Making dynamic modeling effective in economics

Mathematics has been extremely effective in physics, but not in economics beyond finance. To establish economics as science we should follow the Galilean method and try to deduce mathematical models of markets from empirical data, as has been done for financial markets. Financial markets are nonstationary. This means that 'value' is subjective. Nonstationarity also means that the form of the noise in a market cannot be postulated a priori, but must be deduced from the empirical data. I discuss the essence of complexity in a market as unexpected events, and end with a biologically motivated speculation about market growth.