Emails, Jose Gonzalez Altamirano, writer, Keep Austin Wonky blog, April 13, 2016
7:37 p.m.
"All of the data in this TNC and cab debate is so noisy. It's hard to say much with confidence."
"I understand what she is getting at but the share of drivers with incidents is not what the typical rider is thinking about when considering risk. It's fuzzy math. There are a lot of casual TNC drivers; it obfuscates the risk of a trip, of the actual service activity."
"The more relevant measure for risk assessment is the share of trips or or time spent in vehicle that lead to an incident. We also have to account for potential differences in the underlying populations. If almost everybody that takes a taxi is a businessperson going from or to the airport, there might be fewer drunk men and women relative to TNCs. If so, assessing the true risk of each service gets more complicated. And given that there are millions of trips for both services and a relatively small number of reported incidents, it's very hard to feel confident that there are any clear patterns."
"It's not a great situation for data-driven judgments. We have a vague, fleeting assessment of the number of TNC drivers. There hasn't been any new public taxi data in since last fall. Key metrics such as assault incidents and inappropriate touching in the vehicles relies is riddled with under-reporting. Our best public estimate of TNC service levels is based on bullet-point in a political ad that doesn’t explain what exactly constitutes a rider. And the ad is not even formally from Uber or Lyft. The PAC can say anything."
On Wed, Apr 13, 2016 at 7:07 PM, Julio Gonzalez Altamirano wrote:
1. The ad with the "200,000 riders" data point (take a screen shot, might disappear)
https://www.youtube.com/watch?v=NLsXwysXj-w&feature=youtu.be
2. My facebook post on the data (please make sure to read the footnotes)
3.The ABJ cab data