United States Marshals Service’s Post

🤓 I work a lot with data, asking how we can know that public money is achieving what we want it to: did my project create the change, or would it have happened anyway? How much of the change was down to my project, vs other external influences? 🤔 🤔 These questions can be surprisingly hard to answer, especially in complicated policy areas - like reducing reoffending, for example. So when I start a new role, the first thing I do is read, a LOT, to understand the context 🕵♂️📚 This is my first role in Ministry of Justice UK, and I want to share my learning journey as I go. So I'm going to post some data summaries over the next few months. I'll start with the very, VERY basics - below is my first, on prisons in England & Wales: 📊 85,576 people were in prison as of Sept 2023 ☝ The prison population has risen steadily since WW2 - and by 80% in the last 30 years 📆 It is still growing: by 2027, between 93,000 - 106,000 people are expected to be in prison 💷 The average annual overall cost of a prison place is £46,696 📊 Per 100,000 people in England & Wales, 146 are in prison. Globally this ranks middle-ish: higher than Western European nations like France (109), Italy (102) and Germany (67), but much lower than eg: the USA (531) or South Africa (258) I'd love to hear some initial thoughts about these data sets! Are the numbers big / small? What kind of challenges might they present? Is there any other data we should think about alongside them that would give us more context? My next post will probably cover reoffending - comment below if there's anything specific you'd like to know in future! (This is all be published data. DM me for links if you're a complete nerd) #projectmanagement #analytics #data #research #justice

"The mundane artifacts surrounding prisons can be a rich source of data in understanding how the prison institution sees their own role in the rehabilitation and care for inmates or the rather the lack thereof" (p. 88). That's a quote from my #dissertation I wrote to earn my #PhD from Clemson University in 2018. It won the dissertation of the year award from the graduate school too! I wrote everything I could, and I spent hours talking to my two #data friends who wrote the #python code and built my #Tableau visuals. Huge shout out to Ben Webster and Katie Bakewell! --- The premise of my dissertation is a research method I developed: Institutional Genre Analysis. What is it? Using large language models and machine learning to understand and ultimately critique how large organizations use language to get what they want. For my cite of study, I chose prisons because I had recently watched Jon Oliver's HBO segment on it. My passion was data, not prisons. I really wrestled with that distinction as someone who wanted to be a "social justice scholar," so it ended up being quite self-reflective in understanding motive for change. I had a blast writing it. I knew entering my PhD program in 2015 that I wanted to do something with prisons and big data, so everything I did moved me to that, which is why I finished in less than 3 years. :) If you are looking into how you can leverage #ai and #llm in your academic research, please consider citing my dissertation or publication in the Journal of Writing Analytics. Links in the comments!

Welcome to this incendiary and special edition of Light ‘Em Up. Thanks to you, we’re now being actively downloaded in 104 countries, globally. In the interest of always gathering and providing you with the facts, never fiction — we have an exclusive in-depth interview with a high-ranking police administrator with a small, rural police department in the “Magnolia State” just for you! Speaking on the promise of his anonymity, identified only by his title (Assistant Chief of Police), the municipality that he serves has approximately 8,500 people with not a lot of “pass through” traffic in the jurisdiction. It is staffed by approximately 30 accredited officers. The Assistant Chief mentioned that “working in a small town brings to focus the importance of the development of relationships. The good news is — is that at the present crime is low and it is getting lower” in this sleepy, but growing southern town. In this fact-laden episode we explore: ♦ The never-ending challenges of 21st century policing in a small(er) southern town. ♦ Budgeting aspects of small-town policing and the struggle to pay its personnel a competitive wage. ♦ Police “culture” & how retaining staff is a constant struggle. ♦ The findings following an intensive, comprehensive, in-depth study on the importance of prioritizing and reducing unconscious bias among officers entitled Seeing Color in Police Discretion. ♦ Unconscious bias and how this can affect officer discretion, views, and perceptions. ♦ The importance of discretion in law enforcement & the wide latitude officers have in their use of that discretion and how this can be relevant to non-law enforcement industries that you may serve. This episode serves simultaneously as an “Open Forum Discussion Topic” and part one of a lecture series on policing for our Criminal Justice and Forensic Science scholars (Level 3500 Required Listening). Tune in and be educated and empowered: Follow the link in the comment section to the full episode which includes BONUS content for you and you only! #police #civilrights #constitutionalrights #racialbias #bias #policedepartment #criminaljusticereform #criminaljustice #publicsafety #crime #experts #litigation #datascience #dataanalytics #dataanalysis #datadriven #evidencebased #racialequity #racialjustice #arrest #intervention #oversight #pandora #pandorapodcasts

In October 2022 the Global Collaboration of Evidence Based Policing broadcast a fully live studio-quality 24- hour continuous conference out of two hubs – Christchurch, New Zealand and London, United Kingdom. The conference brought together world-leading criminologists from across the globe, academic thinkers, and police professionals from dozens of agencies, to look at the role that evidence, its legitimacy, and policing leadership have in everyday policing – shaping “exceptional” policing, the evidence-based way. The Global Collaboration of Evidence Based Policing has now teamed-up with Policing TV to bring the best of the conference material to a wider audience. In this presentation ‘Bad apples, barrels, orchards: Data science will tell you which’, Dr Tim Cubitt, PhD of Australian Institute of Criminology explores how data science approaches can be used for the detection and management of police misconduct. He discusses how these approaches can be introduced into practice, and what we stand to gain from applying these types of metrics to better understand police misconduct. Watch on PolicingTV https://lnkd.in/eE_S3ijA #lawenforcement #policing #police #professionalstandards #datascience #policemisconduct

This is the best description and discussion about how algorithms have, in my opinion, failed to more effectively accommodate the ongoing evolution in social constructs. For example, why/how should the social norms and judicial processes reflect todays communities by relying on data tainted by biases of 20 years ago? “These machine learning predictions inform a host of decisions, such as: granting bail, outcomes of immigration cases, the kinds of sentences people face (community-based, custodial or suspended), prison security classifications and assignments to rehabilitation programmes. They also help decide the conditions on how people convicted of crimes are supervised in the community and whether or not they can be released early from prison.”

🌟🔥📊 Bureau of Justice Statistics (BJS): Your Source for Criminal Justice Data 🇺🇸🔍 The Bureau of Justice Statistics (BJS), part of the U.S. Department of Justice, is your go-to resource for reliable crime and criminal justice statistics in the United States. Here’s what you need to know: New Releases: Correctional Populations in the United States, 2022: Dive into data on probation, parole, and incarcerated populations. Probation and Parole in the United States, 2022: Explore info about adults under correctional supervision. Crimes Involving Juveniles, 1993–2022: Uncover archived data on juvenile offenses. Data by Topic: Access datasets covering various criminal justice subjects. Create custom visualizations using data analysis tools. Announcements: BJS is hiring! 🌟 https://lnkd.in/gC6ZTXsu

Two great trends are facing societies across the world today, that should be a top priority for policymakers: a) a growing strain on #criminaljustice professionals & #lawenforcement agencies as prison and probation populations grow b) an increasing recognition of the potential for #artificialintelligence and #datascience in historically ‘human’ decision-making, I teamed up with Dr. Miri Zilka (University of Cambridge), Caitlin K. (Technical University Munich) and Jiri Hron (Google DeepMind), to study the exciting merging of the two aforementioned trends: investigating how strategic decisions are made about where, when and how to employ #datascience in #policing, in collaboration with Police Scotland. Read our newest publication here: https://lnkd.in/dXKHhzDx

🌟🔥📊 Bureau of Justice Statistics (BJS): Your Source for Criminal Justice Data 🇺🇸🔍 The Bureau of Justice Statistics (BJS), part of the U.S. Department of Justice, is your go-to resource for reliable crime and criminal justice statistics in the United States. Here’s what you need to know: New Releases: Correctional Populations in the United States, 2022: Dive into data on probation, parole, and incarcerated populations. Probation and Parole in the United States, 2022: Explore info about adults under correctional supervision. Crimes Involving Juveniles, 1993–2022: Uncover archived data on juvenile offenses. Data by Topic: Access datasets covering various criminal justice subjects. Create custom visualizations using data analysis tools. Announcements: BJS is hiring! 🌟 https://lnkd.in/gbACFznR

Interesting analysis on the geographic range of offences for prolific offenders. It implies an offender focused prevention strategy is required than purely a geographic/ hotspot approach

#Throwback #Reshare... Originally posted on 5th October 2023 for new readers who missed this thought... enjoy! Service models don't achieve outcomes, people do.  I've been thinking a lot of about systems and big data recently and how it applies to working with individuals in the prison system.  I've been around the prison system long enough not to be bowled over by glossy service leaflets and infographics outlining the neat journey that someone engaging with a service will go through. In reality, what people who use services will remember is simply:  1️⃣ What you did. 2️⃣ How they are treated by you.  Period. Do you do what you said you would, and do you treat people with dignity and respect, irrespective of whatever task or action you are carrying out. Do you rush conversations or give people the gift of your time and undivided attention. These are the metrics that matter to people who choose to engage with you or the service you're a part of.  Models and systems are, in my mind, primarily about communication and resource efficiency. Are our activities driving us closer or further away from our stated goals, and how do we deploy our resources in the most efficient and effective ways? Models and systems are important, but without #people they are nothing.  This is why artificial intelligence will always play second fiddle to humans, as people inherently like to engage with people, and that's something I'd wager will never change.  #AlexThinkingOutLoud #HMPPS #prison #MOJ #rehabilitation #probation #criminaljustice #jail #criminaljustice #criminology