These are valid questions, but I doubt anyone here will be able to answer them. I know I can't.

I can definitely see how it would be pointless it being voluntary is though. Either have it or not - no point in doing something that won't tell you anything or is open to manipulation.

Yes they could.

But with a mandatory one, those promoting certain answers to support their agenda would be outnumbered and overall results would be less likely to become distorted.

Well, the mandatory long survey was distributed to 20% of Canadian households. The new voluntary survey will be distributed to 33% of Canadian households. I will assume that the makers of the survey will know who these surveys were sent to. They have to come up with a mailing address to send them to.

You need to assume 5 things

1) people will want to form groups in order to skew the survey
2) people will find and devise a plan to do this.
3) the numbers in those groups will be large enough to matter statistically
4) any successful result that a large number of people operating with a purpose to skew data will not be discovered.
5) the skewed data without detection will create a desired effect on government policy.

I think they already exist.

I did actually have mandatory for all in mind, when I made my comment.

It's not quite true that that sort of discrimination is not allowed. The most ludicrous job I've had was "shadowing" an employee of a major utility engaged in the generation of electricy in Ontario. Shadowing is when someone incompetent is on the payroll for reasons unrelated to work, the job still has to be done so an outsider is contracted to do the work. I did it full time for fifteen years or so and part time for another several, I've lost count.

How did this come about? It happened because, when the Rae government came to power it thought the employee base at the utility was inappropriate, 'most everyone was from Scotland. It issued an edict that the staff should more closely reflect the demographics of the province. Illogical as ever, the management picked up the census and set about matching it literally; n% Poles, n% Italians, n% Chinese and so on. At the time I knew an Aboriginal guy who matched a needed skillset but couldn't be hired because that quota was full; they had him on a shadowing contract for five years or better. The place is still stuffed with people who lack merit, though many have since retired.

This was problematic but was, at least, an attempt to do something to break an ethnic monopoly and it was based on real figures. The problem with basing such lunacy on a voluntary survey is that the data will likely be inaccurate; people of specific ethnic groups will be more likely to respond either because they're culturally inclined to fill in forms or because they're organized to complete this specific form. That minority will then be overrepresented where the government has influence in hiring or in the provision of services such as education.

You might, of course, say that the answer is that the government shouldn't meddle in hiring, like that's going to happen. Something very similar happens in many countries, for example, we had to engage a "local partner" in Malaysia (really just a bribe) and, when dealing with local government in a US state with a significant Spanish speaking population we were in danger of losing the sale because we lacked Hispanic staff (like there are a lot of Spanish programmers in Toronto). This last mess was resolved by signing up the then manager of a bar known to you, he's Columbian and has lots of mates who belong to the appropriate culture.

None of this is discrimination based on perceived religion, that seems surreal to me, but given the tangled ethnicity of most people here, it's almost as bizarre; how many half-British, half-Swiss lifeguards should the City of Toronto have?

I'd imagine so. The Charter of Rights and Freedoms is quite hot on equality.

Of the 33% surveyed I wonder what the take up rate will be if its not mandatory? My money, based on observed human nature, is on far less then 60%, which means a smaller sample size, which is never good for data accuracy

People dont need to form organised groups to skew the data, the mere fact that different ethnic groups and social/ cultural groups will have different take up rates skews the data returned on its own, before even getting into special interest groups going out of their way to skew it. For example metis and first nations groups are historically far less likely to volunteer information, so that compromises the data from the start.

The larger picture here is that stats canada is the body that exists to supports the government on all matters statistics related. Why dont the government trust the advice that this body is giving them on how best to gather the data? Is a few people complaining really reason enough to ignore the expert advice of the organisation that deal with statistics? Why is the government hell bent on ignoring the advice of virtually all the statisticians on this subject. On top of that, many premiers and mayors dont think it a good idea, even the governor of the bank of canada is against it!

In 1985, the government established the National Statistics Council to consult with stats canada. Its formal mandate is very brief: it is to "advise the Chief Statistician in setting priorities and rationalizing Statistics Canada programs". In line with other aspects of Canadian policy in relation to statistical activities, a careful balance was attempted between policy relevance and professional independence: members of the Council are appointed by the Minister, but their duty is to advise the Chief Statistician. Its interesting that the minister in this case didnt see fit to run this idea past them before it was implemented


Something in all this stinks, it wont save money, it wont gain better data, and they are ignoring or bypassing their own expert advisors!

Tony Clement says it will address the intrusiveness and compulsory nature of the data collection. But all the data is anonymous, so there really is no intrusion (inconvenience yes, but not intrusion), and the census is still compulsory anyway so Im not at all sure thats true. It seem to me that what he is doing is saving some people the trouble of filling in a form they dont want to and proving to the public servants that he's in charge, and never mind what they think.

Thats objectively false, differential take up rates in different communities can easily be normalized against, and special interest groups deliberately seeking to skew data will never have an impact greater than they're numbers allow.

Ok, assuming that is true, how does that work in practice? I ask not to be pedantic, only out of curiosity. I have worked at 3 different places here in Canada, not once have I had to complete the sort of ethnic profiling, religious pigeon holing questionnaire that is law back home. I could be a devout follower of any number of random cults, none of these people would ever know that.

The only way to have effective employment equality is to have mandatory profiling upon recruitment, information which is then passed to the powers that be. I have seen no evidence of this here.

How are you going to do that if the data collected is anonymous?

The only way to be sure of take up rates among different groups is to have a widespread sample that takes in all groups, and to know that the take up rate is 100% because its compulsory. Kind of like what the majority of statisticians recommend. What they do now in fact...


As for special interest groups, the likely effect is largely dependent on the take up rate. If everyone fills out the form anyway it doesnt matter. If only 1000 people fill out the form, and 900 of them have an axe to grind, then it has potential to have an effect.

Why is the government hell bent on ignoring its own advisers and why didnt it even consult with the national council on statistics about this.

I do not know the answers to your question, so I won't attempt to answer it.

I'm rather glad I started this thread. It seems to have got brains working.

Hmmmm...Government ignoring its own advisors...where have I heard that before?

This is no surprise for this particular government.

I would like to know what is really driving this.

Not really. The Charter is something that Plaintiffs can rely upon in support of their claims. They would claim breach of the Charter and evidence would have to be produced by both sides for the Court to be in a position to come to a conclusion.

For example, a Muslim states that his employer does not allow him to pray and attend religious ceremonies during work hours, but allows such practices for Catholics.

Saying that the company does not employ enough Muslims is unlikely to get the Plaintiff anywhere, unless, s/he was able to show that the employer had an unreasonalbe aversion to Muslims - again, this would have to be proven on the appropriate burden of proof by evidence.

mandatory vs voluntary. The critical info is already known to the gov't. When we start counting skin shades and reacting to that, then we have already failed.