I found this article aptly put my larger concerns about the census into a neat package. However I feel we need to stress the unreliability of much of the census data. Filling the census in for my mother - who couldn't understand how to herself, yet another problem - I witnessed a few issues with privacy. She was loathe to put up information that might get her in trouble with the tax office - legitimately or not, this it is a concern when people are afraid to communicate the truth to their Government.
Further, the questions on identity were strange; notably your "ancestry". This question is just so vague it demeans the notion of personal identity when some people will take the piss out of the question. I could put down anything. Chinese, Native American, French, Yugoslavian for crying out loud!

There may be merit in changing the way we collect data, and the data we collect. But one of the most important benefits of the census is that it gives us a long term picture of trends over time. I would be very sorry to see this lost.

Rhian, I agree that following long-term trends are important, but as if we cannot be clear on what these long-term trends are, is there any point in following them? I don't suggest doing away with population surveys altogether, far from it. But surely a sample of 100% (when as Isabella has pointed out, many people lie anyway) is unnecessary?

Many long-term trends are fairly transparent, like the aging population or geographic shifts in population. Some data that may not seem important now could be significant in future.

Perhaps we don't need 100% sample, though it does give a lot more accuracy when we want to break info down into small samples such as a postcodes. It's also important for collecting data on people who often slip below the radar of more mainstream data collections, for example indigenous people or fly-in fly-out workers.

You make some good points but it is not clear to me how the issue of a reluctance to ask some types of questions is going to be solved by moving away from a universal census - how will changing the method of data collection affect the willingness to pose certain types of questions?

I found the questions nosy and impudent. These are the people who privacy legislation, and here they want to know every kind of personal and private detail.

Their claim that it's the only practical way to gather information is untrue. Imagine if the local video shop, or any private business, sent around a big nosy questionnaire prying into every aspect of your private life, saying it's the only practical way *for them* to gather information. The only appropriate response, even if it were true, would be "So what?".

But of course it's not true for any service that's voluntarily funded. It's only true for government because no-one voluntarily pays for their "services".

Also, have you noticed with government how they keep asking the same questions, you keep answering them, and they keep not knowing the answers? I don't know how many times in life I have filled out a government form asking me for my name, address, date of birth etc. But when I buy something from Amazon, I only have to tell them ONCE. It's like dealing with real live human beings, instead of arrogant ignoramuses.

Even the request to fill out the census form was in bureaucratese. "Your co-operation is sought". That's nice, isn't it? By whom? The Flying Spaghetti Monster?

I used to follow the brainwashed line that providing census information is a good thing for "us". This makes no sense unless you are one of the people who get your money from the government. For everyone else, they intend to use all and any information against you, so they can live at your expense. It's not in your interest to tell them anything.

So I didn't fill it in this time. Just told the itinerant bureaucrat who came to collect the form that we weren't there on census night. No worries. She didn't even collect the form. So I will never fill in a census form again, and I urge everyone else to do the same.