter of Justice Blackmun Goes Public about Roe.” WeNews . URL: http: // www. womensenews. org/ story/ the-nation/ 040229/ daughter-justice-blackmun-goes-public-about-roe 3 Cox, Adam B. and Thomas J. Miles. 2008. “Judging the Voting Rights Act.” Columbia Law Review pp. 1–54.5 Crowe, N.E. 1999. “The effects of judges’ sex and race on judicial decision making on the US courts of appeals, 1981-1996.”.5 Dahl, Gordon B. and Enrico Moretti. 2004. “The demand for sons: Evidence from divorce, fertility, and shotgun marriage.” NBER Working Paper Series .6 Davis, Sue, Susan Haire and Donald R. Songer. 1993. “Voting behavior and gender on the US Courts of Appeals.” Judicature 77:129.5 Dawes, Christopher T. and James H. Fowler. 2009. “Partisanship, voting, and the dopamine D2 receptor gene.” The Journal of Politics 71(03):1157–1171.5 Denny, Kevin. 2008. “Big and tall parents do not have more sons.” Journal of theoretical biology 250(4):752–753. 24 Epstein, Lee, Jack Knight and Andrew D. Martin. 2003. “Norm of Prior Judicial Experience

30 and Its Consequences for Career Diversity on the US Supreme Court, The.” Cal. L. Rev. 91:903.5 Freedman, Deborah S., Ronald Freedman and Pascal K. Whelpton. 1960. “Size of family and preference for children of each sex.” American Journal of Sociology 66(2):141–146.8 Garrett, Major. 2009. “Obama Pushes for ’Empathetic’ Supreme Court Justices.”. URL: http://www.foxnews.com/politics/2009/05/01/obama-pushes-empathetic-supreme- court-justices 2 Gelman, Andrew and David Weakliem. 2009. “Of Beauty, Sex and Power: Too little attention has been paid to the statistical challenges in estimating small effects.” American Scientist 97(4):310–316. 24 Gerson, K. 1993. No man’s land: Men’s changing commitments to family and work. Basic Books. 24, 27 Gottschall, Jon. 1983. “Carter’s Judicial Appointments: The Influence of Affirmative Action and Merit Selection on Voting on the US Courts of Appeals.” Judicature 67:165.5 Green, Donald P., Bradley Palmquist and Eric Schickler. 2004. Partisan hearts and minds: Political parties and the social identities of voters. Yale Univ Pr.5 Howard, Robert M. and Jeffrey A. Segal. 2002. “An Original Look at Originalism.” Law and Society Review 36:113.4 Iacus, Stefano M., Gary King and Giuseppe Porro. 2011. “Multivariate matching methods that are monotonic imbalance bounding.” Journal of the American Statistical Association .6, 16, 24, 26, 27 Just, Richard. 2009. “The Empathy War.”. URL: http://www.tnr.com/blog/the-plank/the-empathy-war 3 Kanazawa, Satoshi. 2005. “Big and tall parents have more sons: Further generalizations of the Trivers-Willard hypothesis.” Journal of Theoretical Biology 235(4):583–590. 24 Kanazawa, Satoshi. 2006. “Violent men have more sons: Further evidence for the generalized Trivers-Willard hypothesis (gTWH).” Journal of Theoretical Biology 239(4):450–459. 24 Kanazawa, Satoshi. 2007. “Beautiful parents have more daughters: A further implication of the generalized Trivers-Willard hypothesis.” Journal of Theoretical Biology 244(1):133– 140. 24 Kanazawa, Satoshi. 2008. “Battered women have more sons: A possible evolutionary reason why some battered women stay.” Journal of Evolutionary Psychology 6(2):129–139. 24 Kane, Emily W. and Laura Sanchez. 1994. “Family status and criticism of gender inequality at home and at work.” Social Forces 72(4):1079–1102. 24, 27 Kastellec, Jonathan P. 2012. “Racial Diversity and Judicial Influence on Appellate Courts.” American Journal of Political Science .5, 29 Kulik, C.T., E.L. Perry and M.B. Pepper. 2003. “Here comes the judge: The influence of judge personal characteristics on federal sexual harassment case outcomes.” Law and Human Behavior 27(1):69–86.5 Lithwick, Dahlia and Sonja West. 2010. “The Unsung Empathy of Justice Stevens: Justice John Paul Stevens is the model for why empathy matters.”. URL: http: // www. slate. com/ articles/ news_ and_ politics/ jurisprudence/

31 2010/ 04/ the_ unsung_ empathy_ of_ justice_ stevens. html 2 Lundberg, Shelly and Elaina Rose. 2002. “The effects of sons and daughters on men’s labor supply and wages.” Review of Economics and Statistics 84(2):251–268.6, 26 Manning, Kenneth L., Bruce A. Carroll and Robert A. Carp. 2004. “Does Age Matter? Judicial Decision Making in Age Discrimination Cases.” Social Science Quarterly 85(1):1– 18.5 Massie, T., S.W. Johnson and S.M. Gubala. 2002. The impact of gender and race in the decisions of judges on the United States courts of appeals. In annual meeting of Midwest Political Science Association, Chicago IL.5 Novak, V. 2003. “Justice Rehnquist, Secret Feminist?”.3 Obama, Barack. 2009. “The President’s Remarks on Justice Souter.”. URL: http: // www. whitehouse. gov/ blog/ 09/ 05/ 01/ The-Presidents-Remarks-on-Justice-Souter 2 Oswald, Andrew J. and Nattavudh Powdthavee. 2010. “Daughters and left-wing voting.” The Review of Economics and Statistics 92(2):213–227.6 Peresie, Jennifer L. 2005. “Female Judges Matter: Gender and Collegial Decisionmaking in the Federal Appellate Courts.” Yale Law Journal 114(7):1759–1892.5 Pollard, Michael S. and S.P. Morgan. 2002. “Emerging parental gender indifference? Sex composition of children and the third birth.” American Sociological Review 67(4):600.8 Reingold, Beth and Heather Foust. 1998. “Exploring the Determinants of Feminist Con- sciousness in the United States.” Women and politics 19:19–48. 24, 27 Rohde, David W. and Harold J. Spaeth. 1976. Supreme Court Decision Making. WH Freeman.4 Rubin, D.B. 1980. “Comment on paper by D. Basu: Randomization Analysis of Experimental Data: The Fisher Randomization Test.” Journal of the American Statistical Association 75:591–593. 14 Scherer, Nancy. 2004. “Blacks on the Bench.” Political Science Quarterly 119(4):655–675.5 Schubert, Glendon A. 1974. The Judicial Mind revisited: Psychometric Analysis of Supreme Court Ideology. Oxford University Press.4 Segal, Jeffrey A. and Harold J. Spaeth. 2002. The Supreme Court and the attitudinal model revisited. Cambridge Univ Pr.4,5 Shafer, Emily F. and Neil Malhotra. 2011. “The Effect of a Child’s Sex on Support for Traditional Gender Roles.” Social Forces 90(1):209–222.6, 27 Sisk, G.C., M. Heise and A.P. Morriss. 1998. “Charting the Influences on the Judicial Mind: An Empirical Study of Judicial Reasoning.” New York University Law Review 73(5):1377– 500.5 Sloane, Douglas M. and Che-Fu Lee. 1983. “Sex of previous children and intentions for further births in the United States, 1965-1976.” Demography 20(3):353–367.8 Songer, Donald, Ahslyn K. Kuersten and Susan B. Haire. 2007. “The United States Courts of Appeals Database.”. 11, 18 Sunstein, Cass R., David Schkade, Lisa M. Ellman and Andres Sawicki. 2006. Are Judges Political? Brookings Institution Press.4

32 Warner, Rebecca L. 1991. “Does the sex of your children matter? Support for feminism among women and men in the United States and Canada.” Journal of Marriage and the Family 53(4):1051–1056.6, 27 Warner, Rebecca L. and Brent S. Steel. 1999. “Child rearing as a mechanism for social change: The Relationship of child Gender to parents’ Commitment to Gender Equity.” Gender and Society 13(4):503–517.6, 27 Washington, Ebonya. 2008. “Female Socialization: How Daughters Affect Their Legislator Fathers’ Voting on Women’s Issues.” American Economic Review 98(1):311–332.3,6,7, 9, 13, 15, 16, 21, 24, 26, 27, 28 Yamaguchi, Kazou and Linda R. Ferguson. 1995. “The stopping and spacing of child- births and their birth-history predictors: rational-choice theory and event-history anal- ysis.” American Sociological Review 60(2):272–298.8 Zuckerman, Alan S., Josip Dasovi´cand Jennifer Fitzgerald. 2007. Partisan families: The social logic of bounded partisanship in Germany and Britain. Cambridge Univ Pr.5 Zuk, Gary, Deborah J. Barrow and Gerard Gryski. 2009. “Multi-User Database on the Attributes of United States Appeals Court Judges, 1801-2000.”. ICPSR06796-v2. Ann Arbor, MI: Inter-university Consortium for Political and Social Research. URL: http: // www. icpsr. umich. edu/ icpsrweb/ ICPSR/ studies/ 6796 10

33

Faculty Biographies

Denny Chin is a Judge, U.S. Court of Appeals for the Second Circuit. Previously he was a Judge, District Court, Southern District of New York, as well as a Law clerk, Hon. Henry Werker, U.S. District Court, Southern District of New York; Private practice in New York City; Assistant U.S. attorney, Southern District of New York; and an Adjunct professor (part-time), Fordham University School of Law. He received a BA from Princeton University and a J.D. from Fordham University School of Law.

Betty Weinberg Ellerin is senior counsel in Alston & Bird LLP’s Litigation & Trial Practice Group in the New York office. Prior to joining Alston & Bird, Justice Weinberg Ellerin served more than 20 years as an Appellate Division jurist. She was the first woman appointed as Deputy Chief Administrative Judge of the State of New York for the New York City Courts. Following that position, she became the first woman appointed as Associate Justice of the Appellate Division of the Supreme Court of the State of New York (First Department). Additional positions include Presiding Justice and Additional Justice of the Appellate Division of the Supreme Court of the State of New York (First Department). Prior to her tenure on the bench, she served as Law Clerk to various State Supreme Court judges. Justice Weinberg Ellerin received her LL.B. in 1952 from New York University, where she was awarded the Florence Allen Scholarship. She received her B.A., cum laude, in 1950 from Washington Square College under the combined college-law program, where she was elected to Justinian, the Pre-Law Honor Society. Judith J. Gische is an Associate Justice of the NY Supreme Court, Appellate Division, First Department since October 2012. Previously she was a Justice of the New York State Supreme Court and a Judge of the Civil Court of the City of New York. She also served as a Judicial Clerk at the Appellate Division of the Supreme Court, Third Department and as an associate attorney at Richenthal, Abrams and Moss. She received a BA from the State University of New York at Buffalo, Magna Cum laude, Phi Beta Kappa, and a J.D. from the State University of New York at Buffalo.

La Tia W. Martin is a Justice, New York State Supreme Court and Supervising Justice of Matrimonial Matters, Bronx County. Judge Martin has also been a Judge in both Civil and Criminal Courts of the City of New York, as well as General Counsel, NYC Office of the Sheriff; Inspector General, NYC Department of Finance, Tax Commission, Financial Information Services Agency, and Office of Payroll; Inspector general, NYC Department of Consumer Affairs; and Assistant District Attorney, Bronx County District Attorney's Office. Judge Martin earned a BS from Boston University and a JD from Rutgers University.

Michael S. Ross is the principal of the Law Offices of Michael S. Ross, where he concentrates his practice in attorney ethics and criminal law. He is a former Assistant United States Attorney in the Criminal Division of the Southern District of New York and also served as an Assistant District Attorney in Kings County. Mr. Ross has been an Adjunct Professor at the Benjamin N. Cardozo School of Law for twenty- four years. He received a B.A. from Rutgers University and a J.D. from New York university School of Law.

Prof. Maya Sen, is an Assistant Professor of Political Science, University of Rochester. Her current research explores the relationship between race and ethnicity and judicial decision making in the federal courts. In June of 2010, she and Jennifer Hochschild were awarded a grant by the Robert Wood Johnson Foundation to study the relationship between genetics, race, and public policy. She is also an affiliate of the Frederick Douglass Institute for African and African-American Studies. She received a JD from Stanford and a PhD from Harvard.