My comment was directed toward economic forecast modeling -- this is what most people use to describe future funding requirements for any given gov program. Though the models have become incredibly more mathematically complex and are internally consistent, they still perform abysmally. Continuing research on them is a fine thing and will generate PhD's for the foreseeable future.

However, given the track record of economic modeling, I'll take the Black Eightball and save money.

Better still, use a general conceptual model. That is a useful approach and provides a fine contextualization for policy. Don't rely on a mathematical model. Everything that a conceptual model is, a mathematical model is not.