Machine Learning Engineer
Hi there. You might remember my last blog post about fake comments in the FCC net neutrality proceeding for a project I did at Metis. In the last couple of weeks since that post, I’ve been working day and night² as a member of the Truth in Public Comments (TiPC) project at the Startup Policy Lab³ here in SF.
We wanted to uncover more facts beneath the suspicions so many others have raised. The best way was to take as direct a measurement as possible — to survey the commenters who provided an email address and ask whether they submitted the comment in the first place.⁴
Here are a few charts from survey results as of 8pm PT on December 10:⁵
We are still working hard on the final report but I just wanted to share with you our preliminary results on each of the campaigns.⁵
On December 14, the FCC will hold a (perfunctory) vote to kill the net neutrality rule, and SPL has agreed to release these preliminary charts in the interests of getting timely information out to the public. Here is their press release on the issue.
A couple more charts on response and bounceback rates.
Stay tuned — SPL will be issuing a preliminary report as more results gathered and analyses and conclusions are confirmed.
¹ My own opinion. Although I am a volunteer leading the data and survey effort for this project at Startup Policy Lab, I’m speaking on my blog as an individual.
² Not ashamed to say that I slept four hours in a nap room at Metis SF on a Sunday night.
³ SPL does some great work on civic tech and other issues at the intersection of law, policy and tech. It’s been tough pulling everything together on such short notice but they have been a steady ship in the storm. Sorry, Charles and Gina, for my frantic late night/early morning slack spam.
⁴ Not the most glamorous data science work, but sometimes old-school methods get the job done.
⁵ Campaign details and aggregate stats on this google spreadsheet.
⁶ I should add a huge caveat that this information is provided without a detailed write-up of methodology and assumptions. That said, I was shocked to see the the stark divide in aggregate stats between pro-repeal and pro-net neutrality campaigns. Our report will give our methodology and assumptions. We will publish the email survey text that we used and will also be publishing relevant code and de-identified data for public verification.
Create your free account to unlock your custom reading experience.