A distribution is fair if everyone prefers his own outcome to that of anyone else. There is no 'envy' in the sense that every agent feels every other is worse off (according to his own system of preferences).
Thus, if I just sold you Manhattan for some glass beads, I don't envy you at all because Manhattan is full of ghosts while glass beads sure are shiny and you're a real sucker to have made the trade. However, as time goes by, and you just keep giving me more glass beads to prove the trade we made was Super-Fair, though I still keep taking your glass beads- coz damn them beads be shiny!- I still can't help feeling more and more envious and more and more of a shmuck and more and more hard done by, even if I believe them Manhattan ghosts gonna git you eventually.
To get round the problem of substantive interpersonal comparisons of Utility- which otherwise made it impossible to speak positively about fairness- a number of approaches were taken to show that Economists could still talk about fairness on the basis of purely bilateral (or indeed Chichilnisky type 'limited arbitrage') considerations and that this could enable Economics as a discipline to move forward by redefining 'fairness' in a manner internal to its own axiom system.
Baumol's classic work on Superfairness- i.e. positions in which all parties suffer zero envy even if something is substracted from them- sought to overcome the obvious objection (nevertheless made by Holcolmbe) that it is unfair to usurp the ordinary meaning of 'fairness' by proving the utility of just such a usurping language game in various topics in Economics- more especially models with individual but no aggregate risk, where it has been shown that ex post and ex ante equity can be satisfied such that envy free allocations exist.
In his book, Baumol defines a notion of incremental Superfairness but shows that allocatively efficient Superfair improvements can be unfair. Thus, for Economics, some scandal lurks for the notion of a golden path Fairness.
Can Tardean mimetics, to which Rene Girard gave a De Maistrean twist, come to the rescue? At first sight, no, not at all- the reverse is the case. We are envious of he whose example comes to be emulated- somehow he has shown his superiority and become the target of 'mimetic rivalry' for others, whereas we ourselves- even if our preferences and beliefs are unchanged- have become invisible or marginalized. More seriously, he can extract a rent or exploit an early adopter advantage while our ability to do the same thing, ceteris paribus, worsens day by day. Initially, it may be, mimetic rivalry causes the envious killing of the 'King' at the apex of Society's system of Tardean imitation, but soon enough a 'phramakos'- a scapegoat- is found from amongst the hoi polloi whose periodic sacrificial killing acts as a sort of innoculation against Violence becoming endemic.
Yet, for a De Maistrean reason, we may accept this dispensation as if not Fair, or even incrementally Superfair, yet Divinely Fair for Christ too was a King and we killed him and ate him and still get a little tipsy drinking his blood and dunno how Girard got all this out of reading Proust but he did okay and like his bestest bro was Mandelbrot so there's probably some heavy duty fractal Maths which shows like how all this is Kantian kosher and euvoluntary and shit.
This follows because in a De Maistrean Universe- as if by mathematical convention, or 'normalization'- there is no aggregate but only individual risk. Thus, glass beads are verily the ghosts of Manhattan and forever is their feast.