Nightly Nuance - May 18, 2021

Hello, my friends. Let’s use a totally different side of our brains than we used in today’s podcast and talk about data and privacy. Many of us just completed the US Census. When we did, we answered lots of questions honestly (presumably) knowing that our data would give rise to hundreds of published statistics. I don’t know about you, but I didn’t even once consider the risks to my privacy of participating in the Census.

The risks are not insubstantial! The Census discovered that just about anyone could analyze data from the 2010 Census, compare it to publicly available information, and trace individual responses back to about 53 million people. YIKES. So, the Census is using a technique called differential privacy--introducing “noise” into the data to try to protect individual responses without compromising the statistical conclusions drawn from the data. And this approach has sparked controversy.

Resources:

Can Differential Privacy Save the Census? (The Markup)

Will New Privacy Changes Protect Census Data or Make Things Worse? (The Markup)

Privacy as contextual integrity (NYU)

The Coup We Are Not Talking About (The New York Times)

Judges hear arguments over contentious Census privacy tool (ABC News)

See acast.com/privacy for privacy and opt-out information.