When they start cracking down on people who complain about the censors, it's time to make some noise.

There is a number commonly cited with regard to Internet censorship offenses in New Zealand. You may see it surfacing in debate on proposed expansions to the Censorship Act or restrictions on ISPs. This from Steve O'Brien of Internal Affairs' Censorship Compliance unit:

Since its inception in July 1996, the Department's Censorship Compliance Unit has investigated more than 500 cases of New Zealanders distributing and trading objectionable material, most of which is child pornography.

One hundred and nine individuals have been convicted in the Courts for offences relating to the trade and/or possession of objectionable material via the Internet, with another 20 cases pending.

A natural inference is that all of those 500 cases relate to actual detected offenses, presumably done by real offenders.

Thing is, it's not clear what that number represents. It certainly four or five times higher than the number of convictions. It's not even the number of prosecutions.

It seems to be the number of investigations which proceed beyond some particular point. And it includes an unclear number of people from whom there was not sufficient evidence to prosecute. Who are, in other words, either not legally guilty or just plain not guilty.

Even the police don't release statistics based on the numbers of people that they think committed crimes.

But this number seems to have been accepted by the Government Administration select committee (and by the media) as having some significance.

With me so far? Right.

Here's the other thing.

When Computerworld staffer Steve Bell wrote an article about this, with his estimates for how inflated the 'investigations' number is, it was supressed.

He writes about the affair in this piece (skip the overladen editorial prose and scroll down to 'Censorship in New Zealand') on an international site called Twentieth Century Inquisition:

I published a story on these figures in Computerworld, 1st November.

Though it reached print, it was pulled from the paper’s website after being up only a day, on the insistence of [Internal Affairs] ... I offered an amended version which retracted the rather nit-picking and questionable “errors”, while insisting on the accuracy of the main argument, but for DIA nothing but deletion of the complete story would do. They have not provided any reason for doubting the figures that lie at the centre of the article, but clearly it suited them to have these suppressed.

I guess that this is an amended version of the article. Dated November 3, it's still on the Computerworld site but impossible to find using the internal search engine. Google also has a cached version.

I realise I'm not in posession of every fact, but it looks to me like the Computerworld editor types caved under pressure. Let's all just take a moment to hope they're ashamed of themselves.

Anyway, Steve Bell has more to say. As an offended party he adds everything he can think of that's wrong with the web monitoring system (and takes action here). Worth a look.

And the central points stand. Firstly that the figure under discussion does not represent a number of people 'caught'. Bell calculates that up to two-thirds of them were cases of insufficient evidence (either of guilt or of any offense). He may not have factored in unsuccessful prosecutions - if so the ratio may be more like a half or one third. But that still reduces it by about 200.

They might, for example, be counting the 111 investigations of peer-to-peer traders, which resulted in 2 convictions (here, again).

Secondly, much as I would love the blame Phil Goff for this, it seems to be Internal Affairs' fault. At the very least they're letting people get the wrong idea. Bell has several instances of the Department either failing to clarify the issue or doing things likely to perpetuate it.

I myself note the Department announcing via one Russell Brown that it 'catches' an 'offender' every three to five days. Or a report on web censorship offenders that in 2004 was updated to contain social, demographic and behavioral information of 185 offenders. Which is more than they've actually had convictions.

The only explanations I can think of are that either they want people, including Ministers, to get the impression that the unit is finding more actual criminals than it actually is, or they're acting on the belief that everyone they investigate is guilty. I favour the latter explaination.

And then, when they get sprung on this, they try to get the article not simply corrected, but removed. I'm aware that this is what everyone tries to do when they get sprung and it's not their fault they got away with it. They only complained. It's not like they sent round a goon squad. But this is government we're talking about. The censors.

If we can't question the actions of the censors, we're in trouble. Child porn is a sensitive issue, and it should be. But no censorship should go unscrutinised. Particularly since the law in question is currently being strengthened, its agents should be open to scrutiny.