As I pointed out before, Edwards is basically saying (in reverse order) that:
Some believe the Census Bureau does too much already. “They waste a share of their budget on studies that no one actually uses,” says Chris Edwards, an economist with the Cato Institute, who cites periodic surveys on such items as the total hog count in the U.S. to prove his point. “A lot of that could be done by the private sector.”
1. the private sector could perform major functions of the Census (which is demonstrably wrong since, as I pointed out before, the private sector has tried to do this repeatedly without ever coming close to matching the quality of government data) ;
2. no one uses reliable data about American agriculture (which is too laughable to waste any time addressing).
What's interesting and depressing here is not just the bad arguments that Edwards made but the valid ones that he didn't (or at least that he didn't make forcefully enough to be quoted). To understand these other arguments, it's useful to think of a simple value function for evaluating government projects
V = Returns - Traditional Costs - Libertarian Costs
Traditional costs are what you normally think of for a project, direct expense, opportunity costs, negative impact on other economic activities. Libertarian costs are the losses of liberty that go with any action where the majority forces the rest of society to take collective action.
Most of us don't give a lot of thought to LC but it's not zero and libertarians are performing a valuable service when they bring it up. In this case Edwards could have made the following valid arguments:
We're financing the census by taxing people who in some cases object to it.
The census is an invasion of privacy by the government.
A relatively small group gets a disproportionate share of the benefits.
I don't happen to agree with those arguments but they are valid and it's worth noting that Edwards chose to use invalid ones instead. We've seen this sort of thing before -- libertarian groups like Cato pushing flawed reasoning rather than make the honest but difficult-to-argue libertarian case. It's a practice that undermines their credibility.
Or it would if anyone cared about credibility anymore.