Methodological individualism has been a colossal failure.  
  
Take for example a recent fad in microeconomics -- game theory. Has   
anyone seen a convincing three person game? So although interesting   
logical puzzles can be solved, little of general practical application   
can arise.  
  
Although applied economists may genuflect towards the method, they don't   
use it. They use stochastic methods.  
  
Even the simple summing of individual demand curves to achieve market   
demand, taught in introductory economics, can only be done with   
extremely restrictive and unstated assumptions. i.e., that no individual   
considers another's preferences or welfare.  
  
Rod Hay