But their pricing model appears to be working most of the time with their market. Sleepers are running with a pretty high load factor, although that is not scientific, just my own observations, and anecdotal evidence. It would be interesting to see real data on that.
I don't have any data that shows what occupancy percentages were, although perhaps if I have some time later today I could calculate some based upon the numbers that I have and knowing how many sleepers per train there are. However, here are the ridership numbers for the last 7 years that show sleeper occupancy is indeed going ever higher, despite the higher prices. And of course we all know that on many trains it's impossible to book a sleeper at certain times of the year.
Note: This data is presented in Amtrak's fiscal year format, that being from October to September. So for example the first year of 2003 represents October 2002 through September 2003.
2003 - 611,308
2004 - 600,021
2005 - 605,010
2006 - 580,149
2007 - 591,023
2008 - 622,243
Note the drop in sleepers occupancy between 2003 & 2004 was due to the loss of several sleepers in various wrecks during that period. Back before all the wrecks including the big one on the AT, Amtrak had 175 sleepers, plus all trains still had dorm cars. Today there are only 160 sleepers listed as active, and no dorm cars on the single level trains which means that revenue rooms are lost to the crew. Note: I didn't include the Superliner Trans/Dorms in those counts, there are 41 of those cars in service both in 2003 and today. Most trains in 2003 did not sell rooms in the Trans/Dorm, whereas most do sell rooms today.
And the drop in sleepers between 2005 & 2006 was due to the cutting of the Three Rivers and its Heritage sleepers, and the loss of the Sunset east due to Katrina. Take out those anomalies and ridership has continued to increase over the years.