ISP Filtering: Evidence-based policy or policy-based evidence?

What constitutes evidence in policy-making? Is it a case of making policy based on evidence, or making evidence to support policy?

Obviously, an ISP filtering trial will provide evidence about the success/usefulness/cost etc of filtering Net content at the ISP. But what about the 'evidence' which suggests that Australians do not want mandatory ISP filtering?

The point I am making is that the use of the term 'evidence-based policy' is purely spin in this instance - in effect, it is policy-based evidence-making.

Why do we need evidence on a policy which nobody wants? The Minister for Broadband was provided with considerable 'evidence' on ABC’s Q&A recently: http://www.abc.net.au/tv/qanda/txt/s2521164.htm.

Indeed, the results of Netspace’s customer poll on compulsory ISP filtering were quite conclusive (with a sample size of 10,000).

But what 'evidence' do I need to suggest that ISP filtering is a case of putting the cart before the horse? The major problem for us lowly citizens is getting access to the evidence when even the federal opposition needs to use FOI laws to get access to the government’s information. What chance does the average citizen have?

I maintain that ISP filtering is a side-issue. Nonetheless, it is good to hear from the Minister for Broadband that the NBN will be rolled out on a fast-track plan where market failure in broadband access is occuring:

The ACCC has confirmed that places with only a single provider of backhaul services include: Geraldton in Western Australia, Mt Gambier in South Australia, Broken Hill in New South Wales, Mildura in Victoria, Mt Isa in Queensland, and Darwin in the Northern Territory.