Another detailed piece on this subject from the NYT suggests that combined printing and distribution costs of print books account for 12.5% of the average hardcover retail price.
-johnk
Hmm...wasn't too far off with my outdated cost figures, coarse rounding of numbers, and disregard for crucial details.

My example put the cost at $5/$30 for 16.66% vs NYT's 12.5% so my $5 cost should be closer to $3.75. Not too far off considering.

Quick question: where did these articles get their numbers? Because I'm beginning to wonder if they're just quoting each other considering how closely their figures match.
Since publishing houses are notoriously jealous about sharing their financial data, I'm assuming they got the $2 ballpark figure from a printing firm? We need to be careful if that's the case. The printing industry has a vested interest in not being painted as the bad guy in this discussion. So I'd be a bit cautious before I accepted such figures from someone "in the trade" as being gospel. Not to say they're wrong. I'm just sounding a note of caution.
I was the CFO at a printing company some years back. One thing you learn very quickly is that it's extremely difficult to accurately estimate "average" costs for a printed piece. Each press job is quoted separately since numerous cost variables in the four main operation areas (pre-press, press, bindery, shipping) have a distinct impact on the quoted price.
Interestingly enough, the press operation itself is the only part that really benefits from volume. Once you're on press (and absorbed the cost of the make-ready and wash-up) it's only press time, paper, and ink.
So the longer the press run, the lower the unit cost of each printed piece. At least up to a point. Ink and paper are relatively cheap. And today's presses can crank out an amazing number of pieces per hour. So many people are surprised to discover it costs very little extra to run 10,000 copies vs 5,000. The actual price of a 10K run will be higher. But it won't be anywhere near double the price of a 5K impression run, all other factors being equal.
Pre-press has probably benefited most from technology. The old days of "film to plate" are gone. And good riddance IMO.
Direct to press from digital is the way it's now done. Good because it eliminates a lot of the old "proofing" work a quality printing shop used to provide. Saves money big time. The unfortunate downside, however, is that there's no longer any content proofing. Whatever gets sent over gets printed. The huge number of typos that make their way into today's printed books is a symptom of that. This never would have been considered acceptable 20 years ago. But apparently the customers don't complain too much, so the publisher's (even the big ones that have the budgets and time to prevent it) are perfectly happy to ignore the problem.
Bindery operations are harder to cost out. Much of the process is sill labor intensive despite automation. And it also has the highest risk factor for a printer. Work spoiled during bindery operations (cutting, folding, binding, etc.) can force a job back on press if the delivery quantity can't be met. When I got out of printing, this was the area of the industry that was being most closely looked at for potential efficiency improvements and increased profitability. I'm sure many technical improvements were made. But from the looks of it, a lot of the cost savings were also gained by employing the age-old tactic of finding a 'sweat shop' somewhere subcontract the work out to.
I could go on and an...

Maybe I'd better spare you.
