Definitions

Character is a contested concept. It can mean different things to different audiences, and can change over time and across culture. One of the most common meanings, as reflected in the Oxford English Dictionary (2016), is “the sum of the moral and mental qualities which distinguish an individual or a people.” Psychologists similarly view character as patterns of thought, emotion, and behavior (Cohen et al. Reference Cohen, Panter, Turan, Morse and Kim2014, 944). Although by definition the term is morally neutral, in common usage it often carries moral connotations. Calling someone a person of character generally implies that he or she is virtuous. Good character is thought to encompass “moral knowing, moral feeling and moral action … knowing the good, desiring the good, and doing the good—habits of the mind, habits of the heart and habits of behavior” (Lickona Reference Lickona1999, 78). Difficulties arise, however, when a person possesses some virtues that we associate with good moral character but is conspicuously lacking in others. Terrorists may have courage, loyalty, and steadfast adherence to principle, but we would not ordinarily speak of them as persons of character (Doris Reference Doris2002, 18).

Psychological Research

Whether people's character is fixed over time and across situations has sparked a rich scholarly debate. In the early half of the twentieth century, the prevailing view among psychologists was that individuals had stable internal traits that could predict their conduct in divergent situations (Allport Reference Allport1937; Funder Reference Funder1994, 125; McCrae and Costa Reference McCrae and Costa1994, 173). As psychologist Walter Mischel (Reference Mischel1968, 1487) noted, “the initial assumptions of the trait-state theory were logical, inherently plausible, and also consistent with common sense and intuitive impressions about personality. Their real limitation turned out to be empirical—they simply have not been supported adequately.” In the 1960s, a wide variety of studies cast doubt on this trait theory of character (Bowers Reference Bowers1973; Roberts and Pomerantz Reference Roberts and Pomerantz2004). A new model of behavior known as situationism emerged. This school of thought has argued that situational factors are often better predictors of behavior than personal traits (Mischel Reference Mischel1968, 146, 177; Doris Reference Doris2002, 2). Situationists see our tendency to attribute behavior to a person's stable character as a “fundamental attribution error” (Ross and Nisbett Reference Ross and Nisbett1991; Harman Reference Harman1999, 2).

Situationists point to a broad array of research. The most famous example involves Stanley Milgram's obedience studies, replicated many times in this and other countries. They showed that about two-thirds of participants were willing to deliver apparently dangerous electric shocks to a subject when instructed to do so (Milgram Reference Milgram1974; Brown Reference Brown1986). In one variation of his experiment, Milgram placed the participant on a team with a confederate who was working for Milgram. When the confederate uncomplainingly obeyed instructions to continue the shocks, 90 percent of the participants went along. When the confederate refused to administer high-level shocks and walked away from the experiment, only 10 percent of the participants complied, again demonstrating the importance of situational influences (Milgram Reference Milgram1974, 116–22).

Studies of helping behavior similarly document the role of contextual factors. In a Columbia study, participants were filling out a questionnaire when they heard a loud crash and a woman's cries of pain in the next room. Seventy percent of participants offered help when they were working alone. When they were working with a confederate who ignored the cries, only 7 percent intervened (Latané and Rodin Reference Latané and Rodin1969). Darley and Batson's (Reference Darley and Batson1973) study of students at Princeton's Theological Seminary found that if they were in a hurry, 90 percent walked by someone slumped in a doorway ostensibly experiencing some distress. If they were not in a hurry, almost two-thirds helped. Even seminary students about to give a lecture on the Good Samaritan were no more likely to help than other students. In fact, when under time pressure, those about to give their talks on helping behavior literally stepped over the victim as they hurried on their way (Darley and Batson Reference Darley and Batson1973). Taken together, these experimental findings call into question the stability of character across different contexts (Mischel Reference Mischel1968, 6).

Although such evidence has sometimes been taken to pit contextual forces against character traits, most contemporary experts “see the two as intricately related” (Monin and Jordan Reference Monin, Jordan, Narváez and Lapsley2009, 347). The mainstream view, sometimes labeled interactionism, understands behavior as a function of the interaction between situation and traits (Endler and Magnusson Reference Endler and Magnusson1976; Funder Reference Funder2006). In effect, this approach finds stable and distinctive patterns underlying what seems to be variability in behavior. For example, by probing for the reasons why an individual might cheat in one situation but not in another, researchers can often find a reliable pattern of conduct (Mischel and Shoda Reference Mischel and Shoda1995; Mischel Reference Mischel, Cervone and Shoda1999), but they need a considerable amount of information about past actions to make accurate predictions of future ethical behavior (Mischel and Shoda Reference Mischel and Shoda1995, 246). Moreover, character qualities are not static even in adults. Some individuals become wiser and more giving over time; others become more selfish (Noam Reference Noam, Noam and Fischer1996; Weissbourd Reference Weissbourd2003). Character can be cultivated, and its openness to change counsels caution before branding someone permanently discreditable (Peterson and Seligman Reference Peterson and Seligman2004, 27).

So, too, people's susceptibility to situational influences argues against overgeneralization about particular character traits. A disloyal spouse is not necessarily a disloyal friend. Presidents can be faithful husbands and unfaithful public servants and the converse is also true, as examples such as Richard Nixon and Franklin D. Roosevelt remind us (Rhode Reference Rhode2016, 156).

In short, individuals can be exemplary with respect to some key traits but not others, and their character-related qualities can vary across situations and evolve over time (Doris Reference Doris2002, 115). That should make us wary about sweeping claims that when it comes to character, individuals either have it or they do not. Yet as subsequent discussion suggests, legal decision makers too often make such misleading generalizations about moral character based on a single unrepresentative “bad act.”

The Traditional Justifications for Character Requirements

Moral character as an occupational credential has an extended history. For lawyers, the requirement dates to the Roman Theodosian Code, a compilation of laws issued sixteen centuries ago. The code mandated that legal advocates be of “suitable character,” with past lives that were praiseworthy (Pharr Reference Pharr1952, 32). In the United States, moral character requirements have long been an integral part of occupational licensing systems. These requirements serve two primary purposes. The first is predictive; the law assumes that the moral character required for a particular position is a stable attribute that licensing authorities can accurately predict based on past behavior. An individual either has it or does not. The research reviewed in Section I casts doubt on this predictive capacity. Ethical conduct is a function of a complex interrelationship between traits and situations. Regulators would need to know a considerable amount about why individuals acted the way they did in the past in order accurately to predict how they will act in the future. Licensing authorities frequently lack that information.

A second purpose of character requirements is symbolic. Requiring applicants to have unblemished backgrounds expresses aspirational norms and serves reputational interests. Members of an occupation want to believe, and want the public to believe, that they are exceptionally ethical. The discussion below casts doubt on the wisdom of letting an occupation's concerns for its reputation trump values of fairness and rehabilitation for applicants.

The Flawed System of Character Review

During America's first two centuries, formal standards for occupations were “few and loose” (Gilb Reference Gilb1966, 60). Professions other than those of law and medicine were seldom subject to state licensing (Gellhorn Reference Gellhorn1956, 106). Then, toward the end of the nineteenth century, regulation increased in response to lobbying by the professions and occupations themselves, many of which were “troubled by competition” (Friedman Reference Friedman1965, 500). Around the turn of the twentieth century, accountants, architects, engineers, and nurses began to come under oversight, and then, vocational licensing spread. By the mid-twentieth century, the United States had more than twelve hundred licensing laws, an average of twenty-five per state (Gilb Reference Gilb1966, 61). Regulation extended not only to established professions, but also to egg graders, guide-dog trainers, yacht salesmen, potato growers, bee keepers, septic tank cleaners, and tree surgeons (Gellhorn Reference Gellhorn1956, 106; Reference Gellhorn1976). Statutes typically required good moral character, which excluded those with felony convictions regardless of the nature of the felony or its relevance to the intended occupation (Gellhorn Reference Gellhorn1976). As law professor Walter Gellhorn noted, a “blanket proscription of this sort seems more vindictively punitive than it does selectively preventive” (Gellhorn Reference Gellhorn1956, 138). Was the public really safer if watchmakers cannot be licensed after conviction of a felony? (Gellhorn Reference Gellhorn1956). And what justified requiring commercial photographers to pass a test for syphilis? (Ga. Code Ann. 1975).

Equally problematic were Cold War requirements of loyalty oaths or prohibitions on subversive conduct by groups as diverse as professional boxers, pharmacists, and piano tuners. Veterinarians in some states could not “minister to an ailing cow or cat unless they [had] first signed a non-Communist oath, thus assuring that they … [would] not indoctrinate their four legged patients” (Gellhorn Reference Gellhorn1956, 130).

Although current licensing processes are no longer concerned with loyalty, they remain inconsistent and idiosyncratic, as well as overinclusive. And it matters. More than eleven hundred occupations are licensed in at least one state, and nearly 30 percent of the workforce is covered by licensing laws (Kleiner Reference Kleiner2006, 148; Department of the Treasury Office of Economic Policy et al. 2015; Cohen Reference Cohen2016, B1; Council on Licensure Enforcement and Regulation 2016). Almost all these occupations require good moral character or the functional equivalent, although for many, the need for character screening is scarcely self-evident. Frog farmers, florists, fortune tellers, street artists, upholstery repairers, and massage therapists (even those for animals) are subject to licensing requirements in some states (Kleiner Reference Kleiner2006, 14; Seville Reference Seville2011; Patterson Reference Patterson2015; Department of the Treasury Office of Economic Policy et al. 2015, 4; Cohen Reference Cohen2016, B5; Consumer Affairs and Business Regulation, Massachusetts 2017).

In most cases, review is fairly cursory, and focuses on whether an applicant has been convicted of a felony or had a license revoked or suspended (Johnson and Campbell, Reference Johnson and Campbell2002). In some jurisdictions, however, the inquiry seems excessively intrusive. For Texas mortgage brokers, licensing officials can consider factors like “reliability,” “mental and emotional stability,” “strong community ties,” and “petty offenses” (Craddock Reference Craddock2008, 464–66). In Iowa, the inquiry into good moral character extends not only to applicants for liquor licenses, but also to their spouses (Iowa Alcoholic Beverages Division 2017). Other jurisdictions impose requirements that seem like mindless formalities. For example, the city of Santa Barbara requires massage therapists to provide five affidavits of good moral character by local residents (City of Santa Barbara 2016).

Although a few professions, particularly law and medicine, engage in much more searching scrutiny, they often focus on conduct that bears only a tangential relationship to professional practice. A case in point involves a Nevada board's decision to discipline a chiropractor who was convicted of involuntary manslaughter for shoving a man at a car wash (Sawicki Reference Sawicki2010, 286). Another example is the requirement that acupuncturists be current in child support obligations as a condition of receiving a license (Dittman v. California 1999, 1032). Of the twenty-two California chiropractors recently disciplined for criminal offenses, only four involved offenses related to chiropractic obligations (California Board of Chiropractic Examiners 2015−2016).

One major effect of these laws is to exclude individuals with a criminal record, regardless of the nature of the crime or its relevance to the intended occupation. At least twenty-seven thousand statutes across the states affect the employment of ex-offenders (Hunt et al. Reference Hunt, Bowers and Miller1974; Cohen Reference Cohen2016, B5). State and federal restrictions bar formerly incarcerated individuals from upward of three-hundred-fifty public employment occupations (Hebenton and Thomas Reference Hebenton and Thomas1993, 111). Under many licensing laws, a felony conviction is an automatic disqualification. Half the states deny a beautician license to former felons (Hunt et al. Reference Hunt, Bowers and Miller1974). In Pennsylvania, a person who had been a health worker for three decades attempted to change jobs and learned that a new law barred him from any healthcare position due to conviction for possession of marijuana thirty years earlier (Butterfield Reference Butterfield2002). In Wisconsin, a grandmother had her home daycare license revoked based on a thirty-year-old misdemeanor conviction for overpayment of public assistance even though she had been effectively running a home for over a decade (Rodriguez and Avery Reference Rodriguez and Avery2016, 1). In Texas, marijuana offenses can disqualify applicants from becoming mortgage brokers (Craddock Reference Craddock2008, 55).

Licensing laws vary in their stringency. A survey by the American Bar Association (ABA) reported over twelve thousand disqualifications of individuals with any type of felony and over six thousand disqualifications based on misdemeanors. Some nineteen thousand exclusions are permanent, and over eleven thousand are mandatory, which deny any discretion for regulatory agencies to grant a license based on mitigating circumstances or rehabilitation (National Council of State Governments, Justice Center 2016). Even where no flat ban exists, many courts and licensing agencies have concluded that anyone who ever committed a crime necessarily lacks the requisite good character for a license (May Reference May1995, 193, 197). In half the states, licenses can be denied due to any kind of criminal conviction, regardless of whether it is relevant to the employment, or how long ago it occurred (Department of the Treasury, Office of Economic Policy et al. 2015, 5).

These license denials are a significant contributor to unemployment, racial inequality, and recidivism (Silva Reference Silva2012, 502; Mullings Reference Mullings2014, 267). Estimates suggest that about twenty million Americans, almost 9 percent of the adult population, and a third of the African American male population, have felony convictions (Shannon et al. Reference Shannon, Uggen, Thompson, Schnittker and Massoglia2011, 11–12). Millions more have misdemeanor records, all of which contribute to joblessness (Appelbaum Reference Appelbaum2015). Nearly seven-hundred thousand Americans leave prison each year, and half will return within three years without having found legal employment (Obbie Reference Obbie2016, B6). In one survey, two-thirds of former inmates were unemployed or underemployed five years after release from prison (Vega Reference Vega2012). The costs are particularly great for racial minorities, who are not only more likely to have convictions, but also to suffer greater adverse employment than whites with similar criminal records (Rodriguez and Avery Reference Rodriguez and Avery2016, 8).

Several thousand licensing restrictions bar individuals convicted only of crimes involving “moral turpitude,” a term subject to varying judicial definitions. One meaning is behavior that violates accepted moral standards of the community (Black's Law Dictionary 1979, 910). Another meaning is “baseness, vileness or depravity” (Petropoulos v. Department of Real Estate 2006). Which definition a court chooses can determine the outcome. For example, in a 2006 California case, an administrative law judge ruled that domestic violence did not involve “baseness, vileness or depravity,” and thus did not constitute a bar to a real estate license (Petropoulos v. Department of Real Estate 2006).

Moral turpitude also functions as a standard for discipline or license revocation in ways that bear little relation to professional practice. Physicians have been disciplined for tax evasion, shoplifting, possession of marijuana, and soliciting sex in a public restroom (In re Kindschi 1958; McLaughlin v. Board of Medical Examiners 1973; Weissbuch v. Board of Medical Examiners 1974; Windham v. Board of Medical Quality Assurance 1980; Sawicki Reference Sawicki2010, 305). Disciplinary officials have often been more concerned with minor criminal offenses, which are easy to prove, than with negligent performance, which is not, but is more relevant to patient care (Sawicki Reference Sawicki2010, 301–04). One sobering study found that criminal misconduct unrelated to patient care tended to be disciplined more severely than misconduct that had a closer connection to competent medical practice (Sawicki Reference Sawicki2010, 305).

Civil rights law imposes some constraints on employment exclusions based on a criminal conviction if they have a racially disparate impact. Such discriminatory exclusions are justifiable only if they meet a business necessity (Equal Employment Opportunity Commission 1987, 2012). As a practical matter, however, licensing boards have found such a necessity even where there is only an attenuated connection between the offense and the job sought (Thompson Reference Thompson2008, 109). California bars ex-offenders from working in real estate or physical therapy; New York prevents ex-offenders from receiving a license as an auctioneer, junk dealer, dental hygienist, veterinarian, undertaker, fire-suppression piping contractor, or bingo operator; Virginia denies licenses to felons in fields involving optometry and funeral homes (Butterfield Reference Butterfield2002, 18; Thompson Reference Thompson2008, 111).

Of course, the problem of categorical employment bans on applicants with criminal records is not just a problem in licensed occupations. James Foreman's recent book, Locking Up Our Own, offers a wrenching example of a young mother pulled over by police for a pretextual motor vehicle violation (Foreman Reference Foreman2017, 188). The officers found two small bags of marijuana in her car, which led to her arrest. Because she had no prior convictions, no charges were brought, but the arrest was enough to deny her permanent employment at a job where she had performed exceptionally well while on probation. “It's company policy,” her supervisor regretfully explained (Foreman Reference Foreman2017, 192).

Such policies, as well as occupational licensing restrictions, demand rethinking, and the inconsistencies in their application should make us question their business necessity. To be sure, our federalist structure tolerates regulatory variation in a wide array of contexts in the interests of local autonomy and experimentation. However, as commentators have noted with respect to occupational licensing requirements, if they were truly necessary to protect public health and safety, one would expect to see more consensus in state laws (Carpenter et al. Reference Carpenter, Knepper, Erickson and Ross2012). The inconsistencies in current rules point up their arbitrary underpinnings.

Mandatory and permanent bans are particularly hard to reconcile with the psychological research discussed earlier. There is no basis for assuming that one illegal act, committed many years earlier under vastly different circumstances, is a good predictor of current threats to the public. The case involving the Wisconsin grandmother is a case in point. To conclude that she lacks character to run a home daycare after she has been doing so effectively for a decade, simply because of a $294 welfare overpayment thirty years earlier, defies basic fairness and common sense (Rodriguez and Avery Reference Rodriguez and Avery2016, 6). Rather, as Section IV argues, we need fundamental regulatory reform that eliminates character inquiries altogether for many occupations and tailors the remaining requirements to criteria directly relevant to performance.

Historical Background

To provide a better sense of how character requirements function in practice, an in-depth exploration of bar admission and disciplinary practices is useful (Rhode Reference Rhode1985, 499–503). As noted earlier, the US legal profession has long been subject to character requirements. However, for its first two centuries, these mandates had little practical significance. Although a few states categorically excluded individuals convicted of certain crimes, the mobility of applicants and the absence of centralized records made character standards difficult to enforce (Chroust Reference Chroust1965, 261). What kept the system workable was that lawyers in the eighteenth and nineteenth centuries generally practiced within small professional communities where reputation was a matter of common knowledge and unethical practitioners were shunned by clients and colleagues.

As the profession grew in size and diversity, this informal approach appeared increasingly inadequate. Between 1880 and 1930, the vast majority of states strengthened character review by centralizing authority in boards of bar examiners and requiring interviews, lengthy questionnaires, and related measures (Gilb Reference Gilb1966, 63). Much of the impetus for more stringent review rested on nativist, ethnic, and anti-Semitic prejudices, as well as anticompetitive concerns during the Depression (Auerbach Reference Auerbach1976).

Pennsylvania had perhaps the most rigorous screening system. Prospective candidates faced a character investigation both at the beginning of law school and when seeking admission to the state bar. The initial interview offered an opportunity to dissuade the “unworthy” from pursuing a legal career (Bar Examiner 1932, 63). The definition of “unworthy” was quite elastic. Those rejected by one county board in 1929 included individuals deemed “dull,” colorless,” “subnormal,” “unprepossessing,” “shifty,” “smooth,” “keen,” “shrewd,” “arrogant,” “conceited,” “surly,” and “slovenly” (Douglas Reference Douglas1929, 703–05). Examiners believed that they could tell from the interviews that candidates lacked “moral … stamina,” and a “proper sense of right and wrong” (Douglas Reference Douglas1929, 703–05). The extent to which comparable biases affected decisions in other states remains unclear. However, the scant evidence available leaves no doubt that concerns with competition during the Depression influenced some bar committees. Examiners frequently argued that “with an overcrowded bar [and] an abundance of candidates who have unquestioned character,” any doubts should be resolved against admission (Bar Examiner 1932, 100, 109). Among those raising doubts were radicals, religious “fanatics,” fornicators, and adulterers (Douglas Reference Douglas1929, 703; Bar Examiner 1932, 65).

Character certification after the 1930s grew more rigorous in form but not in effect. Review became increasingly systematic, but the number of individuals denied admission was minimal, typically fewer than 1 percent (Shafroth Reference Shafroth1949, 194, 198; Brown and Fassett Reference Brown and Fassett1953, 497). Of course, those numbers do not capture the deterrent function of the screening process: a much larger percentage of potential applicants may have been discouraged from even applying to law school or may have been rejected by school admission processes.

Moral character concerns also figured in bar disciplinary procedures. Courts have traditionally asserted power to exclude lawyers for conduct involving moral turpitude, and application of that standard has been ambiguous, inconsistent, and idiosyncratic. A leading California Supreme Court decision unhelpfully concluded that “to hold that an act of a practitioner constitutes moral turpitude is to characterize him as unsuitable to practice law” (In re Higbie 1972, 101–02).

What exactly made someone suitable has been subject to longstanding dispute. During the early part of the twentieth century, habitual drunkards, home brewers, fornicators, and radical political activists fared differently in different courts (Rhode Reference Rhode1985, 552–53). To a 1929 Missouri court, seduction by an unfulfilled promise to marry constituted “baseness and depravity” mandating disbarment (In re Wallace 1929). By contrast, in the preceding year, New Jersey justices found fornication with a fifteen-year-old to warrant only a six-month suspension from the bar, in light of the victim's previously dissolute life and the attorney's reputation as an “upright and moral man” (In re Isserman 1928). In the 1970s, although a Florida lawyer lost his license following a conviction for indecent exposure in a public lavatory, an Indiana practitioner received only a year's suspension for making sexual advances to clients and offering to exchange legal services for nude photographs of one of those clients and her daughter (Florida Bar v. Kay 1970, 379; In re Wood 1976, 133). Inexplicably, it was the latter attorney whose activities were deemed “personal and unrelated” to professional practice.

Throughout the twentieth century, the moral character requirement placed a price on nonconformist political commitments. Conscientious objectors, religious “fanatics,” and student radicals were exhaustively investigated and occasionally denied admission (Application of Cassidy 1944; In re Summers 1945; Application of J. Norman Stone 1955, 771; In re Anastaplo 1961). The greatest concern surfaced during the Cold War, when bar character committees routinely grilled candidates about radical views, including whether they had read Das Kapital or were affiliated with “pinkish” organizations (Brown and Fassett Reference Brown and Fassett1953, 494). Refusal on principle to answer such questions was itself a ground for exclusion (Konigsberg v. State Bar of California 1957; In re Anastaplo 1961). In the mid 1980s, four-fifths of state bars reported that they would or might investigate conduct such as sit-ins resulting in misdemeanor convictions or membership in radical political organizations, although prevailing Supreme Court rulings made clear that such activities could not legitimately have warranted exclusion (Rhode Reference Rhode1985, 568). Such inquiries had a chilling effect on free speech. In one survey, a third of law students reported refraining from activities such as attending political rallies and signing petitions because of impending character reviews (Papke Reference Papke1973, 18–19). In addition to its deterrent effect on constitutionally protected activities, the character oversight process sent a disturbing message about the kind of conformity the bar valued (Kalvin and Steffen Reference Kalvin and Roscoe1961, 178).

My mid-1980s empirical study of bar admission and disciplinary processes demonstrated what the Supreme Court once candidly acknowledged: character requirements are “unusually ambiguous … . [A]ny definition will necessarily reflect the attitudes, experiences, and prejudices of the definer” (Konigsberg v. State Bar of California 1957, 263; Rhode Reference Rhode1985). Perhaps for that reason, the Court has largely avoided specifying what constitutes good moral character. Rather, it has rested with the general observation that any character criteria must have a “rational connection with the applicant's fitness or capacity to practice law” (Schware v. Board of Bar Examiners of New Mexico 1957, 239). The difficulty is that examiners have had inconsistent and idiosyncratic views about what constitutes such a connection. Violation of a fishing license ten years earlier was sufficient to cause one local Michigan committee to deny admission, but in the same state, at about the same time, other examiners on the central board admitted applicants convicted of child molesting and conspiring to bomb a public building (Rhode Reference Rhode1985, 538). Decisions concerning drug offenses were particularly inconsistent. Convictions for marijuana possession were taken seriously in some jurisdictions and overlooked in others; much depended on whether the examiners had, as one put it, grown more “mellow” toward “kids smoking pot” (Rhode Reference Rhode1985, 538).

Other issues provoked comparable disagreement. Examiners differed on how to treat evidence of financial mismanagement, such as bounced checks, bankruptcies, noncompliance with child support obligations, and parking tickets (Rhode Reference Rhode1985, 541). Responses to political activity were equally idiosyncratic. The Secretary of the Arkansas Board of Law Examiners looked at “political dissent with a blink,” while in other jurisdictions, misdemeanor arrests arising out of protest activities would trigger serious concern (Rhode Reference Rhode1985, 543). According to the former Executive Secretary of Manhattan's Character Committee, “sit-ins aren't politics. They involve interfering with the government and the breaking of the law” (Rhode Reference Rhode1985, 543). By contrast, to the California Supreme Court, acts of civil disobedience may reflect the “highest moral courage” (Hallinan v. Committee of Bar Examiners 1966, 87).

States divided evenly over whether they would investigate sexual conduct, sexual orientation, or “lifestyle” (Rhode Reference Rhode1985, 532). Some felt that the activity was not “within their purview” unless it resulted in criminal charges; others would inquire whether it reflected a “contumacious attitude toward the law” (Rhode Reference Rhode1985, 539). Applicants occasionally had to answer questions about the intimate details of their sex lives and living arrangements (Rhode Reference Rhode1985, 579). One official expressed relief that issues such as cohabitation rarely arose: “Thank God we don't have much of that [in Missouri]” (Rhode Reference Rhode1985, 539).

Current Administration of Moral Character Requirements

Contemporary courts and bar examiners continue to divide over the conduct that constitutes grounds for exclusion from the legal profession. Five states prohibit all felons from practicing law; two other state supreme courts have ruled that particular applicants should be permanently excluded (Pobjecky Reference Pobjecky2007, 6, 10). The remaining states consider a wide range of factors that yield inconsistent and idiosyncratic judgments (Rhode et al. Reference Rhode, Luban, Cummings and Engstrom2016, 935–45). Cases involving bankruptcy, marijuana, repeated traffic offenses, driving without automobile insurance, and compulsive gambling all have generated conflicting precedents (Clemens Reference Clemens2007, 277, 285, 287).

Opinions continue to differ over the significance of rehabilitation. For example, in 2014, the California Supreme Court declined to admit Steven Glass, a former reporter who had plagiarized stories in the late 1990s while working for the New Republic. The hearing judge and state bar court had recommended admission, based on evidence including “his employment history, community service, character witnesses, progress in therapy, remorse, and acceptance of responsibility,” all of which provided a “more accurate picture of his moral character than his misconduct of many years ago” (In re Glass 2014, 519). The California Supreme Court, however, was unimpressed, and even ignored the consensus of mental health experts. It reasoned:

By contrast, in another 2014 case, the Washington Supreme Court allowed Shon Hopwood, a man convicted of five bank robberies in 1997 and 1998, to take the bar exam, and, if he passed, to be admitted to practice. While in prison, Hopwood had drafted a pro se petition for a fellow inmate that led to a 9–0 victory in the Supreme Court. Hopwood's exemplary record since prison led to his clerkship on the District of Columbia Circuit Court of Appeals and had caused the judge who sentenced him to twelve years in prison to observe, “Hopwood proves that my sentencing instincts suck” (Mauro Reference Mauro2014).

Another area of inconsistency involves undocumented immigrants. Most states do not require applicants to the bar to disclose their immigration or citizenship status (National Conference of Bar Examiners 2015). However, some states do, and the eligibility of undocumented immigrants has been a matter of dispute. The US Justice Department has opposed admission, relying on a federal statute that prohibits state agencies from granting public benefits including professional licenses to undocumented individuals. An exception to that law exists where state statutes permit such licensure (PRWORA 1996). In 2014, the California Supreme Court relied on such a statute in determining that Sergio Garcia satisfied the good moral character standard although he was in the country without lawful immigration status (In re Garcia 2014). Garcia's personal story is compelling. His parents, Mexican farmworkers, had brought him to the United States when he was seventeen months old. His father and most of his siblings are citizens, but Garcia had been waiting eighteen years, since age eighteen, for a lawful permanent resident visa. Despite the California Supreme Court ruling admitting him to the bar, federal law on undocumented immigrants prohibited employers from hiring Garcia because he lacked work authorization as an undocumented immigrant. Until he finally obtained a green card, he ran a solo practice, handling accident disputes and immigration matters for low- and moderate-income individuals (Solomon Reference Solomon2015, A1; Maclachian Reference Maclachian2017, A1). Half his cases were pro bono, and when other clients could not afford his fees, they sometimes paid him fruit and vegetables (Solomon Reference Solomon2015, A1).

By contrast, Florida has denied admission to undocumented immigrants (Florida Board of Bar Examiners 2014). In response, three former ABA presidents filed a brief with the Florida Supreme Court claiming, without success, that exclusion of undocumented immigrants is a “waste of exceptional talent for our profession” (Gomez Reference Gomez2012, A3). Lawyers such as Garcia suggest that they are right.

Not only is the bar admission process inconsistent and idiosyncratic; it also comes both too early and too late. Screening takes place before most applicants have faced situational pressures comparable to those in practice, yet after applicants have made such a significant investment in legal training that examiners are reluctant to deny admission. The bar refuses to certify fewer than 1 percent of applicants (Rhode Reference Rhode1985, 516; Levin, Zozula, and Siegelman Reference Levin, Zozula and Siegelman2015, 54).

That is not to imply that the public would be better served if more candidates were excluded. Even trained psychiatrists and psychologists have been notably unsuccessful in predicting future dishonesty or other misconduct on the basis of prior acts (Rhode Reference Rhode1985, 559; Kaye Reference Kaye1997, 272–74; Yang, Wong, and Coid Reference Yang, Wong and Coid2010). Untrained bar examiners and judges can hardly do better, particularly given the highly limited information available. Decision makers are frequently drawing inferences based on one or two prior acts committed under vastly different circumstances. Yet the research reviewed in Section I makes clear the problems with such assessments.

The inadequacies of the current character screening process emerged in a recent study that reviewed some thirteen hundred Connecticut bar applicants and their subsequent disciplinary records to determine whether anything at the admission stage predicted subsequent misconduct. The study found that although some factors increased the likelihood of discipline, even those factors were still poor predictors. As the authors explain, “even if some variable (e.g. having defaulted on a student loan) doubles the likelihood of subsequent disciplinary action—a very strong effect—the probability of subsequent discipline for an applicant with a student loan default is still only 5 percent” (Levin, Zozula, and Siegelman Reference Levin, Zozula and Siegelman2015, 52). Taking all the risk factors together, the authors pointed out that “only two individuals were predicted to have a better than even chance of being disciplined, a finding that casts serious doubt on the usefulness of the character and fitness inquiry as a predictor of lawyer misconduct” (Levin, Zozula, and Siegelman Reference Levin, Zozula and Siegelman2015, 79).

A further problem in the bar character process involves the double standard of admission and discipline. From the standpoint of protecting the public, the misconduct of someone already practicing law is more predictive of future problems than the acts of someone not yet admitted. Yet the bar's administration of admission and disciplinary processes has operated on precisely the opposite assumption. Much of the conduct that triggers character investigation of applicants, such as financial mismanagement, psychiatric treatment, minor drug offenses, and political activity, almost never results in disciplinary investigations of practicing attorneys (Rhode Reference Rhode1985, 549).

Moreover, the same inconsistencies and idiosyncrasies that plague the interpretation of good moral character in the admission process reoccur in disciplinary proceedings. Every state has some version of the ABA Model Rules of Professional Conduct, which authorize discipline for a criminal act that reflects adversely on the lawyer's honesty, trustworthiness, or fitness (Model Rules 2015). ABA standards identify eleven aggravating circumstances and sixteen mitigating circumstances that can be relevant in determining sanctions, which permits wildly varying responses to similar offenses (ABA 2005, 26, 28).

A case history of arbitrary interpretations of moral character involves Laura Beth Lamb. She lost her license after taking the bar exam for her abusive husband (In re Lamb 1989). At the time of the exam, she was seven months pregnant. Her husband, who had previously failed two exams, had bouts of rage and depression during which he threw heavy objects and threatened to kill Lamb and her unborn child if she did not take the test in his place. She reluctantly agreed. After an anonymous tip alerted the state bar, she pleaded guilty to felony impersonation and deception. She received a fine, probation, and a requirement of two hundred hours of community service. On losing her job at the Securities and Exchange Commission, she took a position as a legal secretary. She also divorced her husband and received psychological treatment. Despite her therapist's conclusion that Lamb's “prognosis for the future is good provided she remains in therapy,” and that she was unlikely to “do anything remotely like this again,” the California Supreme Court imposed disbarment (In re Lamb 1989, 769). In the court's view, the “legal, ethical, and moral pressures of daily practice come in many forms … [and] may include the sincere but misguided desire to please a persuasive or overbearing client” (In re Lamb 1989, 769). Yet for the court to equate the pressure of an insistent client to that of an abusive, mentally unstable spouse suggests a profound insensitivity to the risks of battering for a pregnant woman (Bacchus, Mezery, and Bewley Reference Bacchus, Mezery and Bewley2004).

In most published disciplinary decisions involving conduct outside a lawyer-client relationship, courts do not even bother to consider the likelihood of its replication in a professional context. It is enough that the conduct threatens the reputation of the profession. A representative example involved Albert Boudreau, a Louisiana lawyer convicted of importing several magazines and a video displaying child pornography (In re Boudreau 2002, 76). Boudreau purchased the items in the Netherlands, where the magazines were lawful and the models were of legal age to be photographed nude. They were underage by US definitions, however. The Louisiana Supreme Court agreed with the disciplinary board that the actions constituted a “stain upon the legal profession,” and clearly reflected on the lawyer's “moral fitness to practice law” (In re Boudreau 2002, 76). Despite the absence of any prior disciplinary record, or any relationship between personal and professional conduct, the court ordered disbarment.

If the goal of such sanctions is to ensure public confidence, a better strategy would be to make the oversight process more responsive to professional misconduct, and less idiosyncratic in its responses to non-professional offenses. It can scarcely enhance respect for bar discipline when lawyers guilty of such offenses receive wildly different treatment, and the focus is professional reputation rather than public protection. Sanctions for drug offenses, tax evasion, and domestic violence now range from reprimand to disbarment, and bear too little relationship to whether misconduct is likely to occur in a professional context (Camarena Reference Camarena2001, 173; Pinaire, Heumann, and Lerman Reference Pinaire, Heumann and Lerman2006, 319; In re Lewis 2007, 730; Florida Bar v. Liberman 2010, 37; “Tax Evasion …” 2010; State of Okla. ex rel. Oklahoma Bar Association v. Smith 2011, 1095).

Historical Background

Good character as a prerequisite for citizenship appeared in the country's first naturalization statute in 1790. The first Congress extensively debated requirements for citizenship and Georgia representative James Jackson introduced the idea that prospective citizens should be required to produce testimonials of proper and decent behavior. His hope was that the title of “citizen of America” would become as “highly venerated and respected as was that of a citizen of old Rome” (Rule of Naturalization 1790). The 1790 Act in its initial form required “good character,” and five years later the term “moral” was added after supporters of the amendment assured their colleagues that it had no reference to religious opinions (Naturalization Bill 1794).

The absence of standards for this requirement led to inconsistent judicial opinions over how good “good character” should be (Persichetti Reference Persichetti1948, 185). In 1878, in the first case to define character, the court suggested that “the average man of the country” was probably as high a standard as could be set (In re Spenser 1878). Judges divided, however, over whether the test was the average man's moral convictions or his actual conduct (United States ex rel. Iorio v. Day 1929, 921; Petitions of Naturalization of F-G-and E-E-G 1956). Behavior that could be relevant in assessing the applicant's “goodness” included adultery and homosexuality (Petitions of Naturalization of F-G-and E-E-G 1956; Petition for Naturalization of O—N— 1964; In re Schmidt 1968). In one adultery case, the court acknowledged evidence from the Kinsey Report suggesting that marital infidelity was widespread, but reasoned that the test is not “what a community does, but rather what the community feels” (Petitions of Naturalization of F-G-and E-E-G 1956, 785). When adultery was “technical,” for example, because the applicant had obtained an invalid Mexican divorce, some, but not all, judges excused the conduct (Petition of R… 1944; Petition of Naturalization of da Silva 1956; Petition for Naturalization of O—N— 1964; In re Briedis 1965).

Cases involving traffic and alcohol-related offenses sparked similar disputes (In re Capozzi 1936; Petition of Donath 1956). During Prohibition, one court denied the petition of an alien who made liquor for his own use. In so ruling, the court conceded that the crime was only a misdemeanor, and that even the “best society” did not frown on home distillers (In re Nagy 1924). Nonetheless, the alien should be excluded for insufficient “attachment to the principles of the Constitution,” including the Eighteenth Amendment, which banned liquor (In re Nagy 1924). In another case that seemed to depart from the average citizen standard, the court rejected the petition of a restaurant owner who resided in a community where liquor could only be sold by the bottle, not by the drink. Although the owner regularly violated the law, so did all his competitors. The court reasoned that although the law was not enforced and all the restaurants in town were in violation, this did “not relieve [the] petitioner of his moral obligation to obey the law” (Petition for Naturalization of Orphanidis 1959). By contrast, in some (but not all) cases involving Sunday closing statutes, courts reasoned that widespread and unprosecuted violations meant that the petitioner's noncompliance was consistent with standards of the average citizen (In re Hopp 1910; United States v. Gerstein 1918).

In an effort to achieve greater clarity and uniformity, Congress in 1952 added a nonexhaustive list of statutory bars to establishing good moral character under the Immigration and Nationality Act. These included gambling, adultery, and being a “habitual drunkard” (Immigration and Nationality Act 1952). The Act did not, however, resolve all ambiguities. For example, the good moral character requirement applied to the five-year period prior to the application of naturalization. However, decision makers could look beyond the five years if there was a relationship between prior acts and current concerns, and courts divided over how close that relationship must be (8 U.S.C. § 1427(a), (e) 2005; 8 C.F.R. § 316.10(a)(2) 2013). Similar disputes surfaced over whether to make an exception for “technical adultery” (Strange Reference Strange1975, 367). In one such case, the court forgave a petitioner who had sexual relations with a woman whom he mistakenly believed to be widowed (In re Petition of Naturalization of Johnson 1968). In the court's view, these were just “two middle aged people contemplating marriage [who] bedded together from time to time” (In re Petition of Naturalization of Johnson 1968). To avoid such difficulties, Congress eventually dropped adultery as an absolute bar to findings of good moral character (Fragomen Reference Fragomen1992, 207).

In trying to make sense of competing case law, one federal judge concluded that “notions of ‘character’ and ‘morality’ are, to say it briefly, diverse. They are compounded of complex, rarely articulated, and subjective premises” (In re Petition for Naturalization of Russo 1966). Another commentator agreed, with the “sorrowful admission that not many definite conclusions can be reached as to what good moral character is under the Nationality Act” (Persichetti Reference Persichetti1948, 193).

The irrationalities of character standards attracted widespread attention during the 1970s when the government sought to deport John Lennon. After the Beatles broke up, Lennon came to the United States on a temporary visa. Although he had once pleaded guilty to possession of a half an ounce of hashish in Great Britain, US Immigration officials temporarily waived that potential ground for exclusion (Hing Reference Hing2013, 33–36). However, after Lennon began performing at rallies to protest the Viet Nam War, President Nixon ordered his removal. The proceedings sparked a flood of letters from eminent musicians, writers, entertainers, and even the Mayor of New York, attesting to the value that his continued presence would bring to the country's cultural heritage (Hing Reference Hing2013, 35 n266). Lennon prevailed, but not because of these testimonials; he was saved by a legal technicality. The British statute that he had violated did not require that his possession of the substance be “knowing,” as mandated by US law (Lennon v. INS 1975, 189–90). To many observers, the fact that Lennon could have been excluded but for this technicality underscored the problems with character standards.

Contemporary Immigration Law

Contemporary immigration law includes a specific statutory or regulatory “good moral character requirement” for certain immigration benefits or relief, and also employs the concept in most other immigration contexts (8 U.S.C. §§ 1255, 1375c 2008; Marks and Slavin Reference Marks and Noonan Slavin2012, 109). As a manual by the Immigration Resources Center notes, this requirement is “increasingly complicated” and “very confusing,” in part because there is no statutory definition (Immigrant Legal Resource Center 2014, §§ 6.2, 6.3). The statute does, however specify certain offenses, including “aggravated felonies” and almost all drug crimes, that will bar a finding of good moral character for purposes of naturalization (Immigrant Legal Resource Center 2014, § 6.2).

Related provisions of immigration law passed in the late 1980s and 1990s also make noncitizens convicted of “aggravated felonies” and other specified offenses subject to deportation (Pub. L. No. 101–649 § 501(a) 1990; Pub. L. No. 100–690, 102 Stat. 4181, 8 U.S.C. § 1101(a) 2014). Although these provisions are doctrinally separate from the moral character requirement, they rest on the same justifications, and are subject to the same objections. They apply regardless of when the felony occurred, and even where there is overwhelming evidence of rehabilitation (Lapp Reference Lapp2012, 1591).

The term “aggravated felony” is misleading because Congress has expanded the list of deportable offenses to encompass crimes that are neither aggravated nor felonies. As currently interpreted, these crimes have included public urination, college drug offenses, and shoplifting of a $10 videogame, baby clothes worth $15, and eye drops and deodorant (Lipton Reference Lipton1999; Rich Reference Rich2000, A1; United States v. Pacheco 2000, 149; Liem Reference Liem2007; Rosenbloom Reference Rosenbloom2007, A13; Menses Reference Menses2012, 850). An alien convicted of turnstile jumping was placed in removal proceedings and detained for three years before the government dropped the case (Menses Reference Menses2012, 775). A host of other misconduct also can justify deportation, including document fraud, drug offenses, and becoming a public charge (Immigration and Nationality Act 1952 § 237(a)). With deportation comes a ban on return to the United States for a specified period, sometimes as much as twenty years, depending on the ground for exclusion (8 U.S.C. § 1182(a)(9)(A)(i) 2013).

The injustices of this standard are clear in cases such as that involving Jose Velasquez, who led what the court termed an “exemplary life” except for one incident in 1980 (Velasquez v. Reno 1999). Then a friend asked if he sold cocaine and Velasquez said no, but suggested that another person might. Although no evidence indicated that he intended to benefit financially, he pled guilty to a charge of conspiracy to sell a controlled substance and was sentenced to probation. Eighteen years later he was placed in deportation proceedings (Velasquez v. Reno 1999; Futterman Reference Futterman1999, 23).

This expansive definition of deportable offenses is not justified by the common assumption that immigrants are driving up the crime rate; aliens, including those who are undocumented, have lower crime and incarceration rates than native-born citizens (Rumbaut and Ewing Reference Rumbaut and Ewing2007; Pérez-Peña Reference Pérez-Peña2017). Moreover, deportation carries life-shattering consequences, and not just for the individual deported. One estimate suggested that some eleven hundred families are separated every day and that in one ten-year period, one-hundred thousand children lost a parent (Knight Reference Knight2010; Menses Reference Menses2012, 847). In 2015 alone, Immigration and Customs Enforcement (ICE) removed over thirty-one thousand aliens who had at least one US-born child (ICE 2016a, 2016b).

Crimes of moral turpitude also prevent naturalization and constitute grounds for deportation if the crime was committed within five years of admission to the United States (8 U.S.C. § 1182; Immigration and Nationality Act 2006; Menses Reference Menses2012, 799). Again, this category has proven expansive, and has included petty larceny, lewdness, knowingly issuing a bad check, and contributing to the delinquency of a minor (Marks and Slavin Reference Marks and Noonan Slavin2012, 102–03; Menses Reference Menses2012, 800). In commenting on the idiosyncratic interpretations of moral turpitude, Supreme Court Justice Jackson concluded “that there appears to be universal recognition that we have here an undefined and undefinable standard” (Jordan v. de George 1951, 235).

For purposes of naturalization, the good moral character requirement excludes a broad range of individuals, including those who evade child support obligations, violate marijuana laws, or have extramarital affairs that tend to destroy an existing marriage (8 C.F.R. §–316.10(b)(3) 2013; Faddah v. INS 1977, 496; Grunbaum v. District Director 2012, *5–6; Lapp Reference Lapp2012, 1610). Some individuals have been penalized for seeking naturalization because the process reveals a prior offense that triggers deportation hearings. In one celebrated case, Qing Hong Wu applied for naturalization and revealed that he had been convicted of muggings at the age of fifteen (Lapp Reference Lapp2012, 1601). His subsequent behavior had been exemplary and a dozen years after his conviction he had worked his way up from a data entry clerk to vice president for Internet technology at a national company. Immigration officials not only denied his application for naturalization, but also placed him in detention and began removal proceedings. Only after a public outcry and a pardon from New York's governor did officials reverse their decision and grant his citizenship application (Bernstein Reference Bernstein2010, A20; Lapp Reference Lapp2012, 1601).

Wu is not an isolated case, and legal aid attorneys have reported that some clients are deterred from applying for naturalization due to concerns that old offenses could trigger deportation proceedings (Bernstein-Baker Reference Bernstein-Baker2007, 376; Lapp Reference Lapp2012, 1613). For example, a Dominican Republic émigré who has been a lawful permanent resident for forty years has been afraid to file for naturalization because of two criminal offenses more than a quarter-century old: possession of a weapon in the fourth degree and disorderly conduct (Lapp Reference Lapp2012, 1613).

The injustices of using minor criminal offenses as a proxy for moral character were underscored by a 2016 New Yorker profile of Noemi Romero who was brought to the United States by undocumented parents when she was three (Saunders Reference Saunders2016, 57). Although she wanted to apply for legal status through the Deferred Action for Childhood Arrivals (DACA) program, she could not afford the $465 application fee. To obtain a job at a local Vietnamese grocery store, she borrowed her mother's Social Security card. She landed the job, but several months later, the store was raided. She was charged with aggravated identity theft and forgery. While she was held in jail for two months, her lawyer arranged a plea bargain to reduced charges. She accepted, not realizing that her felony conviction would make her permanently ineligible for DACA. Although she was not immediately deported, her life was in limbo: she “can't work and can't go to college, although she has lived virtually her whole life in the US, and has no reason to go back to Mexico and nowhere to live if she's sent there” (Saunders Reference Saunders2016, 57). For undocumented individuals such as Romero, Trump administration policies are likely to make a bad situation worse.

Immigration and Character Under the Trump Administration

The moral character requirement has taken on a new urgency under the Trump administration. In the 2015 announcement of his presidential candidacy, Trump characterized many Mexican immigrants as rapists, and announced that he would build a wall between the United States and Mexico (Time Staff 2015). Later that year, he called for a complete ban on Muslims entering the country (Trump Pence 2015).

President Trump has attempted to follow through on both promises, and his actions reflect inaccurate assumptions regarding the character and productivity of immigrants. Less than a week after assuming office, Trump signed an executive order to hire five thousand additional Border Patrol agents and construct a wall along the US-Mexico border. “We are going to get the bad ones out,” he told the press (Diamond Reference Diamond2017). The order claims that “illegal immigration” has placed a significant strain on “the local communities into which many of the aliens are placed” (Trump Reference Trump2017). This claim ignores the $11 billion that these individuals pay in state and local taxes each year (Gee, Gardner, and Wiehe Reference Gee, Gardner and Wiehe2017).

Trump's 2017 executive orders preventing citizens of seven Muslim-majority nations from entering the United States rests on equally problematic assumptions about the link between religion, nationality, and moral character (Kessler Reference Kessler2017; Trump Reference Trump2017). The national security concerns that Trump invoked to justify the ban ignore the fact that none of the recent mass shootings or terrorist attacks in the United States were perpetrated by individuals from the nations covered by the executive order (Thrush Reference Thrush2017). In its original form, the executive order gave priorities for refugee claims to religious minorities, such as Christians, in predominantly Muslim countries, on the apparent theory that only Muslims have terrorist tendencies (Thrush Reference Thrush2017). After a federal court found that the ban, after a full trial, would likely be found to constitute unconstitutional discrimination, the administration issued a revised version eliminating the religious exemption (Lawfare 2017; Trump Reference Trump2017). Its premise is that the risk of even one person from these countries entering the United States is “unacceptably high” (Trump Reference Trump2017). In July 2017, the Supreme Court allowed relatives of American residents to enter the United States but temporarily upheld the ban's broad restrictions against other refugees, leaving the travel ban largely in force pending oral argument in October (Liptak 2017).

In another 2017 executive order, Trump also dramatically expanded the groups of undocumented immigrants who would be a priority for deportation. Those groups included anyone with a criminal conviction or criminal charges pending, and anyone who, “in the judgment of an immigration officer … pose[s] a risk to public safety or national security” (Trump Reference Trump2017). Under this order, not only would individuals guilty of petty offenses be subject to removal, but so too would individuals who were charged but never convicted of any crime, or who just seemed threatening to immigration authorities. Once again, the assumption underlying this order appeared to be that any conviction or arrest was sufficient evidence of bad character to justify deportation. This assumption adds to a climate of fear within immigrant communities, particularly given the Trump administration's stepped-up arrest rate for immigration offenses. During the first one-hundred days of the administration, that rate was up by over a third compared with the same period last year (Rhodan Reference Rhodan2017, 4).

The fear resulting from these measures carries enormous collateral costs. One is the reluctance of victims of crimes to report them to the police due to concerns over deportation. Survivors of domestic violence and sexual assault have been left particularly vulnerable. Since Trump took office, police departments and nonprofit organizations have seen a dramatic decline in immigrant women reporting sexual violence and seeking protective orders (Medina Reference Medina2017, 1, 21). Although victims of abuse are eligible for special visas if they cooperate with prosecutors, many women are reluctant to take the chance (Medina Reference Medina2017, 21; Pashman Reference Pashman2017). And their concerns are not unfounded. Courthouse arrests have become increasingly common, such as one involving an El Paso woman taken into custody by immigration agents moments after she had received a domestic violence protective order (Medina Reference Medina2017, 21).

Another toxic effect of Trump policies is the stress it creates for the six million US children who are themselves citizens but who are at risk of having an undocumented parent deported (Schochet Reference Schochet2017, 2). Persistent levels of such high stress can permanently impair children's cognitive and emotional development (National Scientific Council on the Developing Child 2010, 9; Schochet Reference Schochet2017, 11). Of course, the fear arising from Trump's immigration policies is not limited to those with criminal records or other moral character difficulties, but the stepped-up deportation of such immigrants underscores the importance of rethinking character requirements.

Occupational Licensing

One way of addressing problems in moral character requirements for occupational licenses is to reduce the number of occupations subject to licensing. This strategy would have other benefits. As a recent Brookings Institute report concluded, “the literature provides little evidence that stricter licensing regimes lead to improved quality of services” (Kleiner Reference Kleiner2015, 3). What the research does unequivocally show is that such regimes lead to higher prices for consumers, unwarranted burdens for licensees seeking to move across state lines, and unjustified exclusions of those with criminal records. Society as a whole would benefit if states conducted cost-benefit analyses of the need for licensing in selected occupations and the advantages of less restrictive alternatives, such as registration or voluntary certification systems (Kleiner Reference Kleiner2015, 3). The federal government could jumpstart this review process by providing financial incentives and establishing best practices (Kleiner Reference Kleiner2015, 4–5). A few states have begun moving in this direction. For example, Arizona recently removed licensing requirements for yoga teachers, geologists, and citrus fruit packers (Cohen Reference Cohen2016, B5). More states should follow suit.

More states could also adopt the approach of jurisdictions that have abandoned the good moral character requirement altogether and replaced it with a more narrowly tailored question asking whether applicants have “engaged in conduct warranting disciplinary action against a licensee” (Minnesota Statute Annotated § 147.02 [West 1989]; Maine Revised Statute Annotated Title 32 § 3271 [West 1991]). This approach should also be coupled with a review of disciplinary sanctions to ensure that they are available only for conduct reasonably related to job performance.

Another approach would be to retain the moral character requirement but to eliminate categorical bans based on criminal records and require individualized assessments. A model statute by the National Employment Law Project suggests that such assessments should consider:

1. • The nature and gravity of the offense;

2. • The nature of the job;

3. • The length of time that has passed since the offense;

4. • The circumstances surrounding the offense, including the age of the offender and contributing social conditions; and

5. • Evidence of rehabilitation, including subsequent work history and character references (Rodriguez and Avery Reference Rodriguez and Avery2016, 31–32).

This approach builds on reform efforts in a growing number of states as well as guidance by the Equal Employment Opportunity Commission that calls on employers to limit criminal background checks to job-related convictions (Equal Employment Opportunity Commission 2012, 11, 16). At least nineteen states and several dozen cities have passed “ban the box” statutes that prevent employers from asking about job applicants' criminal records and eliminating ex-offenders from consideration before conducting any individualized assessments (Appelbaum Reference Appelbaum2015; National Employment Law Project 2017). Some states also allow ex-offenders to demonstrate rehabilitation or require licensing officials as well as employers to make individualized assessments of job applicants' conviction records rather than relying on blanket prohibitions (Reentry Policy Organization 2003, 301). Under some laws, individuals who make a showing of good moral character can receive certificates of rehabilitation, which effectively lift statutory bars.

Of the states requiring individualized assessments of employee history, New York provides the strongest protection. It prohibits employers from using criminal convictions to deny employment unless “(1) there is a direct relationship between one or more of the previous criminal offenses and the … employment sought; or (2) … The granting of the employment would involve an unreasonable risk to property or to the safety or welfare of specific individuals or the general public” (New York Correction Law 1998; New York Executive Law 1998). Unlike laws in other states, this statute also sets forth specific criteria for determining the relationship between the job and the criminal offense. These criteria include the seriousness of the offense, the relationship between the job and the crime, the time that has elapsed since the offense, the age of the applicant at the time of the offense, the rehabilitation of the applicant, and the state's interests in protecting the public and the business (New York Correction Law 1998). Accordingly, in one New York case, the court held that the city could not deny a custodial job to an individual convicted of manslaughter and drug offenses because the crimes were not related to the employment and the applicant did not present an unreasonable risk to persons or property (Soto-Lopez v. New York City Civil Service Commission 1989). States could substantially improve the administration of the good moral character requirement by following this approach for licensed occupations.

Courts could also demand that licensing authorities more closely link character inquiries to actual job responsibilities. Litigation involving private employers suggests a useful framework. For example, in 2007, an African American man challenged a transportation authority's policy of excluding anyone ever convicted of a violent crime from working as a paratransit driver for disabled passengers (El v. Southeastern Pennsylvania Transportation Authority 2007). Forty-seven years earlier, the plaintiff had been convicted of second-degree murder arising from a gang-related fight when he was fifteen years old. He claimed that the employer's categorical ban had a disparate impact on racial minorities and the EEOC agreed. The employer responded that the policy was justified by business necessity, relying on expert testimony suggesting that those convicted of violent offenses were somewhat more likely to reoffend. On a motion for summary judgment, the court accepted that justification and upheld the policy, but suggested that the result might have been different if the plaintiff had produced evidence indicating that after a certain number of years, convictions ceased to be reliable predictors (El v. Southeastern Pennsylvania Transportation Authority 2007).

A growing body of research could supply such evidence. It indicates that most detected recidivism occurs within three years of an arrest and almost always within five (Blumstein and Nakamura Reference Blumstein and Nakamura2009). One large-scale study of eighty-eight thousand offenders found that it was possible to predict the point at which the likelihood of an arrest for someone with a prior offense would decline to the same rate as for the population generally. For burglary, it took 3.8 years; for aggravated assault, it took 4.3 years (Blumstein and Nakamura Reference Blumstein and Nakamura2009, 12).

Such research should inform decision making by courts, legislatures, and licensing authorities in considering how long a prior conviction should have relevance. But as noted earlier, character determinations should also include other factors, such as the circumstances surrounding the offense, the age at the time it was committed, whether it is job related, and evidence of rehabilitation. The goal should be to be to make moral character inquiries in licensing more consistent with what we know from research on moral character in practice.

Lawyer Admission and Discipline

The same goal should guide reform of lawyer admission and disciplinary proceedings. Courts, bar examiners, and disciplinary authorities should require a direct and substantial relationship between legal practice and any potentially disqualifying conduct. In assessing that relationship, decision makers should consider factors such as the remoteness and seriousness of the conduct, its relationship to the specific duties of lawyers, the rehabilitation of the individual, and any other mitigating circumstances. Under those criteria, bar examiners should not categorically exclude undocumented immigrants. Nor should bar admission and disciplinary decision makers disregard psychological research and mental health testimony in proceedings such as those involving plagiarist Stephen Glass, and domestic violence survivor Laura Beth Lamb, where there is little likelihood for future misconduct.

In all cases, the preeminent concern should be public protection, not professional reputation. Resources now squandered in a vain attempt to predict lawyer misconduct would be better spent on detecting, deterring, and remedying it. As former Supreme Court Justice Robert H. Jackson noted in a related context, a vague standard like moral turpitude invites caprice and clichés (Jordan v. de George 1951, 239). A profession committed to due process and fundamental fairness should aspire to do better.

Immigration

In the words of John F. Kennedy: “Immigration policy should be generous; it should be fair; it should be flexible” (Kennedy Reference Kennedy2008, 50). By this standard, current policy falls far short. As a 2009 Human Rights Watch Report put it:

Although critics have argued with some force that Congress should eliminate the moral character requirement entirely, that seems politically unlikely (Lapp Reference Lapp2012, 1630–31). A more plausible alternative would involve four reforms. First, Congress should dramatically restrict the offenses that decision makers can consider in naturalization and deportation contexts. Only truly serious crimes should matter. Second, Congress should restore discretion to individual judges to consider the entire set of facts bearing on character before an individual is deported. These include the factors enumerated above, such as the nature and circumstances of the offense, the time that has elapsed since the offense, the age at which was committed, and evidence of mitigation and rehabilitation. Third, Congress should follow the example of countries such as the United Kingdom and specify a period for the relevance of an offense, after which it would no longer have adverse consequences if the applicant can demonstrate rehabilitation (UK Border Agency 2010, 15; Lapp Reference Lapp2012, 1634). Fourth, either legislation or judicial decision making should establish a clear set of norms for courts and immigration officials in making character determinations. The approach of the European Court of Human Rights in deportation contexts is instructive (Sherlock Reference Sherlock1998; Cook Reference Cook2003, 316). It specifies relevant considerations, including the seriousness of the offense, the individuals' records since the offenses were committed, the amount of time the individuals have lived in the deporting country, whether their families reside there, whether they have any ties to another country, and the likelihood that they could successfully reestablish family life in another country (Cook Reference Cook2003, 319). A standard similar to this approach is already specified for certain immigration cases, and should be more broadly applicable. Individuals whose criminal convictions now prevent them from discretionary relief should have a more holistic review of their entire records.