Chivalry and the Law: How Women Get Off
The chivalry thesis, as it relates to criminal justice, is the idea that women are more often spared punishment for their crimes than men, in part, because a predominantly male legal system favors women, a prejudice stemming from an inculcated or ingrained sense of chivalry. When the judges (or magistrates), police, prosecuting attorneys, and other individuals in our legal systems are female, it is claimed that this bias is reduced, therefore resulting in less clemency toward women in the criminal justice process. This thesis was originally presented by Otto Pollak in the 1950 book The Criminality of Women, but unlike his thesis, which was rooted primarily in biological causes, this piece will argue for the presence of a chivalry bias in criminal justice from a principally sociological perspective.
While the existence of a chivalry bias in various criminal justice systems, such as those of the United States and Britain, has long been disputed, more recent research has contended and sought to confirm it, though employing different (yet interchangeable) terminology. In 2004, Max M. Schanzenbach, from the Northwestern University School of Law, made the case for the actuality of a "paternalistic or chivalrous bias" in a paper called "Racial and Gender Disparities in Prison Sentences: The Effect of District-Level Judicial Demographics." Significant among his findings was that, where gender disparities were concerned: "The greater the proportion of female judges in a district, the lower the gender disparity for that district." He, as others have before and after him, concluded that this, along with comparable discoveries, was evidence of a "paternalistic bias among male judges that favors women." This concept and his findings, while mainly offered under an alternative designation, are the same as the arguments and evidence for the presence of a chivalry bias in criminal justice.
Additionally, self-report studies have, for many years, revealed that the disparity between the sexes in criminal behavior is not as great as the differences in arrest and conviction statistics. In British law, for instance:
"Although women who stepped far outside expected gender roles (through the use [of] violence towards children, for example) were prosecuted severely, most crimes committed by women were likely to be dealt with by less formal judicial procedures, such as informal arbitration and summary prosecution, or at the Quarter Sessions courts, and such cases do not appear in the Old Bailey records."
A similar state of affairs has existed during much of U.S. history.
But before we shift on to inspecting at length more self-report survey findings, data on arrest and conviction disparities, and a few feasible solutions, let us examine what chivalry was, what it means today, and what role it may play in the criminal justice system, as well as some broader theories surrounding the issue.
Notions of chivalry from the medieval period have continued to the present in some ways and melted away in others; initially intended as a set of "rules to live by" for Christian knights, this code of conduct contained elements of patriotism, honor, and to a lesser extent, guidance regarding interactions with women, which some aver began as observations of the importance of the Virgin Mary in the Catholic belief system. Today, the sole aspect that the majority of people, men in particular, tend to remember about chivalry is that women must be treated with reverence and respect.
The distorted view of women and men in regard to criminal behaviors dates back even to the earliest days of sociological theory and research, which women were, prior to the emergence of feminism, not largely a part of as subjects. Indeed, when women were considered, the prevailing notion was that crimes committed by females were more exclusive to the individual, stemming from a defect of their own personal biology, as opposed to something which all women were potentially capable of. Late 19th century thinkers such as criminologist and physician Cesare Lombroso and Italian historian, journalist, and novelist Guglielmo Ferrero made cases based on such reasoning. In their 1898 book The Female Offender, the pair theorized that women's biology has a tendency to prevent them from engaging in criminal behavior. Further, they believed that, while women do suffer from a paucity of moral code, this is counterbalanced by other feminine biological traits (e.g., underdeveloped intelligence, maternity, and piety).
In 1950, writer and professor of sociology Otto Pollak, the originator of the chivalry thesis (as noted above), contended that women were skilled at concealing their crimes due to biology, arguing in addition that women had grown efficient at lying to men by way of practice at hiding their discomfort and pain during periods and false interest and pleasure during sex. Author David Jessel and scientist Anne Moir, in 1997, attempted to support the biological argument as they discovered that a significant percentage of women are let off on charges of violent acts due to PMS.
Different theories, some of which antedate Otto Pollak's work, centered on the divergent roles of the sexes in relation to their children. Talcott Parsons, in a 1937 book entitled The Structure of Social Action, asserted that because children were most frequently raised by the female parent, this provided a role model for girls which was denied boys, at least in the modern industrialized world where men had to "go off" to work rather than take boys out to the field to learn farming.
More current research likewise supports this notion in a broader context. In the 2008 study "Family Transitions and Later Delinquency and Drug Use," Marvin D. Krohn, Gina Penly Hall, and Alan J. Lizotte hypothesized that family transitions such as divorce will trigger a greater tendency for drug abuse, both directly and indirectly, through effects which such transitions have on school performance, discipline, and household stability. In addition, their work focused on the greater impact that these transitions have on boys than on girls due to the fact that boys are more likely to externalize their anger at the situation, and that in broken families, the mother is more likely to retain control of the children. The belief that the mother-girl bond is stronger than the mother-boy bond was hypothesized in their research to produce this outcome.
Interestingly, however, the actual results of this study did not support the gender difference which was hypothesized, although they did, unsurprisingly, show an overall relationship between family disruption and subsequent drug abuse.
Enter the Chivalry Thesis
Though Otto Pollak was the originator of the chivalry thesis, the faults of his overly biological approach and explanations rendered the validity and effectiveness of his conclusions lacking. But by taking a more sociological approach, it is possible to evidence the idea of a chivalry bias in various criminal justice processes.
In a 1995 study called Young People and Crime, John Graham and Benjamin Bowling, of the British Home Office Research and Statistics Department, conducted a national random sample of 1,721 young people aged 14 to 25, with a booster sample of 808 respondents from ethnic minorities of the same age. In this study, participants were asked whether they had ever committed any of a range of offenses from vandalism to serious property and violent crimes. The study was extensive in scope and covered sex, age, and racial variables. In their findings, they reported that 31% of females and 55% of males admitted to at least one of 23 offenses throughout their lifetimes, and 12% of females and 28% of males admitted to at least one in the last year. This revealed that sex differences in criminal behavior were not as great as the disparities in crime statistics, and although the authors did not promote a "chivalry thesis" directly, their data could, and has been interpreted to, support such a thesis.
Then, in 2000, Claire Flood-Page, Siobhan Campbell, Victoria Harrington, and Joel Miller, also of the Home Office, performed a second sweep of the Youth Lifestyles Survey. In their report, 37% of females and 57% of males admitted to at least one of 27 offenses in their lifetimes, and 11% of females and 26% of males admitted to at least one in the last year. They likewise discovered no significant change during the six years between the first survey in 1992/93 and the second survey from 1998/99 in the percentage of women or men who admitted to committing offenses in the past year. These outcomes were congruent with the Graham and Bowling findings in suggesting some difference in the probability of being convicted between the sexes.
Furthermore, a number of studies have endeavored to measure the variance in sentences for women and men that might be due to a chivalry bias. One of the more striking instances of a marked disparity comes from a 2001 work called "Racial, Ethnic, and Gender Disparities in Sentencing: Evidence from the U.S. Federal Courts" by professor David B. Mustard, of the University of Georgia. Without controlling for variables such as criminal history and level of offense, his review found that: "The average sentence for males is 278.4 percent greater than that of females (51.5 versus 18.5 months)." Later in the text, he noted that "after controlling for the offense level, criminal history, district, and offense type, . . . females received 5.5 fewer months than males." According to Mustard, the average sentence length is 46 months, which evaluated at the mean, evinces that "males receive 12 percent longer terms than females." He also remarked that "females receive even shorter sentences relative to men than whites relative to blacks." Lastly, much like Max M. Schanzenbach, he argued that while literature on discrimination regularly declares that "females are objects of discrimination and receive worse outcomes," in sentencing: "women receive better outcomes, consistent with women's being treated paternalistically in court." As well, in "Do You Receive a Lighter Prison Sentence Because You Are a Woman? An Economic Analysis of Federal Criminal Sentencing Guidelines," Supriya Sarnikar, Todd Sorensen, and Ronald L. Oaxaca (2007) observed what they suspected was a 30% gender gap.
But in a more recent U.S. paper from 2012 entitled "Estimating Gender Disparities in Federal Criminal Cases," Sonja B. Starr, a professor of law at the University of Michigan Law School, inquired: "Do otherwise-similar men and women who are arrested for the same crimes end up with the same punishments, and if not, at what points do their fates diverge?" Importantly, she noted that this is not a simple measure of the final verdict in trials which ascend to completion. That number, as measured by previous studies, would be lower. Instead, Starr took into account intermediate decisions made along the way (to the extent that they can be traced), such as discoveries of evidence by the police and decisions to prosecute or not by prosecuting attorneys.
Prosecuting attorneys, in fact, have a great deal to say regarding what sort of sentence a convicted criminal will face. In a case involving a couple, such as a husband and wife, committing a serious crime, they might select to reduce the charges against the woman by claiming that it was the man who planned the crime, and that the woman had little choice in the matter. Sentencing guidelines were, in the past, exceptionally broad at times, beginning with no jail sentence and rising to tens of years. These ranges were later more tightly constrained, but that had the effect of prosecutors adjusting the charges which they brought against defendants, as alluded to above, to achieve the same effect:
"In 2003, Congress passed the PROTECT Act. The Act sought to eliminate virtually all departures from the Guidelines by creating a reporting mechanism for the judges who were not 'compliant.'
The result was a division of labor that gave extraordinary power to prosecutors who could effectively determine sentences, either by what they charged in the first instance or what they held in reserve for the sentencing 'real offense' determination. It also gave power to Congress, which could also determine sentencing outcomes through mandatory minimum sentences or its edicts to the Commission."
These evolving factors over the years have made precise determinations regarding the totality of gender bias in the law arduous to come by. Prior to Starr's work, several studies attempted to arrive at an estimate, and concluded values, as partially referenced already, ranging from 10% to 30% longer sentences for men than women.
Yet these studies only considered the bias as it might exist from the point of arrest until final disposition of any given case. Starr analyzed the data in terms of several points at which such a bias could be applied, and found that being male can contribute in sum to a 63% increase in sentence length on average.
(It should be mentioned here that a case decided by the U.S. Supreme Court in 2005, United States v. Booker, 543 U.S. 220 (2005), and three subsequent cases, restored a certain degree of freedom to judges when it comes to sentencing, but it is too soon to tell if this will deepen or lessen the extent of unequal sentences for women and men.)
Since Roman times, Lady Justice has been depicted wearing a blindfold and carrying a set of scales to represent the objectivity applied by the court system. The facts are "weighed," and justice is not distracted by the superficialities of emotion, hence the blindfold. Yet we know that in the real world, justice does not function so flawlessly.
In the modern era, it has been accepted that fairness begins, not in the courtroom, but in the legislatures and councils which produce the laws in the first place. While laws against murder and robbery would not typically specify the sex of the perpetrator, there were, until recently, still laws on the books that assumed that sexual aggression was always committed by a male and that the victim was always female. Alterations to both U.S. and British law have now largely eliminated personal pronouns such as "he," "she," "him," and "her," replacing the gendered terms with sex-neutral language such as "perpetrator" and "victim" in legal definitions of rape and sexual abuse. But those modifications have not yet entirely filtered into the consciousness of the public and media, and we often see people, including journalists, leap to assumptions about the sexes of victims and perpetrators even when so doing compromises the argument which they are attempting to make. Even when laws are discussed by legislators and the courts, two of the most substantial elements of criminal justice, the use of sex-neutral language has still not fully taken hold.
Unfortunately, when it comes to studying crime from a sociological perspective, these prejudices of the past can still thwart our best intentions. We lament the fact that so many crime studies rely exclusively on surveys, frequently anonymous ones, but our criminal justice systems do not perceive themselves as the servants of sociologists. The reason that sociologists must so regularly rely on surveys is that crime statistics are, in some cases, either unreliable (different states report differently), or unavailable (police and court records having been expunged for one reason or another).
Because, mathematically, the number of convictions cannot exceed the number of prosecutions, and the number of prosecutions cannot exceed the number of arrests, we witness in the public record good statistics about convictions, less perfect data regarding arrests, and less data about those who were acquitted or arrested and then released for reasons other than acquittal. Since there will always be individuals who are falsely accused, and even more who are mistakenly arrested, this seems an expectable outcome as far as data collection is concerned. It just doesn't assist us in understanding crime as well as we would desire to "in the trenches." The answers to questions such as: "How many women committed a robbery but were freed as a first offense, and how does this compare with men in the same or similar circumstances?" can, for the most part, only be conjectured about.
As imperfect as the data that we have is, we must strive to make sense of it. Concurrently, we must hope that the quality of the data improves in order that more crimes can be prevented in the first place, rather than placing the emphasis, as we do now, on after the fact detection, conviction, and punishment.
What About the Children?
One of the several facets of these biases (against men) is the presumption that women take primary responsibility for the raising of children, an easy assumption to make until a few decades ago, but far less so today. Sentencing guidelines and the laws themselves do not have an explicit manner of handling this concern regarding the welfare of children. Were that so, a great deal more transparency would be possible. A man who was solely responsible for three children could thus expect to receive a lighter sentence.
Without such unambiguous guidelines in the law, it is likely that child welfare is simply taken into account at an almost subconscious level under the presupposition that it is principally the obligation or occupation of women. This extends beyond a mere notion of chivalry alone. As it becomes more and more common for men to assume roles as caregivers, this form of bias may well start to correct itself, just as gender balance on the bench will, to a degree, correct the bias due to chivalry (or "paternalism").
How Can It Change?
The first substantial development to recognize here is that this issue is changing. FBI arrest data from 2003 to 2012 show that arrest rates have decreased for males while increasing slightly for females overall, with significant differences in some categories. We can anticipate that politicians and pundits will do everything that they can to maneuver in front of the parade and work to convince us that they are leading it, but demographic shifts which began decades ago continue to bear fruit when it comes to the pursuit of equality in the present.
Women have nearly caught up with men in law schools today, and while we would hope to maintain a fair and close balance there when one is attained, this has and will continue to gradually populate the bench - and on down through the legal system - with women who are, at least based on what the evidence indicates, more likely not to share men's greater biases of chivalry or paternalism (women who might, in fact, overcompensate against these drives or beliefs). Moreover, if the trend for more and more men to be responsible for the raising of children endures, this too could result in that pendulum swinging the other way for a while, but clearly we are not there yet.
Sex-neutral laws, more specific sentencing guidelines which take men's growing share of childcare into greater consideration, and even seemingly unrelated factors such as efforts to clear overcrowded prisons by endeavoring to reform criminals rather than incarcerate them could all have an equalizing effect when it comes to gender disparities in criminal justice.
This article has been republished with updates to (August 11, 2017):