A far simpler observation is that Australian cities and populations have grown, and with them suburban sprawl, roads and commute time. Families are pushed furth

A far simpler observation is that Australian cities and populations have grown, and with them suburban sprawl, roads and commute time. Families are pushed further and further out in order to afford homes, where there are few cycling facilities (or many other facilities worth cycling to), and cycling on the road here is a borderline suicide pact. As you allude, pedestrian safety has likely improved as pedestrians simply disappear from the increased number of roads, and in those few inner city areas where there is pedestrian density, this has pushed back both against cars and bicycles. In my neighborhood and city for example, speed limits are dropping and cars are being excluded from some streets. Driver safety has likely improved due to concerted efforts, car improvements and everyone increasingly being stuck in gridlock. Furthermore, if this theory is correct, it may also explain increase cycling fatality rates, as those remaining cyclists are presumably cycling further and sharing roads with traffic. Indeed, the best place to look to try to figure this out in Australian terms is probably Canberra. It's the only vaguely populated Australian city with relatively small commutes and a complex network of actual bicycle paths. But as far as I know, Canberra has had relatively steady bicycle use, and relatively high rates amongst children. You would also need to explain why your hypothesised cause results in a steady continued decline, as opposed to a sudden once-off effect or rate shift as often happens with many such policy implementations, whereas my explanation does fit such data (assuming that the data is as you say). In short, I just don't think the evidence for your hypothesis that it's the helmet laws as opposed to continued geographic/demographic shift holds up under scrutiny.