You might be interested in this book [0]. One idea from the author is that in astronomy, you model the behavior of a set of objects, where each individual object or "atom" of your simulation (planets, rocks) has a behavior simple and easy to model. In economics and social science, the "atoms" of your simulation are human beings, and there is no good intrinsic model for the behavior of a single human beeing.

One of the observation of the author is that models work better if the human beings accept to restrict their behavior and follow simple rules. Another observation is that in social science, it is hard to design models that do a prediction significantly better than "tomorrow will behave the same as yesterday".

[0]: http://www.ens-lyon.fr/en/article/research/your-life-numbers...