Strange stats in The Observer

There's a freebie magazine with The Observer today all about sex and attitudes to it. In the survey section there are odd results when people have been asked about the threat to children from paedophiles. Here they are -

Do you believe that children are more at risk from paedophiles than they were 10 years ago?

Yes: 80%
No: 20%

Twenty years ago?

Yes: 72%
No: 28%

Fifty years ago?

Yes: 66%
No: 34%

If I'm reading this right, doesn't this mean that people think that the past was riskier?

If you answer the question 'no' then surely you must think that in the period specificed the risk was the same s now, or perhaps even greater. So if we say the risk today is 'high' then just 20% of people think the risk 10 years ago was also 'high' (or greater), but 28% think the risk 20 years ago was at least the same, and if you look back 50 years a third of us think the risk was at least the same as today. So that suggests that people answering these questions think that the threat from paedophiles had been dropping since the war until a spike in the last decade (coinciding with the 'election' of the ZaNuLieBore junta!).

I say it's odd because I suspect people didn't think they were answering the questions in a way that gave that impression, they probably wanted to express their view that things had got riskier recently. Would be interesting to find out what order the questions were askd in to see if this made a difference, but more broadly I guess it demonatrates what a waste of space these types of surveys are.