Counting Crimes: An Obsolete Paradigm

Curated on Posted on by Stefaan Verhulst

Paul Wormeli at The Criminologist: “To the extent that a paradigm is defined as the way we view things, the crime statistics paradigm in the United States is inadequate and requires reinvention….The statement—”not all crime is reported to the police”—lies at the very heart of why our current crime data are inherently incomplete. It is a direct reference to the fact that not all “street crime” is reported and that state and local law enforcement are not the only entities responsible for overseeing violations of societally established norms (“street crime” or otherwise). Two significant gaps exist, in that: 1) official reporting of crime from state and local law enforcement agencies cannot provide insight into unreported incidents, and 2) state and local law enforcement may not have or acknowledge jurisdiction over certain types of matters, such as cybercrime, corruption, environmental crime, or terrorism, and therefore cannot or do not report on those incidents…

All of these gaps in crime reporting mask the portrait of crime in the U.S. If there was a complete accounting of crime that could serve as the basis of policy formulation, including the distribution of federal funds to state and local agencies, there could be a substantial impact across the nation. Such a calculation would move the country toward a more rational basis for determining federal support for communities based on a comprehensive measure of community wellness.

In its deliberations, the NAS Panel recognized that it is essential to consider both the concepts of classification and the rules of counting as we seek a better and more practical path to describing crime in the U.S. and its consequences. The panel postulated that a meaningful classification of incidents found to be crimes would go beyond the traditional emphasis on street crime and include all crime categories.

The NAS study identified the missing elements of a national crime report as including more complete data on crimes involving drugrelated offenses, criminal acts where juveniles are involved, so-called white-collar crimes such as fraud and corruption, cybercrime, crime against businesses, environmental crimes, and crimes against animals. Just as one example, it is highly unlikely that we will know the full extent of fraudulent claims against all federal, state, and local governments in the face of the massive influx of funding from recent and forthcoming Congressional action.

In proposing a set of crime classifications, the NAS panel recommended 11 major categories, 5 of which are not addressed in our current crime data collection systems. While there are parallel data systems that collect some of the missing data within these five crime categories, it remains unclear which federal agency, if any, has the authority to gather the information and aggregate it to give us anywhere near a complete estimate of crime in the United States. No federal or national entity has the assignment of estimating the total amount of crime that takes place in the United States. Without such leadership, we are left with an uninformed understanding of the health and wellness of communities throughout the country…(More)”

The 2021 Good Tech Awards

Curated on Posted on by Stefaan Verhulst

Kevin Roose at the New York Times: “…Especially at a time when many of tech’s leaders seem more interested in building new, virtual worlds than improving the world we live in, it’s worth praising the technologists who are stepping up to solve some of our biggest problems.

So here, without further ado, are this year’s Good Tech Awards…

One of the year’s most exciting A.I. breakthroughs came in July when DeepMind — a Google-owned artificial intelligence company — published data and open-source code from its groundbreaking AlphaFold project.

The project, which used A.I. to predict the structures of proteins, solved a problem that had vexed scientists for decades, and was hailed by experts as one of the greatest scientific discoveries of all time. And by publishing its data freely, AlphaFold set off a frenzy among researchers, some of whom are already using it to develop new drugs and better understand the proteins involved in viruses like SARS-CoV-2.

Google’s overall A.I. efforts have been fraught with controversy and missteps, but AlphaFold seems like an unequivocally good use of the company’s vast expertise and resources…

Prisons aren’t known as hotbeds of innovation. But two tech projects this year tried to make our criminal justice system more humane.

Recidiviz is a nonprofit tech start-up that builds open-source data tools for criminal justice reform. It was started by Clementine Jacoby, a former Google employee who saw an opportunity to corral data about the prison system and make it available to prison officials, lawmakers, activists and researchers to inform their decisions. Its tools are in use in seven states, including North Dakota, where the data tools helped prison officials assess the risk of Covid-19 outbreaks and identify incarcerated people who were eligible for early release….(More)”.

“If Everybody’s White, There Can’t Be Any Racial Bias”: The Disappearance of Hispanic Drivers From Traffic Records

Curated on Posted on by Stefaan Verhulst

Article by Richard A. Webster: “When sheriff’s deputies in Jefferson Parish, Louisiana, pulled over Octavio Lopez for an expired inspection tag in 2018, they wrote on his traffic ticket that he is white. Lopez, who is from Nicaragua, is Hispanic and speaks only Spanish, said his wife.

In fact, of the 167 tickets issued by deputies to drivers with the last name Lopez over a nearly six-year span, not one of the motorists was labeled as Hispanic, according to records provided by the Jefferson Parish clerk of court. The same was true of the 252 tickets issued to people with the last name of Rodriguez, 234 named Martinez, 223 with the last name Hernandez and 189 with the surname Garcia.

This kind of misidentification is widespread — and not without harm. Across America, law enforcement agencies have been accused of targeting Hispanic drivers, failing to collect data on those traffic stops, and covering up potential officer misconduct and aggressive immigration enforcement by identifying people as white on tickets.

“If everybody’s white, there can’t be any racial bias,” Frank Baumgartner, a political science professor at the University of North Carolina of Chapel Hill, told WWNO/WRKF and ProPublica.

Nationally, states have tried to patch this data loophole and tighten controls against racial profiling. In recent years, legislators have passed widely hailed traffic stop data-collection laws in California, Colorado, Illinois, Oregon, Virginia and Washington, D.C. This April, Alabama became the 22nd state to enact similar legislation.

Though Louisiana has had its own data-collection requirement for two decades, it contains a loophole unlike any other state: It exempts law enforcement agencies from collecting and delivering data to the state if they have an anti-racial-profiling policy in place. This has rendered the law essentially worthless, said Josh Parker, a senior staff attorney at the Policing Project, a public safety research nonprofit at the New York University School of Law.

Louisiana State Rep. Royce Duplessis, D-New Orleans, attempted to remove the exemption two years ago, but law enforcement agencies protested. Instead, he was forced to convene a task force to study the issue, which thus far hasn’t produced any results, he said.

“They don’t want the data because they know what it would reveal,” Duplessis said of law enforcement agencies….(More)”.

Crime Prediction Software Promised to Be Free of Biases. New Data Shows It Perpetuates Them

Curated on Posted on by Stefaan Verhulst

Article by Aaron Sankin, Dhruv Mehrotra for Gizmodo, Surya Mattu, and Annie Gilbertson: “Between 2018 and 2021, more than one in 33 U.S. residents were potentially subject to police patrol decisions directed by crime prediction software called PredPol.

The company that makes it sent more than 5.9 million of these crime predictions to law enforcement agencies across the country—from California to Florida, Texas to New Jersey—and we found those reports on an unsecured server.

The Markup and Gizmodo analyzed them and found persistent patterns.

Residents of neighborhoods where PredPol suggested few patrols tended to be Whiter and more middle- to upper-income. Many of these areas went years without a single crime prediction.

By contrast, neighborhoods the software targeted for increased patrols were more likely to be home to Blacks, Latinos, and families that would qualify for the federal free and reduced lunch program.

These communities weren’t just targeted more—in some cases they were targeted relentlessly. Crimes were predicted every day, sometimes multiple times a day, sometimes in multiple locations in the same neighborhood: thousands upon thousands of crime predictions over years. A few neighborhoods in our data were the subject of more than 11,000 predictions.

The software often recommended daily patrols in and around public and subsidized housing, targeting the poorest of the poor.

“Communities with troubled relationships with police—this is not what they need,” said Jay Stanley, a senior policy analyst at the ACLU Speech, Privacy, and Technology Project. “They need resources to fill basic social needs.”…(More)”.

How Courts Embraced Technology, Met the Pandemic Challenge, and Revolutionized Their Operations

Curated on Posted on by Stefaan Verhulst

Report by The Pew Charitable Trusts: “To begin to assess whether, and to what extent, the rapid improvements in court technology undertaken in 2020 and 2021 made the civil legal system easier to navigate, The Pew Charitable Trusts examined pandemic-related emergency orders issued by the supreme courts of all 50 states and Washington, D.C. The researchers supplemented that review with an analysis of court approaches to virtual hearings, e-filing, and digital notarization, with a focus on how these tools affected litigants in three of the most common types of civil cases: debt claims, evictions, and child support. The key findings of this research are:

  • Civil courts’ adoption of technology was unprecedented in pace and scale. Despite having almost no history of using remote civil court proceedings, beginning in March 2020 every state and D.C. initiated online hearings at record rates to resolve many types of cases. For example, the Texas court system, which had never held a civil hearing via video before the pandemic, conducted 1.1 million remote proceedings across its civil and criminal divisions between March 2020 and February 2021. Similarly, Michigan courts held more than 35,000 video hearings totaling nearly 200,000 hours between April 1 and June 1, 2020, compared with no such hearings during the same two months in 2019.Courts moved other routine functions online as well. Before the pandemic, 37 states and D.C. allowed people without lawyers to electronically file court documents in at least some civil cases. But since March 2020, 10 more states have created similar processes, making e-filing available to more litigants in more jurisdictions and types of cases. In addition, after 11 states and D.C. made pandemic-driven changes to their policies on electronic notarization (e-notarization), 42 states and D.C. either allowed it or had waived notarization requirements altogether as of fall 2020.
  • Courts leveraged technology not only to stay open, but also to improve participation rates and help users resolve disputes more efficiently. Arizona civil courts, for example, saw an 8% drop year-over-year in June 2020 in the rate of default, or automatic, judgment—which results when defendants fail to appear in court—indicating an increase in participation. Although national and other state data is limited, court officials across the country, including judges, administrators, and attorneys, report increases in civil court appearance rates.
  • The accelerated adoption of technology disproportionately benefited people and businesses with legal representation—and in some instances, made the civil legal system more difficult to navigate for those without...(More)”.

An Obsolete Paradigm

Curated on Posted on by Stefaan Verhulst

Blogpost by Paul Wormelli: “…Our national system of describing the extent of crime in the U.S. is broken beyond repair and deserves to be replaced by a totally new paradigm (system). 

Since 1930, we have relied on the metrics generated by the Uniform Crime Reporting (UCR) Program to describe crime in the U.S., but it simply does not do so, even with its evolution into the National Incident-Based Reporting System (NIBRS). Criminologists have long recognized the limited scope of the UCR summary crime data, leading to the creation of the National Crime Victimization Survey (NCVS) and other supplementary crime data measurement vehicles. However, despite these measures, the United States still has no comprehensive national data on the amount of crime that has occurred. Even after decades of collecting data, the 1968 Presidential Crime Commission report on the Challenge of Crime in a Free Society lamented the absence of sound and complete data on crime in the U.S., and called for the creation of a National Crime Survey (NCS) that eventually led to the creation of the NCVS. Since then, we have slowly attempted to make improvements that will lead to more robust data. Only in 2021 did the FBI end UCR summary-based crime data collection and move to NIBRS crime data collection on a national scale.

Admittedly, the shift to NIBRS will unleash a sea change in how we analyze crime data and use it for decision making. However, it still lacks the completeness of national crime reporting. In the landmark study of the National Academy of Sciences Committee on Statistics (funded by the FBI and the Bureau of Justice Statistics to make recommendations on modernizing crime statistics), the panel members grappled with this reality and called out the absence of national statistics on crime that would fully inform policymaking on this critical subject….(More)”

When Machines Can Be Judge, Jury, and Executioner

Curated on Posted on by Stefaan Verhulst

Book by Katherine B Forrest on “Justice in the Age of Artificial Intelligence”: “This book explores justice in the age of artificial intelligence. It argues that current AI tools used in connection with liberty decisions are based on utilitarian frameworks of justice and inconsistent with individual fairness reflected in the US Constitution and Declaration of Independence. It uses AI risk assessment tools and lethal autonomous weapons as examples of how AI influences liberty decisions. The algorithmic design of AI risk assessment tools can and does embed human biases. Designers and users of these AI tools have allowed some degree of compromise to exist between accuracy and individual fairness.

Written by a former federal judge who lectures widely and frequently on AI and the justice system, this book is the first comprehensive presentation of the theoretical framework of AI tools in the criminal justice system and lethal autonomous weapons utilized in decision-making. The book then provides a comprehensive explanation as to why, tracing the evolution of the debate regarding racial and other biases embedded in such tools. No other book delves as comprehensively into the theory and practice of AI risk assessment tools….(More)”.

Manipulation As Theft

Curated on Posted on by Stefaan Verhulst

Paper by Cass Sunstein: “Should there be a right not to be manipulated? What kind of right? On Kantian grounds, manipulation, lies, and paternalistic coercion are moral wrongs, and for similar reasons; they deprive people of agency, insult their dignity, and fail to respect personal autonomy. On welfarist grounds, manipulation, lies, and paternalistic coercion share a different characteristic; they displace the choices of those whose lives are directly at stake, and who are likely to have epistemic advantages, with the choices of outsiders, who are likely to lack critical information. Kantians and welfarists should be prepared to endorse a (moral) right not to be manipulated, though on very different grounds.

The moral prohibition on manipulation, like the moral prohibition on lies, should run against officials and regulators, not only against private institutions. At the same time, the creation of a legal right not to be manipulated raises hard questions, in part because of definitional challenges; there is a serious risk of vagueness and a serious risk of overbreadth. (Lies, as such, are not against the law, and the same is true of unkindness, inconsiderateness, and even cruelty.) With welfarist considerations in mind, it is probably best to start by prohibiting particular practices, while emphasizing that they are forms of manipulation and may not count as fraud. The basic goal should be to build on the claim that in certain cases, manipulation is a form of theft; the law should forbid theft, whether it occurs through force, lies, or manipulation. Some manipulators are thieves….(More)”

Virtual Juries

Curated on Posted on by Stefaan Verhulst

Paper by Valerie P. Hans: “The introduction of virtual or remote jury trials in response to the COVID-19 pandemic constitutes a remarkable natural experiment with one of our nation’s central democratic institutions. Although it is not a tightly controlled experimental study, real world experiences in this natural experiment offer some insights about how key features of trial by jury are affected by a virtual procedure. This article surveys the landscape of virtual jury trials. It examines the issues of jury representativeness, the adequacy of virtual jury selection, the quality of decision making, and the public’s access to jury trial proceedings. Many have expressed concern that the digital divide would negatively affect jury representativeness. Surprisingly, there is some preliminary evidence that suggests that virtual jury selection procedures lead to jury venires that are as diverse, if not more diverse, than pre-pandemic jury venires. Lawyers in a demonstration project reacted favorably to virtual voir dire when it was accompanied by expansive pretrial juror questionnaires and the opportunity to question prospective jurors. A number of courts provided public access by live streaming jury trials. How a virtual jury trial affects jurors’ interpretations of witness testimony, attorney arguments, and jury deliberation remain open questions….(More)”

What Data About You Can the Government Get From Big Tech?

Curated on Posted on by Stefaan Verhulst

 Jack Nicas at the New York Times: “The Justice Department, starting in the early days of the Trump administration, secretly sought data from some of the biggest tech companies about journalists, Democratic lawmakers and White House officials as part of wide-ranging investigations into leaks and other matters, The New York Times reported last week.

The revelations, which put the companies in the middle of a clash over the Trump administration’s efforts to find the sources of news coverage, raised questions about what sorts of data tech companies collect on their users, and how much of it is accessible to law enforcement authorities.

Here’s a rundown:

All sorts. Beyond basic data like users’ names, addresses and contact information, tech companies like Google, Apple, Microsoft and Facebook also often have access to the contents of their users’ emails, text messages, call logs, photos, videos, documents, contact lists and calendars.

Most of it is. But which data law enforcement can get depends on the sort of request they make.

Perhaps the most common and basic request is a subpoena. U.S. government agencies and prosecutors can often issue subpoenas without approval from a judge, and lawyers can issue them as part of open court cases. Subpoenas are often used to cast a wide net for basic information that can help build a case and provide evidence needed to issue more powerful requests….(More)”.