Fred99:
I agree with you on that.
There's a huge problem with (essentially subjective) "self assessment" used for collecting data. It's not just that publicity / media coverage of "issues" might influence an individual's response now - rendering it inaccurate as a "snapshot", it's also that publicity / media coverage is likely to change for any number of reasons between this "snapshot" and the next - so it's not going to be possible to make any honest conclusion about any trend, any need for future action, or to measure success of any actions underway now.
I can see the point you are making, however, I think in a pool of 3 million responses, you'd expect the majority to be honest in their responses and so long as the questions weren't too intrusive and were carefully worded, you could get some indications of some things perhaps.
Not sure how you would position this, but potentially asking about likelyhood of trying additional children, or asking people what aged they would aim to stop work. Do you intend on purchasing a investment property? Not exactly scientific, but could give an indication on intentions.