Weigh Benefits and Costs of Predictive Policing Before It’s Too Late
Your article “Black Box Justice” [Sep|Oct 2017], about Professor Richard Berk’s algorithms that predict crime, provided good coverage of so-called “predictive policing,” especially in presenting details about how these algorithms work. This gives insight into the civil liberties trade-offs that give many people pause, as Andrew Ferguson L’00 notes in his book The Rise of Big Data Policing, cited in the article.
Not all uses of these algorithms are transparent. Commercial predictive software used by police departments includes HunchLab, from the Philadelphia-based Azavea Corporation; and PredPol, from the California-based company of the same name, probably the most widely used such package. These algorithms are generally not open to scrutiny and so add to the sense of opaque decision-making about human beings.
At this early stage, the benefits of predictive policing need to be carefully weighed against its costs before it becomes irretrievably cemented into our justice system without sufficient oversight.
Sidney Perkowitz G’62 Gr’67, Atlanta
Transparency is also a concern for Berk, who argues that criminal justice algorithms should be open-source. —Ed.
Number of “High Risk” Prisoners Raises Fairness and Due Process Questions
As a regular volunteer in New York State correctional facilities for both secular and religious programs, I read your article on the use of computer algorithms to predict crime with concern. I have seen the despair of men who have been denied parole over and over again, after having made their best efforts to change. Because of the rising suicide rate in New York State prisons, volunteers were recently required to receive training in recognizing signs of possible suicide. From my conversations with and observations of men who have been denied parole multiple times, I believe this rise is, in part, due to the unreasonable denial of parole. New York State does not use computer algorithms in its process, to my knowledge, although it has a risk assessment questionnaire not available to the public. Fairness and due process questions are raised by the large number of prisoners seen as high risk as compared to the actual risk of offending mentioned in the article.
I have some thoughts about what might predict the likelihood of a person committing another crime. Some of these are well known, and, I assume, are part of Professor Berk’s algorithms: the more educated, the less likely to recidivate; the older, the less likely; the closer the family ties, the less likely; the more positive program participation, the less likely. Knowing these characteristics means that correctional systems should be acting on this knowledge, and they are not.
Other factors may not have as much research. It is my observation that participation in religious groups is enormously helpful, and some of the chaplains I have known have provided incredible support and are role models for the prisoners under their care.
If the prison allows those who are incarcerated to take positive leadership roles inside the prison, these men and women are both less likely to return to prison and more likely to exercise leadership once they are released. They also become positive role models for others. This leadership, however, makes those running the prisons very anxious as uncomfortable issues may be raised, and the fear of gang activity is high. It seems to me that the level of suspicion that positive leadership engenders, and the fear of having hierarchy questioned, keeps facilities from accurately assessing what is happening and seeing the difference between gang leadership and positive activity, which may often diminish gang activity.
For me the use of these algorithms allows the system to abdicate responsibility for ensuring that what works is being implemented and maintaining a system that is costly in human and financial terms. Other countries do it better.
Pamela Wood CW’62, New York
Bad News, Good News
I read with great interest Ben Yagoda’s explanation of how the abominable UPenn moniker came to be [“Expert Opinion,” Sep|Oct 2017]. I always figured it came from outside sources that dubbed us UPenn based on their familiarity with UMass and UConn, and it just stuck. But to know that it was self-inflicted—Aaarrgghhh!!!
On a brighter note: I really like the new look.
Tom Lamont W’68, Ridgewood, NJ
As a longtime reader and supporter of the Gazette, I confess the changes made in the Sep|Oct 2017 issue have left me underwhelmed to say the least.
Changes to enhance readability, clarity, and functionality are always welcome; however, these changes accomplish nothing other than scream … Change for Change’s Sake!
The old adage, “If it ain’t broke, don’t fix it,” would seem applicable in this instance. A particular example is the manner in which “Alumni Notes” is nearly indistinguishable from “Obituaries.”
My recommendation is to go back to the former format.
Gerald Citron W’55, Scarborough, NY
Taking Advantage of All Our Names
Many years ago, in the late 1980s, I was thrilled to find an “unofficial” baseball cap that read UPenn Quakers. I immediately snatched up that hat, as many folks in my home region of Greater Boston do seem to identify more with the name UPenn. Why not take advantage of all the names of our University? Why not actively use Penn and UPenn simultaneously (as well as Pennsylvania and U of Penn/U of P for that matter). Any way to allow more students and alumni to identify with the school—with the name they prefer—is desirable. The name Penn is great, but challenges occur when one does a web search with the word Penn. I am not an expert in branding, but it seems to me that if the majority of certain geographical cohorts identify the school as UPenn while the majority of Philadelphians call it Penn, why not keep the option to use both names on apparel and other official items?
Ari B. Kaufman C’ 89, Ashland, MA
Attending to Social Justice Is Essential for the Welfare of Children
My colleague Richard J. Gelles [“Expert Opinion,” Jul|Aug 2017] points out that the Code of Ethics of the National Association of Social Workers (NASW) calls on social workers to “challenge social injustice.” He then reaches the conclusion that social justice blinds social workers to the sight of tortured children and prompts them to coddle abusive parents, whom they supposedly view as their true clients.
To support his claim, Gelles offers only the fact that children sometimes die horrible deaths even after they are “known to the system.” The one case he cites involved criminal wrongdoing by caseworkers—something that is hard to blame on the NASW Code of Ethics.
In fact, attention to social injustice is not the cause of these horror stories. To the contrary, attending to social justice is essential for the welfare of children. First, the real reason for missed cases of extreme child abuse is almost always because underprepared, undertrained caseworkers have overwhelming caseloads. They lack the time to investigate cases properly, so they make terrible errors in all directions. What is overwhelming them? False allegations, trivial cases, and the confusion of poverty with child “neglect.” Such cases are far more common than the horror stories Gelles cites.
Second, ignoring social injustice causes horrendous harms to children. One need only read the July 21, 2017, New York Times story on foster care as the new “Jane Crow,” documenting case after case of needless removals of children, largely because of their poverty, to see how the system destroys individual families and undermines entire communities of color. Or consider the multiple studies that have found that 30 percent of America’s foster children could be home right now if their families just had decent housing. Or the study showing that, the NASW Code of Ethics notwithstanding, caseworkers given otherwise identical hypothetical scenarios are more likely to consider a child “at risk” if the family is black. Other studies reveal it takes less risk for caseworkers to remove a black child from the home than to remove a white child.
So it’s no wonder enduring the trauma of a child maltreatment investigation has become the norm for black children. A recent study found that a majority of black children will suffer this trauma before they reach age 18. As I argued in my book Shattered Bonds: The Color of Child Welfare, this excessive and discriminatory disruption of black families constitutes a form of racial injustice.
All this harm is compounded when children are consigned needlessly to the chaos of foster care. Because most cases are nothing like the horror stories, two massive studies, involving more than 15,000 cases, found that in typical cases children left in their own homes fared better even than comparably maltreated children placed in foster care.
Philadelphia tears apart families at the second highest rate among America’s 10 largest cities—triple the rate of New York City and quadruple the rate of Chicago. Yet Philadelphia’s appalling rate of child removal did not prevent the tragedy Gelles uses to make his case. Research confirms that taking away more children does not reduce child abuse fatalities. In contrast, when Alabama was forced by a lawsuit to rebuild its entire system to emphasize family preservation, independent court-appointed monitors found that child safety improved.
More broadly, a system attentive to social justice would generously support families’ ability to care for children, reduce both parental and societal child maltreatment, and avoid the trauma to children caused by needlessly tearing them away from their homes. Social justice matters to child welfare not because it’s what parents want but because it’s what children need. For social workers to do anything less would be unethical.
Dorothy E. Roberts, faculty, Philadelphia
Mourn Lost Tools, but Undemocratic Work World Is Sadder
As a historian of science and engineering focused on the history of instruments and metrology in American mass production, I was delighted to come across “The Tools of Man,” Frank Burke’s tender account of the traditional industrial measurement devices and practices now fading into obsolescence [“Alumni Voices,” Sep|Oct 2017]. There is no doubt that deep engagement with tools and techniques in manufacturing jobs is the norm for a far smaller proportion of workers today than would have been the case in past decades.
However, Burke’s essay sidesteps the deep, structural causes of these diminished learning and work experiences: automation, deskilling, outsourcing, and other commitments to maximizing owner and investor profit. Stressing Americans’ shrinking sense of “thrift” and “pride in workmanship” hides these pernicious forces, all of which make industrial efficiency and rewarding shop-floor jobs appear to be incompatible goals for the nation (which they are not).
It should also be noted that past educational and work opportunities that fostered “interest and talent” in the mechanical arts followed prevailing ideas of racial, gender, and physical and intellectual differences to provide only some (white, male, identifiably abled and heterosexual) American workers with the chance to develop technical skills of high value to the economy.
A priori ascriptions of low mechanical aptitude made women and African American citizens, for example, unworthy of stimulating and remunerative manufacturing jobs. Such inequities never disappeared and are strongly reasserting themselves in 2017.
I appreciate Burke’s memorialization of the sadly departed artifacts and satisfactions of US manufacturing occupations, but we should be most dismayed by the insistently undemocratic world of work that departure reflects.
Amy E. Slaton G’90 Gr’95, Narberth, PA
Wrong on Rachel Carson
In the story “Science and Error” [“Gazetteer,” Sep|Oct 2017] on Paul Offit’s book, Pandora’s Lab: Seven Stories of Science Gone Wrong, writer JoAnn Greco expressed surprise that Rachel Carson was included in what she refers to as “Offit’s rogues’ gallery.”We are surprised as well!
According to Greco, Offit claims that Carson “did real harm” by writing Silent Spring because soon after it was published the insecticide DDT was “banned for all uses, including controlling mosquitoes.”
Yes, DDT spraying reportedly initially brought the number of cases of malaria way down in Sri Lanka—which the book uses as an example of “harm”—and elsewhere. Unfortunately, as Carson warned inSilent Spring and Robert Rudd reported in 1964’s Pesticides and the Living Landscape, DDT’s widespread use over time could and repeatedly did lead to the development of resistance to it by target insect populations, including mosquitoes, thus reducing or eliminating the effectiveness of DDT as a pest control agent. According to Rudd, resistance to DDT in mosquitoes was detected as early as 1947 and again in 1952.
Furthermore, in 2016 Sri Lanka announced the elimination of malaria (no new endemic cases during the three prior years), under a program not tied to widespread broadcast spraying of chemical insecticides to control mosquitoes. The government focused on rapid diagnosis and drug treatments for the malaria parasite, especially in children, preventing its transmission by mosquitoes to other people. This, combined with bed nets, insecticide spraying indoors, on-site strategic mosquito preventive measures, and the country’s strong political commitment, involving teamwork, public education, and an enlightened public health system, all contributed to the program’s success. This successful strategy shows a more sustainable way to eliminate malaria than use of massive quantities of drugs and insecticides until resistance renders them ineffective and they need to be replaced.
Diana Post V’74, Silver Spring, MD
… And Linus Pauling, Too?
Allowing Paul Offit’s own opinions masquerading as established facts to go unchallenged has done a great disservice to the Gazette’s readers and their own health. Your writer should have placed Offit’s own clearly editorial views in context and advised readers to do their own research before accepting his unchallenged views so readily. The tip-off to what amounts to an appreciation for such controversial views should have been Offit’s one-sided case against one of our country’s true scientific geniuses, Linus Pauling, which the article left unchallenged.
Neil B. Kauffman WG’82, Swarthmore, PA
In Defense of Silk Stockings and Socialism (and Socialism)
I write in response to the letters by Richard Masella and William Buck in the Sep|Oct Gazette attacking the book Silk Stockings and Socialism by historian Sharon McConnell-Sidorick, which Julia M. Klein wrote about in the previous issue [“Arts,” Jul|Aug 2017].
I cannot possibly deal with all of the distortions in Mr. Masella’s screed. However: The myth that “reds” ran the State Department and various governmental programs in the post-World War II period was never true and has long since been debunked. The idea that an infiltration of “socialist power” somehow led to the Cold War, and ultimately to the Korean and Vietnam Wars, is ludicrous. Finally, as a member of Democratic Socialists of America, I can assure him that the charge that we are hostile to personal freedom, family, religion, etc., are untrue, although it is a fact that socialists do prioritize human rights over property rights.
Mr. Buck’s notion that segregation ended in the South because industry ran out of white workers is quite wrong. As a participant-observer and historian of the civil rights movement (the research for my Penn PhD was on the “Negro” sit-in movement of 1960), I think the evidence is clear that the major reason segregation has been largely ended is because of the struggles of the African American population and their allies in the streets, the courts, and in politics.
As for his claim that “In 35 years no worker in the South wanted a union”: I happened to be in North Carolina in the spring of 1961 doing fieldwork for my dissertation. At the request of the socialist newspaper New America, I went to Henderson to report on a Textile Workers Union strike at the Harriet-Henderson Mill, then in its third year. The workers had voted to join the union in 1943 and achieved a contract; but the company chose not to renew, working conditions deteriorated, and in 1957 the workers voted to strike. The company turned to strike breakers, eight union officials were convicted on flimsy dynamite charges (later pardoned), and the strike was broken. Most of the strikers, including women, were older, did not get their jobs back, and never found work again.
There is voluminous historical evidence that tens of thousands of workers in the South, in settings from tenant farming to modern factories, have sought unions, and there is a similar ton of evidence that employers have done everything in their power, legal and otherwise, to stop them.
Martin Oppenheimer Gr’63, Franklin Township, NJ
Diversity Has Not Meant Mediocrity in Any Sense
Am I to infer from J. M. Leone’s letter in the Jul|Aug 2017 issue [in which he asserted that “whenever you see the word diversity … substitute the word mediocrity”] that if the Penn campus was lily white it would be a far superior university? Is Leone implying that my presence on campus somehow devalued the institution? Have our three female presidents done such great harm to the University that only a white male, alone, can fix it? Should we tell incoming Provost Wendell Pritchett, an African American, that he has no place at Penn?
It takes a fanatical imagination to turn an extremely positive statistic [Penn’s acceptance rate of 9.2 percent] into a tale of a failing university. For sure, Penn’s acceptance is towards the high end for the Ivies, due in large part to the fact that we have more slots to fill. Beyond that, what has occurred at Penn over the past 20-30 years is truly remarkable and should make us all proud.
It is obviously difficult for Mr. Leone to understand, too, that diversity means more than accepting students with darker complexions than his. Diversity includes gender, age, religion, sexual orientation, gender identity, physical capabilities, cultural backgrounds, belief systems, geography, etc.
I for one am eternally grateful for the diversity that I encountered as a student that was not present in my native, practically all-white, small town. I continue to appreciate the efforts of former Interim Penn President Claire Fagin Hon’94, who focused on re-engaging our alumni people of color. I am still humbled that I was afforded the honor of working alongside Trustee Emerita Gloria Twine Chisum Gr’60 Hon’94 to establish the James R. Brister Society, which not only has become a strong advocacy group for people of color on the Penn campus (faculty, staff, and students) but has served to increase the presence of people of color on Penn’s board of trustees, the schools’ boards of overseers, and other highly responsible alumni positions—while also bringing in more alumni contributions for the University.
I am so sorry, Mr. Leone, that you allow your personal prejudices and biases to cloud your view. We should all be proud of the university Penn has become and its continued evolution.
D. Anthony Bullett C’83, Huntingdon, PA
How People Talk I
What an embarrassment to her education at Penn is Ashley Parker’s habit of speech—and she a reporter for The Washington Post and an analyst for MSNBC! In the two-page Gazette article on Parker [“Alumni Profiles,” Sep|Oct 2017] I counted five misuses by her of the word like:
“I feel like politics kind of has everything.”
“I’d be like …”
“[An] editor called me and was like …”
“I was like, ‘Yeah …’”
“You’re exhausted, but you’re also like …”
For the sake of Parker’s future in oral reporting, I hope that she can learn to speak English like a Penn graduate should.
Robert D. Kaplan L’61, Sarasota, FL
How People Talk II
In reading the article “Ron Gold’s Second Act” [Sep|Oct 2017], I was surprised and disappointed to see the F word unedited. Is this a recent decision and what is the rationale for this? I thought the Gazette might be on a higher standard.
Elizabeth Kanzawa W’80, Tokyo
While we don’t make a fetish of reproducing every stammer, false start, and redundancy in direct quotes, we do strive to preserve the flavor of natural speech—which includes the misuse/overuse of like and the occasional profanity. —Ed.
“Black Box Justice” [Sep|Oct 2017] incorrectly stated that the Philadelphia adult parole and probation department (APPD) has continuously included zip code data in its determinations about detaining and releasing criminal offenders. In fact, the APPD removed zip code variables in May 2016, after using them for seven years.