As Duncan notes, in his subsequent press release which is now on the link, the issue is whether the requisite economic data can be gathered in less intrusive ways (without presumably sacrificing in a material way accuracy or quality) then a census survey, that if you refuse to answer, results in a 5K fine potentially.
If there is, then it seems to me that Duncan has the upper hand in the balancing test between the needs for the polity to have the data for policy making purposes, and concerns about individual privacy and the burdens imposed on such individuals.
We seem to be putting up a lot of threads lately where there is more to the story than what the article source (or headline), or the spin placed thereupon here, would imply. I don't bother with most of them, but on this one, I thought I would dig a bit deeper since it all seemed so absurd, as to be implausible.
The reason Census data is preferable to privately collected data is that the mandatory nature of responding to it eliminates selection bias. When a telemarketer calls houses at random, they're only able to speak to those who are willing to speak to them and complete the interview. The potential for whether or not you respond being correlated with some unobserved factor makes using that data in meaningful ways problematic. It might be good enough to sell Snuggies. It's certainly not good enough to plan multi-billion dollar infrastructure projects or assess the impact of the Affordable Care Act.
The information you give out in the ACS is not "personal." Your name is not being attached to it. People aren't going to look your name up in some database, find out how many toilets you have in your house, where your address is and somehow use that information against you. It gets assigned an observation code and merges into the big sea of data that is incredibly useful to people in both the private and public sectors. I fail to see how that's any more intrusive or burdensome than the information your bank and your credit cards have about you (which is indeed personal).
