Making Data Great Again
summary
Describes building an Integrated Data Infrastructure (IDI) in New Zealand and a US Administrative Data Research Facility that links health, education, justice and welfare records, uses Jupyter notebooks and virtual RDCs for collaborative measurement, and trains agency staff; a linked-analysis finds only 3% stable employment among ex‑offenders across eight quarters; an Allegheny County prediction system—evaluated at Stanford—reduced racial disparities and improved maltreatment screening; argues utility, timeliness and measurement decisions matter as much as privacy and that linkage, cleaning and thousands of coding choices drive results.
AI-written summary · about
chapters (22)
- 0:00Intro
- 3:09Why measurement is important
- 5:41New types of data
- 7:48New Zealand example
- 10:12Risk factors
- 11:21Integrated data infrastructure
- 13:42Data in the center
- 16:42Legal framework
- 19:11Data is messy
- 24:26Looking at the data
- 26:42Looking at the linkages
- 27:39Realworld example
- 30:30Privacy tradeoff
- 32:38The Census
- 37:03Statistics to Inform Policy
- 44:00Measuring Privacy Neutrality
- 45:30Impact Evaluation
- 47:20Virtual RDC
- 47:40Collaboration
- 48:20Data Discovery
- 53:07Utility
- 56:22Feinberg Problem
related talks
01h 01mData Con LA 2020 - Friday Keynotes
28mData 4 Development
41mUsing Data Riches: A Tale of two projects
43mData-Driven Innovation in the Federal Government.
44mManagement of Data Communities
27mNavigating Data's Next Great Shift
41mAccelerating privacy-enhancing data processing
42mBuilding a data infrastructure for large municipal government