A brief, but interesting introductory article on the types of accounting rules that are legal but which promote bad investment in the long term appears here (pdf).   The summary, in the form of a joke, follows:


 Three accountants interview for a job.  When asked to evaluate 2 plus 2, the first responds “four,” and the second, thinking it’s a trick question, says “five.” But the third, whom they hire on the spot, says: “what do you want it to be?” I wasn’t laughing, however, when I learned that the behavior of candidate three lies within generally accepted accounting practice.


The article provides an example of the “trouble [that] can arise when single numbers are substituted for distributions.”  But I think there is likely to be a more general explaination of the problem.  Also, as the article sights there is the general problem of: 


  …But when they need to estimate the NPV of a project for an accountant, they will inevitably be asked for a number.


I can’t help but think this is an example of the ‘bad math’ that Chris Macrae keeps mentioning in his more lucid moments (1, 2, 3) when he say, for example here:

about 10 year ago … in big management consultancies through the 90s I was appalled at a maths error that was systematically devaluing trust, rewarding those who imaged over reality including conflict-makers and short-term speculators -the so called Unseen Wealth Intangibles crisis as it was then called, the Inconvenient Truth crisis as some sustainability mapmakers now call it. (typos corrected from the original quote)

Somewhere in all of this there is a lesson to be learnt from economics and the need to look beyond the short-term to the long-term impacts of the rules within systems.