The Search for a Fair Society: The Difference b/w a Service Contract & a Creation Contract or Why a Meritocracy Isn’t Fair

October 7. What a hard day. Well, grab a coffee and let’s talk about a realistic utopia of a fair society because without a clearly formulated goal you never get where you want to.

So, Rawls’ argues for a standard of fair equality of opportunity: an equal chance for all to develop natural skills and abilities because a formal equality of opportunity – the idea that jobs and other positions should go to the most qualified candidate – alone will not ensure equality of outcome. His argument is supported by the motivation every human alive under the sun ever had to contemplate cooperation with other humans; something that requires of the participants to limit their freedom to self-generate their model of the world, their reality, as they seem best suitable for themselves and align their individual models by using a shared dataset. They could survive alone. They do it because they want a chance to realize the high payoff random searches offer, the highest possible payoff of any kind of search. Cooperating offsets the risk that an individual alone can’t take.

So far, so wonderful. After reading all that you might conclude that a meritocracy then fits best as a model of a fair society. Those who are at the same level of talent and ability, and have the same willingness to use them, have the same prospects of success regardless of their initial place in the society, and after that success and influence depend strictly on merit i.e. intelligence paired with effort. But that’s not the case.

Why? When the old aristocracy is replaced by a new aristocracy of ‘merit’, those at the bottom are condemned to suffer the double indignity not only of being poor but of being looked down on and being held entirely responsible for their low status. In the end the privileged elite would pull away from the rest, and the least well-off would find it impossible to maintain self-respect. The hubris of society’s educated elite paired with the humiliation of the left-behinds would be a dangerous, explosive cocktail of discontent. And it would be again a situation in which not all get out of the cooperation what made them join the group in the first place. Why then, when the contract doesn’t make any sense for them as it only puts burden on them, should they continue to fulfill their responsibilities that the shared dataset puts on them? A shared dataset that they have lost any say over as it contains a stipulation that not just success but also influence depends strictly on merit. The only way they will act as if they use it in their individual world generation processes will be by force or the threat thereof. And that’s the clearest sign that society is most definitely not fair as a fair society doesn’t need violence or the threat thereof to enforce its structures and rules, the dataset it is based on.

In the end, this again goes back to the idea that society is a dataset, a social contract that people agree to not for survival but to offset the too high risk random searches pose for an individual. It’s a contract about actions taken by the members and what should happen in case the high payoff of random searches enabled by the contract materializes for one of the members. It’s not a contract in which the members ever promise success. In modern law this is the difference between a service contract, in which a party promises its services and gets paid irrespective of the success, and a creation contract, in which one party commits to creating something and is only paid after inspection and approval. A fair society is a service contract which the implementation of meritocracy would turn into a creation contract. And that would make it for most people impossible to fulfill their obligations from the outset. Out of no blame of their own.

Why? Because people only ‘deserve’ or can be blamed for what they are responsible for. But we are all the result of a natural lottery. Talent just as class, race, gender, health etc. is nothing we have any control over. Thus, no result, even when we would all start from a truly equal playing field, could ever be justly called ‘deserved’ or could be blamable. Therefore, it mustn’t have any influence over our lives or determine how much influence we are given over the dataset that allows and commits us to cooperation.

So then, is the solution perfect equality of outcome? No, because there need to exist incentives that encourage people to not just develop and apply their talents and skills in any way that fits them – what they are free to do yet will not entitle them to more than the basic support – but to develop and apply their talents and skills to the good of the group. In the end society, our willing cooperation, is an exercise in harnessing diversity. It’s all about keeping the balance between aligning the self-generated models of the world of the members to the point the pursuit level of cooperation is possible and keeping these models diverse enough that the individual searches but also the cumulated search of the group remains random. And that’s where Rawls’ Difference Principle comes in. But that’s for future posts.

Any thoughts? Tell me. Tell all. Since our models of the world change with every new information we gather and the models are only accessible to others when we communicate them, we are part of never-ending negotiations that require constant conversation and debate.

To watch this post as a video, go here.

#science #history #reality #society #philosophy #WorldGeneration #fairness #information #Rawls #OriginalPosition #equality #mind #self #brain #thinking #exploring

Previous
Previous

The Search for a Fair Society: Rawls’ Difference Principle ft. Self-Generated Models of the World

Next
Next

The Search for a Fair Society: Making the Case for Fair Equality of Opportunity ft. Self-Generation of Our Reality