Thanks, MO, for accompanying grim NHTSA stats with some perspective from the AMA, rather than letting them go unchallenged as you have in the past.
Here's more perspective from a June 2001 NHTSA report, "Recent Trends in Fatal Motorcycle Crashes": In 1990, the Motorcycle Industry Council found that 3.7 million street motorcycles were "in use". That same year, NHTSA reported that 3244 riders died. Eight years later, motorcycles in use had climbed to 4.8 million, but deaths had dropped to 2294. In other words, during that span, 1990 to 1998, not only did total deaths drop 29% but the fatality rate dropped from 877 deaths per million bikes to 477. In other words, riding got safer by 46%.
What has happened since 1998 is a mystery, at least to me, because I can find no more current "motorcycles in use" data than 1998. This figure is absolutely crucial for determining whether the increase in fatalities is due to increased average risk or to an increase in the number of riders who choose to accept that risk.
If the AMA wants to make themselves useful in this debate, the can sponsor annual motorcycling demographic surveys to help counter the stream of bad news about motorcycling coming from NHTSA.
By the way, the 48% drop in deaths from 1990 to 1999 mentioned in your lead paragraph is wrong. The 1990 toll was 3244 and the 1999 toll 2472, a decrease of 24% (NHTSA, "Recent Trends in Fatal Motorcycle Crashes", June 2001, p9).