That comment paper isn't actually debunking that specific survey. It's just essentially saying that proving that voter ID laws suppress minority votes is hard.
The phone survey of less than a 1,000 people is still a perfectly valid piece of evidence. The methodology wasn't obviously flawed and the delta was so large that the sample size was more than enough.
Here's another survey from 2012 that shows a similar number.
Sure as your comment paper points out, you can't use nation wide data to prove state level causation, but there's still good evidence that minorities are much less likely to have government ID.
It's not a shocking result either. We know that not having ID correlates with low income, and minority status also correlates with low income (on average, sure there are some minorities that actually have higher income).
The phone survey of less than a 1,000 people is still a perfectly valid piece of evidence. The methodology wasn't obviously flawed and the delta was so large that the sample size was more than enough.
Here's another survey from 2012 that shows a similar number.
http://www.projectvote.org/wp-content/uploads/2015/06/AMERIC...
Sure as your comment paper points out, you can't use nation wide data to prove state level causation, but there's still good evidence that minorities are much less likely to have government ID.
It's not a shocking result either. We know that not having ID correlates with low income, and minority status also correlates with low income (on average, sure there are some minorities that actually have higher income).