Choosing Where to Eat and Vacation

It struck me during lunch yesterday that despite many signs that people are generally risk averse (such as their behavior on Deal or No Deal), when making non-monetary choices it seems that people are quite risk-loving. Take, for example, where to have lunch (or what to eat once you've chosen a restaurant). Living in Manhattan with so many restaurants around, I've seen that many people have strict preference to eat somewhere they haven't eaten before; this exploration is all of the fun to them. But why? It's not in the hopes of finding someplace they love and can keep frequenting- they'd rarely come back anyway. They may well know that the restaurant they went to last month has a great dish that they loved, but they'd still rather eat somewhere new. In a classical utility framework, this implies a very, very convex utility curve over meal quality. (I should point out that I am not like this; after a short time in the city, I had identified several places I like to eat. I'd be happy to go to those restaurants all the time, while only occasionally trying something new, in the hopes of identifying another place I could add to my short-list. And, of course, once in one of these restaurants, there's only one or two things I'd order...)

Of course, the monetary stakes are pretty low here; even in the worst case, you've lost only a few bucks and had a less than satisfying meal. Plus, you'll have another meal in a few hours so it's really no big deal. But this behavior also holds with bigger choices, such as the choice of where to vacation. People generally only have the chance to take a vacation a couple of times a year, and the cost of a vacation can be very high. So, you eventually pick somewhere, go there, and have a great time. The beaches are fantastic, the food is good, stuff is cheap, the locals are nice. You have a great time. When do you go back? Probably never. Despite the fact that the cost of a vacation is very high, both in terms of having such a limited number of them and the monetary expense, people tend to keep trying new places.

I guess one utility-theory explanation is that there is, in fact, very diminishing marginal utility to going to a restaurant or a Caribbean island. (And thus, you get a story that the utility function is really concave in that dimension.) Once you've been, it's not nearly as fun the second time. Maybe this is so; but I know that I would have a fantastic time at restaurant X or vacation destination Y. So why not go back?


UPDATE: In comments, Bryce points to his blog. There, he makes a great point that I left out: the creation of memories, which (a) make these things durable goods and (b) gives a very strong argument for first experiences being much more valuable. Presumably, even the memory of a terrible meal has a lot of value. This speaks to the diminishing marginal utility of an experience as the rationale for "experimentation."

I'm not as sure of this logic when it comes to a vacation though. I wonder, are people that return to a vacation spot more likely to have done it because they liked the place or because they didn't do something the first time?