We could draw a bunch of conclusions that the data doesn't point at, sure. Show me the fatality rate per crash by vehicle manufacture year and I'm betting we'd see a steady trend downward.

Here's another graph that shows a change around 2010 that I could lay over the data that would correlate a rise in car sales with a rise in pedestrian deaths

And since the fatalities are in absolute numbers and not rate, the number of fatalities per car on the road might have been holding steady or going down. (Note that that graph has no sources, so you shouldn't take it at face value anyway)

Not trying to defend cars, just trying to call out massaging data to fit what one has already decided is true.