  Steve,
I think the economic theory crowd offer a more objective alternative to the reasonable person standard. They base their view largely on the brilliant formula developed by Judge Learned Hand in United States v. Carroll Towing Co. (1947). The formula is often presented as a way for jurors to think about the reasonable person, but it's seldom followed correctly.
Here's the formula: B < PL = injurer is negligent.
 B = defendant's cost burden for taking precautions
 P = probability that injury will occur
 L = the cost of the loss.
Take a generic product liability case for example. (B) Say it costs* Honda $50 to add a seatbelt (let's just say) to each car. (P) The probability of a fatal Honda accident that a seatbelt would have prevented is 1/10,000. (L) The cost of the fatality is $1,000,000 (I'm just keeping it simple!). So $1,000,000 > 50/.0001, and Honda is liable (or is considered "unreasonable").
It's an impressive formula because it tells us not just whether the defendant is liability but how much he should spend in the future to avoid being liable (i.e., to be considered "reasonable"). In this case, if the other figures hold, Honda would have to spend just under $100 per seatbelt to avoid being liable. Once the price of the seatbelt goes past $100, Honda should not be liable for not adding it. Adding that seatbelt would be excessively cautious or "inefficient," as economists like to say; it would effectively injur the injurer. (*we're really talking about marginal costs, not total costs here, but I don't feel like writing that out.)
Two caveats: First, we have actuarials to help us figure out these costs, some of them pretty exact, others rather flexible, which is where we would still probably need to call on a jury for help, even though their grasp of statistics and actuarial tables probably sucks to high heaven. Second, above is the simplest version of the Hand formula. I've seen way more complex variations.
Jordan

