I think our concern only comes in when there is some language that points to the potential interference of the executive in statistical methods.
How data is compiled, how and when it is released, and what methods are being used are quite different in my view from the basic question about what sort of information is needed. That is a very legitimate thing to express on behalf of the executive branch.
That sort of language, as far as I've seen at least, is what you'd find, say, in the Australian laws of statistics. There's some explicit indication that the minister can ask for certain statistical products to be developed, but there is nothing that says that he or she will tell the chief statistician how to come about producing that information. I think that's the crucial point.