Dharma and Fairness
It is common knowledge that almost all social upheavals around the world have been a fight for fairness. Or were they really about fairness?
To answer this of course, we need to define what is fairness. Unfortunately, this is where things start going out of hand. In my class on negotiation theory we study at least six definitions of fairness -- many of them contradicting one another!
Consider this example. Suppose some people agree to meet up for something. And they all agree that they will meet in the house of one of the persons. It is clearly unfair, right? Everybody else, except the guy at whose house the meeting is to be held, needs to commute in traffic, while the lone guy gets to relax at his house waiting for others to arrive.
However, when we were students pursuing our theses, we routinely went to meet our professors at their house to discuss our research on holidays and it never occurred to us that it was unfair. No, it was not because there was a "hierarchy" with professors bullying students (I studied in post unification Germany which was very keen to attract students, and the professors were not only inspiring, but also "chilled out" -- if that's the right word).
The reason we found the setup fair was that as students we had a singular goal -- to make progress towards our thesis. And commuting several miles for a meeting was just a small cost towards the benefit of making progress on our theses. While the professor had several students and had several goals to pursue -- not just research goals, but also administrative and financial goals to keep the department and lab running. The intrinsic cost that he would have to pay to commute through traffic for meeting his students would be much higher.
The thing here to note is that utility and cost have subjective elements, even though there is a price tag associated with the object. If a kg of tomatoes cost Rs. 100/- (Ha!) and I buy tomatoes by paying Rs. 100/- what it means is not that the value of a kg of tomatoes is Rs. 100/-. What it means is, for the buyer, a kg of tomatoes is more valuable than Rs. 100/- while for the seller, Rs. 100/- is more valuable than a kg of tomatoes.
Hence, a system of fairness based on objective valuation may not actually be considered fair.
Here is another example connotation of fairness -- the property of Pareto optimality. A system comprising of multiple rational agents is said to be in a state of Pareto optimality, if no agent can change what they are doing, to get a better utility, without hurting the utility of some other agent.
A queue for example, is in a state of Pareto optimality. People standing in a queue can choose to either remain in the queue, or cut the queue and go straight to the counter to get for themselves a better payoff (lesser waiting time). However, while they get a better payoff, all others in front of them in the queue would be worse off by this action of theirs.
Pareto optimality, is hence seen as yet another example of fairness.
However, consider this example. A society that practices slavery is also in a state of Pareto optimality. And if Pareto optimality were to be the measure of fairness, then slaves seeking freedom, or workers seeking better and more human working conditions, would hurt the prospects for their masters or for the management, making it unfair.
So when is Pareto optimality fair and when is it unfair?
Consider yet another example -- a game called the "Battle of the sexes".
A couple wish to go on a date and they have between them two options -- a musical concert or a cricket match. The boy likes to go to the musical concert, while the girl likes to go to the cricket match (of course!). If they both decide to go to the musical concert, then the boy would have "won" the battle -- not only are they going on a date, they are going to his preferred choice. If instead they choose to go to the cricket match, the girl would have "won" the battle. They of course, have a third choice -- to call off the date and go on to the concert or the cricket match separately. In which case, they are both equal -- but they are not on a date, which is what had started the whole exercise.
The above is an example of a choice between a system state that is "equally poor" or "unequally rich", where the poorest in the unequally rich state is richer than the richest in the equally poor state.
So in this case, do we favour equality over collective wealth, or collective wealth over equality? (The answer is not that simple -- what if unequally rich state, the richest was orders and orders of magnitude more richer than the poorest?)
I can give several more examples of fairness, all of which have a "Yes but.." exception, where the very definition of fairness can be used to create a system that is blatantly and visibly unfair.
To answer this of course, we need to define what is fairness. Unfortunately, this is where things start going out of hand. In my class on negotiation theory we study at least six definitions of fairness -- many of them contradicting one another!
Consider this example. Suppose some people agree to meet up for something. And they all agree that they will meet in the house of one of the persons. It is clearly unfair, right? Everybody else, except the guy at whose house the meeting is to be held, needs to commute in traffic, while the lone guy gets to relax at his house waiting for others to arrive.
However, when we were students pursuing our theses, we routinely went to meet our professors at their house to discuss our research on holidays and it never occurred to us that it was unfair. No, it was not because there was a "hierarchy" with professors bullying students (I studied in post unification Germany which was very keen to attract students, and the professors were not only inspiring, but also "chilled out" -- if that's the right word).
The reason we found the setup fair was that as students we had a singular goal -- to make progress towards our thesis. And commuting several miles for a meeting was just a small cost towards the benefit of making progress on our theses. While the professor had several students and had several goals to pursue -- not just research goals, but also administrative and financial goals to keep the department and lab running. The intrinsic cost that he would have to pay to commute through traffic for meeting his students would be much higher.
The thing here to note is that utility and cost have subjective elements, even though there is a price tag associated with the object. If a kg of tomatoes cost Rs. 100/- (Ha!) and I buy tomatoes by paying Rs. 100/- what it means is not that the value of a kg of tomatoes is Rs. 100/-. What it means is, for the buyer, a kg of tomatoes is more valuable than Rs. 100/- while for the seller, Rs. 100/- is more valuable than a kg of tomatoes.
Hence, a system of fairness based on objective valuation may not actually be considered fair.
Here is another example connotation of fairness -- the property of Pareto optimality. A system comprising of multiple rational agents is said to be in a state of Pareto optimality, if no agent can change what they are doing, to get a better utility, without hurting the utility of some other agent.
A queue for example, is in a state of Pareto optimality. People standing in a queue can choose to either remain in the queue, or cut the queue and go straight to the counter to get for themselves a better payoff (lesser waiting time). However, while they get a better payoff, all others in front of them in the queue would be worse off by this action of theirs.
Pareto optimality, is hence seen as yet another example of fairness.
However, consider this example. A society that practices slavery is also in a state of Pareto optimality. And if Pareto optimality were to be the measure of fairness, then slaves seeking freedom, or workers seeking better and more human working conditions, would hurt the prospects for their masters or for the management, making it unfair.
So when is Pareto optimality fair and when is it unfair?
Consider yet another example -- a game called the "Battle of the sexes".
A couple wish to go on a date and they have between them two options -- a musical concert or a cricket match. The boy likes to go to the musical concert, while the girl likes to go to the cricket match (of course!). If they both decide to go to the musical concert, then the boy would have "won" the battle -- not only are they going on a date, they are going to his preferred choice. If instead they choose to go to the cricket match, the girl would have "won" the battle. They of course, have a third choice -- to call off the date and go on to the concert or the cricket match separately. In which case, they are both equal -- but they are not on a date, which is what had started the whole exercise.
The above is an example of a choice between a system state that is "equally poor" or "unequally rich", where the poorest in the unequally rich state is richer than the richest in the equally poor state.
So in this case, do we favour equality over collective wealth, or collective wealth over equality? (The answer is not that simple -- what if unequally rich state, the richest was orders and orders of magnitude more richer than the poorest?)
I can give several more examples of fairness, all of which have a "Yes but.." exception, where the very definition of fairness can be used to create a system that is blatantly and visibly unfair.
*~*~*~*~*
You guessed it right if you are thinking that fairness cannot be defined only in terms of payoffs of the players involved. There is more to the definition of fairness than just the self-interest functions of the players.
And that is the element of sustainability -- that dharma thing again!
Look back at all the social upheavals of history. Were they really about fairness, or were they about sustainability?
Slavery -- or the trading of our liberty for some concrete benefit -- has existed for millenia and it was also rationalized away by weird logic. A weak person for instance, had two choices -- struggle for survival or become a slave of a stronger person and make the other guy stronger, so that he can take care of both of them. Trading of one's freedom for a life of safety was seen as a most rational thing to do.
Except, beyond a certain point the configuration becomes unsustainable.
We trade our freedom for safety or convenience all the time. Be it using Gmail and telling google all about ourselves, or passing through an X-ray scanner in airport security and answering embarrassing questions about the contents of our bags to airport security.
As long as the extent to which we give up our liberty is bounded (by place, time and type of liberty) it is still fine. But when this trade becomes unbounded, we get into systemic stability issues.
Which is what is the core issue concerning privacy in the digital age. The question of privacy is not about fairness per se. The argument for fairness can be countered by several examples where we voluntarily give up our personal information.
The core issue is of sustainability of basic human values and dignity in a system where every information about them can be recorded in high-fidelity forever.
*~*~*~*~*~*
So, let me say this again. Unless we develop a comprehensive "Theory of Being" we cannot hope to find real solutions to the digitally connected world of the 21st century. We will just keep harping on fairness rhetoric without gaining any fundamental insight about what is the real problem.
Comments