Screened & Scored in D.C.
EPIC spent 14 months investigating the D.C. government’s use of automated decision-making systems. This report aims to shed light by providing a view of the many ADS that shape the course of District residents’ lives.
The D.C. government outsources critical governmental decisions to automated decision-making systems in areas such as public benefits, healthcare, policing, and housing. As a result, District residents are surveilled, screened, and scored every day. But because of weak government transparency laws, opaque procurement processes, the power and influence of tech vendors, and the decline in local journalism, it has been difficult to uncover the details of how many automated decision-making systems are used in government programs.
EPIC spent 14 months investigating the D.C. government’s use of automated decision-making systems. This report aims to shed light by providing as comprehensive a view as possible of the many automated decision-making systems that shape the course of District residents’ lives.
Screened & Scored in D.C. Panel Discussion
On September 21, EPIC hosted a panel about the use of automated decision-making systems throughout the District of Columbia. EPIC Scholar-In-Residence Virginia Eubanks moderated a panel of National Fair Housing Alliance’s Dr. Michael Akinwumi, Upturn’s Natasha Duarte, CLASP and JustTech’s Clarence Okoh, and EPIC Counsel Ben Winters. As part of EPIC’s ongoing Screening and Scoring Project, the panel wove topics from in EPIC’s Screened and Scored in the District of Columbia report with the panelist’s work and expertise.
FOI Documents
As part of the research for Screened & Scored in D.C., EPIC sent a number of Freedom of Information records requests to D.C. agencies and the U.S. Department of Veterans Affairs. We have also included records released from the D.C. Pretrial Services Agency obtained from previous work related to automated decision-making systems in the criminal justice context. EPIC will continue to update this section as new information is released. We are still awaiting responses from the following D.C. agencies: D.C. Public Schools, Dept. of Health Care Finance, Child and Family Services Agency, and the Dept. of Transportation.
-
2019 Validation Study
March 8, 2020
-
2019 Predictive Bias Report
March 8, 2020
-
2019 Letter Showing a List of Factors Changed after 2015 Risk Assessment
March 8, 2020
-
2017 Family Resource Guide
Nov. 30, 2021
-
Structured Decision Making Demographic Breakdown
Nov. 30, 2021
-
2017 Care Planning and Coordination Handbook
Nov. 30, 2021
-
2020 Risk Offense Chart
Nov. 30, 2021
-
2020 Presentation Examining the Structured Decision Making Risk Assessment Tool
Nov. 30, 2021
-
2018 Memorandum Clarifying Aspects of the Structured Decision Making Tool
Nov. 30, 2021
-
2020 Office of Research and Evaluation’s Memorandum on the Structured Decision Making Tool
Nov. 30, 2021
-
2017 Structured Decision Making Overview and Procedures Presentation
Nov. 30, 2021
-
2017 Proposed DYRS Risk Re-Assessment Scoresheet
Nov. 30, 2021
-
2016 Structured Decision Making Overview
Nov. 30, 2021
-
Youth Level of Service/Case Management Inventory (YLS/CMI) 2.0 User’s Manual
Nov. 30, 2021
-
Pondera Proposal
July 19, 2021
-
Pondera Contract
July 19, 2021
-
D.C. OCP Solicitation Letter: Fraud Case Management and Data Analytics System Subscription
July 19, 2021
-
D.C. DHS-Pondera Bilateral Modification Agreement
July 30, 2021
-
DCAS Application Intake Intelligence Gathering Spreadsheet
May 20, 2022
-
Customer Service Scripts Provided by Salesforce
May 20, 2022
-
Customer Service Scripts Provided by Coaching Content
May 20, 2022
-
2018 RentGrow Contract 26A
August 21, 2021
-
2018 RentGrow Contract 26B
August 21, 2021
-
Sample Ineligibility Letter
August 21, 2021
-
Proposed Ineligibility Letter
August 21, 2021
-
Draft Applicant Screening Procedures
August 21, 2021
-
Applicant Family Selection Criteria
August 21, 2021
-
Final Eligibility Determination Processing
August 21, 2021
-
Final Rule Notice for Clarification and Guidance on Screening and Eviction for Drug and Criminal Activity
August 21, 2021
-
Procedures for Processing LRSP Referrals and DCHA Eligibitility Determination
August 21, 2021
-
Bids and Proposals Related to the Contract Between D.C. Dept. of Health Care Finance and Liberty Healthcare
March 28, 2022
-
2017 Dun & Bradstreet Supplier Risk Manager (SRM) Contract
March 7, 2022
-
Dun & Bradstreet Supplier Risk Manager (SRM) OneLogin Migration Procedures
March 7, 2022
-
Dun & Bradstreet Supplier Risk Manager (SRM) OneLogin Migration Service Alert
March 7, 2022
-
Bids and Proposals Related to the Contract Between D.C. Dept. of Employment Services and Pondera
Novemeber 29, 2021
Automated Decision-Making Systems in D.C.
Issue | Agency | Tool | Goals/Decisions Made |
Education | University of the District of Columbia | Aspire, Accomplish, Take on the World (A.AC.T) — EAB, Inc.’s Risk Assessment for Student Guidance | To increase graduation rates by predicting which students are not likely to graduate and alerting school support staff |
Education | Office of Deputy Mayor for Education | EdScape — a set of interactive visualizations and downloadable datasets. | To (1) act as a usable source of information that informs how and where new schools, programs, or facility capacity is needed, and (2) provide the public with the same information as policymakers for transparency |
Education | Office of Deputy Mayor for Education | EdStat — a statistical model | To increase school attendance |
Health | D.C. Office of Veterans Affairs | COVID-19 Prognostic Tool Developed In-House | To inform COVID-19 treatment and clinical training by automatically generating a 120-day mortality risk score for patients based on age, BMI, preexisting health conditions, and vital signs |
Health | D.C. Department of Health | Prescription Drug Monitoring Program | To reduce misuse and diversion of prescription drugs by monitoring when and how often certain drugs (opioids, benzodiazepine, etc.) are prescribed |
Health | D.C. Medicaid/Department of Health Care Finance | Electronic Visit Verification (EVV) by Sandata Technologies LLC | To ensure people receiving home-based healthcare are receiving proper support (and being billed accordingly) by tracking data about beneficiaries, their caregivers, and the services rendered |
Health | D.C. Department of Health Care Finance | InterRAI-Home Care (InterRAI HC) | To facilitate continuing medical care and flag health risks through assessments capturing and evaluating patient data |
Housing | D.C. Housing Authority | RentGrow — a risk-assessment algorithm | To screen out applicants for housing based on predictions of who won’t make payments on time |
Housing | Interagency Council on Homelessness | Vulnerability Index and Service Prioritization Decision Assistance Tool (VI-SPDAT) — a statistical tool developed by OrgCode Consulting, Inc. | To assist case workers in determining who gets housing assistance first and what that assistance looks like |
Housing | Interagency Council on Homelessness | Service Prioritization Decision Assistance Tool (SPDAT) — a statistical tool developed by OrgCode Consulting, Inc. | A more comprehensive tool than VI-SPDAT used on individuals who are presumed to be highly vulnerable but score too low on the VI-SPDAT to qualify for permanent supportive housing |
Housing | D.C. Department of Buildings | Proactive Inspection Program Risk-Based Algorithm | To make housing code violation inspections more efficient by using factors like a building’s age and landlord’s history of violations to determine the which houses should be inspected when |
Economic Opportunity | D.C. Department of Human Services | Pondera—FraudCaster & CaseTracker | To monitor and manage SNAP referrals and identify potential fraud by analyzing recipient data and generating fraud risk scores |
Economic Opportunity | D.C. Department of Employment Services | Pondera—FraudCaster & CaseTracker | To identify potential improper payments and fraud in the unemployment insurance program by compiling data about each recipient and generating an algorithmic risk score |
Economic Opportunity | Office of Contracting and Procurement/Department of General Services | Dun and Bradstreet’s Data Universal Numbering System (D-U-N-S) | To score companies bidding for government contracts that is similar to credit scores using proprietary ratings based on predictive analytics, as well as past suits, liens, and judgments |
Economic Opportunity | D.C. Department of Health Care Finance (DHCF) | District of Columbia Access System (DCAS) | To integrate and streamline the Medicaid, SNAP, and TANF benefits process through automated, real-time verification of eligibility information |
Criminal Justice | D.C. Department of Youth and Rehabilitative Services | Structured Decision Making (SDM) Tool, developed by DYRS and the Annie E. Casey Foundation | To better rehabilitate youth offenders by predicting how likely they are to re-offend and informing how restrictive their placement will be (e.g., whether they will be placed in a juvenile facility) using factors like prior adjudications, school attendance regularity, and peer relationships |
Criminal Justice | D.C. Department of Youth and Rehabilitative Services | Child And Adolescent Functional Assessment Scale (CAFAS) and Pre-School and Early Childhood Assessment Scale (PECFAS) | To improve youth treatment and rehabilitation by assessing youth offenders’ day-to-day functioning across different life skills to help determine their placement and treatment during (or instead of) commitment to a facility |
Criminal Justice | Department of Forensic Sciences | Automated Fingerprint Identification System (AFIS) | To facilitate forensic investigations by analyzing fingerprints and alerting investigators when a fingerprint matches an existing record in the AFIS database |
Criminal Justice | Metropolitan Police Department (MPD) | Automated License Plate Readers (ALPRs) —cameras equipped with technology that selectively finds license plates, reads them, and sends the info to a central database. | To automatically capture license plate numbers, store them in an MPD database, and compare them to a “hot list” of wanted license plates |
Criminal Justice | Metropolitan Police Department (MPD) | Shotspotter | To detect gunshots through acoustics in real-time and alert authorities |
Criminal Justice | Metropolitan Police Department (MPD) | Predictive Policing (funded by Department of Justice) | To inform policing efforts through predictive data analytics |
Criminal Justice | Metropolitan Police Department (MPD) | TrapWire | To analyze citizens’ reports of “suspicious activity” |
Criminal Justice | Metropolitan Police Department (MPD) | D.C. Gang Database | To track suspected gang members, direct surveillance and police activity toward tracked individuals and increase sentence severity for those convicted |
Criminal Justice | D.C. Department of Transportation (DDOT) | D.C. StreetSafe —Automated Traffic Enforcement (ATE) | To automatically detect and record traffic violations, which human officers review to issue fines |
Criminal Justice | D.C. Department of Pretrial Services | Pre-Trial Risk Assessment Instrument (RAI) | To recommend appropriate release conditions for criminal defendants by generating individual recidivism risk scores using 43 factors from five categories—criminal history, current charge, criminal justice system status, drug test results, and social/demographic attributes |
Criminal Justice | D.C. Sentencing Commission | The Guidelines Reporting Information Data (GRID) System and Guidelines Scoring System (GSS) | To monitor sentencing trends and inform sentencing guidelines by integrating arrest, court, and criminal history data and calculating criminal history scores |
Criminal Justice | D.C. Superior Court’s Family Court, Social Services Division, D.C. Child Guidance Clinic | Structured Assessment of Violence Risk in Youth (SAVRY) | To inform juvenile sentencing decisions by evaluating 24 factors, including an offender’s criminal history, social factors, and demographic information, and assigning offenders a recidivism risk score |
Resources
-
What’s in a name? A survey of strong regulatory definitions of automated decision-making systems
Ben Winters, EPIC | Oct 2022
-
Pondera’s Fraud Prediction Algorithms for Public Benefits
EPIC | 2022
-
Screening & Scoring Project
EPIC
-
The Shakedown State: Fraud Detection and Predicting the Past
Virginia Eubanks, EPIC | 2022
-
EPIC’s Testimony Supporting D.C. Council Bill B24-0558, The Stop Discrimination by Algorithms Act (“SDAA”)
EPIC | 2022
-
Guides for advocates trying to fight and fix public benefits technology systems
Benefits Tech Advocacy Hub
Support Our Work
EPIC’s work is funded by the support of individuals like you, who help us to continue to protect privacy, open government, and democratic values in the information age.
Donate