Head injuries when people are not wearing helmets are not rare at all. Even if they were, surely it is better to be safe rather than sorry, I cannot work out Steve's objection to helmets.
The 'science' tends to work like this:
There is evidence that a helmet can reduce the level of injury arising from a relatively low-impact blow to the head, such as falling from bike onto the floor. Depending on the specifics of the helmet design and the impact angle, however, there is a small chance of an increase in the level of injury, which increases further if the impact is not on a flat surface (such as the road) but on something angled (such as the lip of a car bonnet, or the edge of a kerb). Overall, though, a helmet appears to be generally of benefit as protection in the event that a low-level impact occurs.
For higher strength impacts there is very little evidence that wearing a helmet makes any overall difference to injury severity or rates.
There is evidence to show that other road users give less space to cyclists wearing helmets, and cyclists wearing helmets are prone to take more risks as a result of the sense of protection they derive from having the helmet, and so the likelihood of a low-impact injury is increased.
Different individual research examples put slightly different rates for all of these findings, and as such the extent of the error bars on any judgement is such that there's no clear evidence of which is more or less safe.
For children, where drivers are more likely to give adequate space and the mass of the rider makes a lower impact collision more likely, it's a little clearer that wearing a helmet is overall of benefit, but it's far from crystal clear even in those instances.
O.
