It seems like antibiotics like Colistin are uncontrolled in China, and perhaps very cheap? It seems like madness to use these last resort drugs on farm animals.
In particular, it does not explain why people perceive their self interest the way they do and it does not explain how those perceptions change. It does not explain how people come to find common interest with others and does not explain how someone's notion of self and notion of society commingle and interact. These things are the definitions that let you talk about rational behavior in the first place.
This is by no means a defect of the concept. A microscope is not defective just because it is not a telescope, but it is still just a microscope. If you magnify a festering wound enough, it ceases to look so gruesome.
For instance, in a little simplified example, each farmer would prefer to not abuse antibiotics because they deeply care about their health and maybe even about the future of mankind. But in a competitive environment, the first farmer to use more antibiotics will gain advantage over the others, so some other farmers - maybe afraid of losing the money to feed their children - will start using more antibiotics too to stay competitive (and those who refuse on ideological grounds will simply get outcompeted). Soon, everybody is abusing antibiotics even though no one wanted it (maybe the first defector was greedy, or maybe he really needed the money). The end result is the same (competitive advantage gained by antibiotics cancels out when everyone is using them), everyone is collectively worse, and no one can break the cycle.
That's why we need external actors, like governments, to enforce some coordination. But real-life governments too often fail at that.
I linked to a powerful microscope in a parallel subthread.
The competitive environment you mention is the large scale structure of society itself. Those external actors you mention are the very forces that define the environment and define the context in which a farmer determines their self interest. These structures do not arise from the aggregate of individuals working independently, but from the coherent actions of ensembles of actors. The very notion of 'self' is ill-defined and subject to change as social forces construct the way people perceive the relationship between their bodies and the rest of the world. There is no state of nature.
It's essentially a big and complicated version of a simple game: if you press this button, you will get $10, but everybody else on the planet will lose $0.00001. Do you press the button?
The rational choice is to say yes. The benefit to you far outweighs your cost. It even far outweighs the combined costs to all the people you know.
In short, externalities make people choose bad things in a rational way. It has pretty much nothing to do with how society is structured. If you want to change it, then you need to figure out a way to make the costs to the individual outweigh the benefits to the individual.
Since we live in the same universe and have to deal with some of the consequences of each others' actions, there's some natural correlation between our utilities, but we live in civilizations which have far more social structure built on top of this basic structure, and these social structures inform and structure our possible preferences and the relations between them.
In fact, these social structures can have preferences and play their own games! The United States government, for example, is an entity whose preferences and actions are embodied in the coherence of its constituent individuals, and its structure informs the games and preferences of its individuals in a feedback loop we like to call bureaucracy. Does the large-scale behavior of the US government explain the choices of its constituents, or vice versa? Clearly the answer is neither and both: They are two ways of looking at the same dynamical structure.
In one view, things can look perfectly rational. In the other view, things can look stark raving mad. I don't think either point of view is distinguished as canonical.
I'd say that's not the point. The point is, individual selfish choices aimed at producing good outcome for the choicemaker start to produce bad outcomes for the choicemaker when everyone else makes the same choice too, but at this point it's too late to leave the ship.
A canonical toy example would be working overtime. Say everyone is working 8 hours, but you decide to stay one hour longer. The boss notices and gives you a bonus for your dedication to company. Your co-workers notice, some of them decide to stay an hour longer too. Since the bonus pool is limited, you all get somewhat smaller bonuses. Ultimately so many people stay the extra hour that it turns into a new normal; 8h/day workers become seen as slackers and are pushed to 9h/day mode to avoid angering the boss. And since it's a normal mode, nobody gets bonuses. End result: you work 9h/day and get paid the same as you used to be before.
The article I linked[0] has much better examples of this happening in real-life.
The point is, things can look "stark raving mad" pretty much objectively, because the end result is bad in the choicemaker's value system. It's easy to get into a coordination trap and incredibly hard to get out of it.
[0] - for convenience, http://slatestarcodex.com/2014/07/30/meditations-on-moloch/
I agree that neither point of view is canonical. That is basically my point: it's only madness from a certain perspective, and it's completely rational from another. Neither is right nor wrong, it's simply that the different perspectives are in conflict.
In context of the large scale structure of society, one that promotes large scale exploitation for the benefit of an individual, it makes sense to profit now and let others take care of the costs accrued by your actions.
There is no clear economic disincentive to steer people away from making this decision despite the negative outcome it will eventually have for everyone including the individual.
So in China, and every country there are a few relevant players. Farmers, farm owners, politicians, and scientists.
Farmers (laborers) have to survive and most work wages so they arent accountable. Farm owners have to compete and if they dont they go out of business and someone will take their place so they're not really accountable. But its the law that makes the rules of the game and so it is the responsibility of the political class to avoid designing "tragedy of the commons" scenarios.
There are at least SOME good-willed politicians who dont want things like antibiotic-apocolypse. The problem is that they have a ton of scientists whispering different things in their ears so they dont know what issues are big threats and which are scientists trying to get funding into their own field of study. Its therefore also the responsibility of the scientific community to effectively communicate with the political class.
See? At least some people out there arent perfectly rational!
Government is ostensibly the system that prevents external costs, and imposes rules that skew what's rational for individuals today to be closer to what's rational for society tomorrow, but in practice it fails to do so much of the time.
Also, per tradition, let me post a handy reference on coordination problems:
I'm not trying to deflect blame or avoid responsibility. I'm simply saying that this "madness" comes from individual people making completely rational choices. Widespread antibiotic use on farms is both awful and completely sensible for the individuals involved.
That doesn't mean there is no responsibility or that we're helpless. It just informs the remedies. It means that attempts to get people to act rationally will fail, because they already are. If you want to change things, then you need to make it so that the rational choice for the individual is also the best choice for society as a whole. And that's typically where the laws come in.