| This is a longish diary, but I am trying to more formally explore for myself the problems of regulation.
Graduate training is a lot like brainwashing. The student thinks her/his professors are smart and know what they are talking about. They all use models, day after day, and the essence of graduate training is that the students model themselves after the way that their professors think. Do they stop and ask "Is this the right way to think about the real world?" Of course not. For one thing, the simple models ARE essential. The world is complicated and we need to simplify it down to essential elements to begin to understand how it works. And this approach works very well for chemistry and physics, because the systems they try to understand are basically rather simple systems.
Even in economics we have gained a tremendous amount of understanding about how humans behave in response to scarcity and incentives. But social systems are almost infinitely more complicated than physical systems. There are no fixed parameters we can rely on. What we come out with are at best qualitative descriptions of how humans behave. The only numbers we use are statistics, with fairly large uncertainty bands around our best estimates.
So the student graduates and gets a job that encourages them to use their graduate training - the simple models - on the job, on a "professional" job. After all, a farmer would not want to hire a Ph.D. in economics to work on the farm; the student thinks they deserve a "professional" salary and they probably aren't used to working with their hands, and certainly not as hard as the farmer does. The student wants to use their brain, not their hands; that's why they went into graduate school, after all, and that's what fascinates them, not the hours under a hot sun or pouring rain.
My point in all this is that they come to think that using simple models is the right and only way to think about the world. That is just the way their brains have been programmed to work. After all, the models are in many circumstances really productive of a better understanding. And also, many of our grad students are people who have gone from undergrad school directly into grad programs, with no years of work history in the real world.
Another problem is that these models are all mechanistic. They are just more or less complicated formulas, with a few, simply-interracting variables. Write down the model, plug in the numbers and see what you come up with. They respond the same way every time you run the model.
Now some economic models are VERY complicated: they have (literally) hundreds of variables and equations. You might say they are so complicated that no one really understands how they work. Funny thing, though, is that they can only predict the short-run future, say the next quarter or two or three. But the longer the prediction time horizon, the worse their predictions. Because the world is constantly changing, which means the starting points for prediction in these models are constantly changing, or that their parameter values are constantly changing. So their predictions become worse and worse. Trying to predict what will happen 2 years from now is virtually impossible, other than in rather general terms.
As Joel Salatin notes, the four key principles of industrialized production are specialized, simplified, routine, and consistency. This reduces production to essentially a physico-chemcial formula that can be managed by physical scientists (or the MBAs who try to ape them).
But the real world of working people is so much more complicated than the models of social scientists. The circumstances of every person's life, and of every firm, are unique, that is to say, different from the circumstances of everybody else's life or firm. But farms are biological systems, and for that matter, so are the millions of small and large businesses. For them the key principles are diversification, complexity, flexibility, and living (read: randomness).
Now simple rules can be very useful, as per your example, tracking inventories. But even there they are approximations because of mistakes and failures to properly record numbers. I know, because the other day my wife had to talk with our pharmacist about being shorted some pills she has to take. And from his records he concluded that he should give her, free, 50% more pills than she had been shorted. So even relatively simple inventory procedures can be wrong.
So these people trained in using simple models - whether economists or lawyers - try to write laws and rules and reg's that cover the many different circumstances of a wide range of firms. But of course they are using their simple modelling frame of mind that says all farms are farms, they are mechanized systems that can be easily manipulated by changing the rules (the models) that govern them. They don't recognize that there are many classes of farms when categorized by incomes (revenues or net incomes?), kinds of products, production methods (CAFO's or free range), etc. And that the health consequences are very different for these different kinds of farms. And so are the economic consequences.
These problems are worsened when you consider the effects of big money: rules written for dairies will typically be written in a way that favors the large dairies at the expense of the small dairy. Or for fear of health consequences, rules will be written requiring that all milk sold must be pasteurized thru dairies, it cannot be sold directly off farm.
Then consider that the rules must be applied and enforced by regulators, and more especially by in-the-field inspectors. These people are not necessarily the best trained or educated. They will have been trained in basic procedures, NOT in flexibility. They are looking for "specialized, simplified, routine, and consistent" rules and operations. They will want to minimize the demands on themselves, to keep the rules as applied in the field simple. They cannot handle diversification, complexity, flexibility, and randomness (or chance). And they have the power of God over producers; they can create an hellacious amount of trouble if they want to. Again we have simple rules simply applied, creating difficult obstacles for producers. See the writings of Joel Salatin for many examples.
And too often the benefits of the rules are presumed rather than demonstrated. Physics and chemistry run experiments constantly to demonstrate that their back-of-the-envelope models are right. Social scientists cannot run experiments for the most part. How often are laws or regulations tested by looking carefully at data to see if the rule is yielding the desired results? And then shutting down the regulatory program if it is not improving public health and safety?
While very good arguments can be made for government regulation (I made such arguments for most of my professional life), writing good rules or setting up effective incentive systems can be very difficult for all the above reasons (unfortunately, I did not give equal time to such complications). You can't write a good rule or set up an incentive system if you don't truly understand the nature of the industry you are trying to regulate. That's why these NAIS hearings are so important; they have to be inundated with comments pointing out the many different circumstances of different farms and the problems the NAIS will pose for them. |
