NHTSA released their early numbers for 2013 fatalities and for the 3rd year in a row, the number of bicycle fatalities has gone up, though this time only by 9 fatalities (or 1.2%). The number stands at 743, but will likely be revised upward later (It usually is. I don't know why. Is it people who die later from their injuries or is it corrected reporting errors?) The estimated number of injuries on the other hand dropped back down to 48,000 same as it was in 2011.
US Pedalcyclist fatalities 1975-2013
It's important to note that this isn't a count of all bicyclist fatalities per year. It is only fatalities that occur on certain types of roads that involve cars. It does not include cyclists who die in single bike crashes; in crashes with pedestrians or other cyclists; on sidewalks, parking lots or bike paths or on federal roads like the GW Parkway. By my count (post forthcoming, so for now you'll just have to take my word for it) this data undercounts total cyclist fatalities by as much as 16%, and possibly more, meaning actual deaths are probably closer to 862.
No one took me up on my bet that the number of bike fatalities would come down this year, but if they had, they would have won (if only barely). I'd make the bet again though for next year. The trendline continues to point down, even as exposure continues to increase. Below I've used bike commute rates as a proxy for exposure since it's the best data available.
US Pedalcyclist Fatalities and Commute Rate 1975-2013
And of course, the commute rate is the percentage of Americans biking, not the number of Americans biking. If we look at the raw number of bike commuters from ACS data, the two trends look even more different
And if we plot the number of fatalities divided by the number of bike commuters to get a proxy for the fatality rate we see that biking is becoming safer over time, contrary to the reporting following the release of the GHSA report.
And finally, here's the trend for bicycling injuries, which also shows cycling getting safer.