Hassan Masum, David Zaks, and Chad Monfreda provide a good introduction to the valuation problem in their latest WorldChanging post on "ecosystem goods and services". Here is how they sum it up:
Ecosystem Goods and Services Series: Valuation 101, February 12, 2007: … Monetary estimates are easy to comprehend, but hide assumptions, approximations, and simplifications. They're also valid only at particular scales, and measure only certain kinds of value. On the other hand, they integrate information about supply and demand, however imperfectly — something more data-oriented valuation such as indicators or ecosystem models don't do explicitly.
A marriage of the two may be the way forward. When monetary values are used, misleading precision should be eschewed, and model assumptions clearly stated and summarized up front. And where monetary values are not useful, we need to develop comprehensive models with understandable summary statistics and indicators, that can be used directly to make policy tradeoffs without going through the intermediary monetary stage.
It's easy and comfortable to add up costs and benefits in a single, seemingly universal framework, and then to compare net benefits of different avenues of action to choose which way to go. But sometimes costs and benefits can't be translated into single values — too much is lost in the transition.
A more nuanced idea of valuation will let us use the powerful tools of financial planning where appropriate. And where financial planning is inappropriate, we can use indicators, models, and simulation approaches to compare the benefits of alternative courses of action in terms more directly linked to human well-being. At the end of the day, the challenges in really comprehending the true value of resources or the environment (or of an idea or social capital) should be approached with humility…as a long process of advancement, as our knowledge, wisdom, and experience grow.