Geofence warrants are a fairly new concept mostly involving data from Google.
Privacy advocates say they violate civil liberties. For example, the @ACLU found that law enforcement was using geofence data to track Black Lives Matter protesters in 2016. themarkup.org/ask-the-markup…
California is one of few states where law enforcement agencies must disclose geofence warrants to a state dataset.
We looked at that dataset—as well as a geofence transparency report from Google—and found the numbers didn’t add up.
Only 186 warrants that contained the word “Google” had been reported to the state of California in 2020.
Google, on the other hand, said that the company received 1,909 geofence warrant requests from California in that same year.
2019 data revealed a similar pattern.
There are many reasons this could be happening. One factor that both privacy experts and the California Department of Justice highlighted is that agencies can delay reporting a warrant if it’s sealed from the public.
And there are material consequences. The lack of reporting makes it hard for civil liberties advocates to keep tabs on the warrants and fight them.
“... if you don’t know about them, how do you get involved in the cases?” said Jennifer Lynch of @EFF. themarkup.org/privacy/2021/1…
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Our colleagues at Germany’s @SZ used Citizen Browser data to uncover the messaging that made its way into voters’ news feeds during the country’s recent election cycle. getrevue.co/profile/citize…
They found voters of the far-right AfD party were more likely to see posts attacking issues like climate change, migration, and COVID-19 from their party leaders.
Meanwhile, voters from other parties were generally served coverage on those topics from established media outlets.
We’re thrilled that our tools are being used to reveal how polarization on Facebook is playing out beyond the United States.
This is a huge milestone for our small nonprofit newsroom—mind if we indulge in a quick recap of our recent work? ⬇️
This story from @darakerr was one of @ToddFeathers’ favorite pieces of journalism this year.
“An example of investigating an industry that tries to turn people into data and turning it around by using data to show the tragedies that attitude can create.” themarkup.org/working-for-an…
“This story investigates a system that upholds segregation through arbitrary and inconsistent rules. I especially appreciated students’ perspectives.” themarkup.org/news/2021/05/2…
NEW: Amazon placed items from its house brands and exclusives ahead of competitors with better customer ratings and more sales, @adrjeffries and @leonyin found after examining the results of nearly 3,500 popular product searches. themarkup.org/amazons-advant…
Take Amazon’s Happy Belly Cinnamon Crunch cereal, for example.
It had four stars and 1,010 reviews, but Amazon gave it the number one search result spot, ahead of Cap’n Crunch, which had five stars and 14,069 reviews.
We found that knowing only whether a product was an Amazon brand or exclusive could predict in seven out of 10 cases whether the company would rank the item first in search results.
Have you heard of D.R. Horton, Lennar Corporation, or PulteGroup Inc.? These are our nation’s largest home builders.
They have some things in common beyond new construction: owning mortgage companies that denied applicants of color at higher rates than their White counterparts.
Let’s start with the widest disparity: DHI Mortgage, which finances homes built by parent company D.R. Horton, the nation’s largest home builder.
It was 160% more likely to deny Black applicants and 100% more likely to deny Latinos than similar White applicants.
Nationally, our analysis found, the mortgage industry was 40% to 80% more likely in 2019 to deny home loans to people of color than to White people with similar financial characteristics.
We also found significant disparities in 89 metropolitan areas, spanning every region of the country, from Boston, Mass., to Riverside, Calif.