Models. They work on things that have great physical rigor.

Having worked with the results of environmental models and modelers over the years, being told that a model provides a certain output makes me less likely to buy-in.

My view of modeling is wonderfully expressed in a book entitled, "Useless Arithmetic", http://www.amazon.com/Useless-Arithmetic-Environmental-Scientists-Predict/dp/0231132123

When economics can predict a singular event, I'll listen to economic models. Until then, I consider them marvelously complex art.