Do you have a source for this HJ? It would be usefull to bookmark a link for future use of that kind of statistic.
I can't remember all of them but here are a few:
Rumar (1985) using British and American data (quoted by Tom Vanderbilt) 95% of crashes
Lum & Reagan (1995). "Interactive Highway Safety Design Model: Accident Predictive Module". Public Roads Magazine. 93% of crashes
There was a
DfT report a couple years ago, carried out by
RTL which gave the 87% figure. The surprise with that study was how few
accident crashes, involving cyclist and drivers, were attributed to the cyclist, only about 5%. It also showed that RLJing cyclist are involved in very few crashes (not that that is a justification for RLJing).
There are other studies as well.
It is also worth noting that an
RAC survey found most British drivers who'd been in a crash did not believe themselves to be at fault. No great surprise there...