Sharing for the discussion.
If what separates physics from economics in eyes of the scientific method is the inability to test hypotheses in the latter field, then this article goes some way in refuting that premise.
I'll go one further. Does anything that dogs economics as a scientific endeavor not dog medicine?
THAT'S WHAT HAPPENS WHEN YOUR MATH SAYS AN ECONOMIC COLLAPSE IS LESS LIKELY THAN THE SUN SUDDENLY GOING OUT I’m troubled by the sense among skeptics that disagreements about the answers to certain questions suggest that economics is a confused discipline, a fake science whose findings cannot be a useful basis for making policy decisions.
Yes. Whereas biological systems evolve slowly enough such that what worked treating my grandfather will likely work upon me, economic theories are based in environmental contexts that can shift dramatically over time, and significantly alter the outcome. That's not to say that economic experiments aren't informative or worthwhile, but there are political, technological, and cultural influences that make the results less generalizable. Economics is a soft-science. Medicine is far less soft, because it is based upon biology, which is much harder.Does anything that dogs economics as a scientific endeavor not dog medicine?
Please know that I'm trying to agitate for a discussion and not merely being obtuse. Does medicine not suffer from political, technological, and cultural influences, particularly influences that affect the generalizability of their results? One could argue that psychiatry, for example, suffers from an enormous pressure to further legitimize their field, pressure to find and publish statistically significant research, while they struggle to handle the lack of reliability in their data. As for political and technological influences, isn't the DSM-V substantively more a reflection of American conceptions of health than say, Chinese or Iranian? Are American conceptions more objective, or is it unfair to describe the DSM-V as American? This is not dismiss the underlying biological nature of medicine (nor to criticize searching for objectivity in the face of towering subjectivity). But I see similarities in the hard science of biology as I do in the measurable behavior and amassable statistics of human economic activity. N.B. I am also the very first in line to point out that the squishiness of economics as it stands today is woefully un-predictive, if not downright dangerous -- a la Stalinism or the North Korean economic dogma -- and that a lot of professional economists are nothing more than a willing pardoner class, selling indulgences to the highest bidders from the "job creators".
You're certainly correct that medicine, as such, lies at the interface of biology and social science. Even biology itself, exclusive of medicine, can be greatly affected by sociology. There is a great book on the topic called The Mismeasure of Man by Stephen J. Gould in which he recounts the vast undertaking in the late 19th c to "prove" the superiority (defined arbitrarily, of course) of whiteness. This was at a time when blacks were suddenly free and many immigrants from parts of Europe other than Britain, Germany, France and Scandinavia were suddenly entering the US in great numbers, especially Jews from Eastern Europe. In response "Science," led by the Swiss-American head of zoology at Harvard, Louis Agassiz, "proved" that there was a racial hierarchy to man, with (to no one's surprise) Britons and Germans being at the top, followed by the various groups all the way down to blacks and Native Americans, the inferiority of whom closely correlated with the relative darkness of skin. This work was the "scientific" basis of immigration quotas, and eventually was used as a founding principle of eugenics (which of course helped justify the Holocaust). So no, biology is of course not immune to social leanings, and of course medicine, which is part biology and part sociology, is even more prone. Edit: As scientists, I think the best we can do is recognize that we're not input/output machines. I don't know if we can do any better than that, but I do spend a hell of a lot of time thinking about it, which I hope is a good start.
Are we not, though? Or are the inputs and outputs so multitudinously abundant that it's impossible to measure? This reminds me of a Dave Foster Wallace quote. thoughts and associations can fly through your head. You can be in the middle of a creative meeting at your job or something, and enough material can rush through your head just in the little silences when people are looking over their notes and waiting for the next presentation that it would take exponentially longer than the whole meeting just to try to put a few seconds’ silence’s flood of thoughts into words. This is another paradox, that many of the most important impressions and thoughts in a person’s life are ones that flash through your head so fast that fast isn’t even the right word, they seem totally different from or outside of the regular sequential clock time we all live by, and they have so little relation to the sort of linear, one-word-after-anotherword English we all communicate with each other with that it could easily take a whole lifetime just to spell out the contents of one splitsecond’s flash of thoughts and connections, etc. — and yet we all seem to go around trying to use English (or whatever language our native country happens to use, it goes without saying) to try to convey to other people what we’re thinking and to find out what they’re thinking, when in fact deep down everybody knows it’s a charade and they’re just going through the motions. What goes on inside is just too fast and huge and all interconnected for words to do more than barely sketch the outlines of at most one tiny little part of it at any given instant.Edit: As scientists, I think the best we can do is recognize that we're not input/output machines.
I know that you know as well as I do how fast
I think much of modern neuroscience, and all of "neuropsychology," have far more in common with phrenology and crainiometry than they do with physics or chemistry. That's really the salient point I want to make. Science really isn't about discovering the truth. It's just a systematic way of asking and answering questions. If the system you devise if flawed, then the answers will be flawed as well. They questions we ask, they way in which we ask them, and how we interpret the answers, are affected greatly by our upbringing. This has been proved many times over, but somehow each generation thinks that they're the one who has shed this baggage.
I'm not reading a DFW quote because fuck that guy but the serious answer to your question: This gets to the problem of modeling. I've always been a cynic of statistics because a successful statistical model is predicated on accounting for all the relevant inputs and outputs in order to get a meaningful answer. In engineering, this is easy when you have a broad basis of data from which to model, and tricky when you don't. Take airplanes: we know the stress characteristics of steel in a boundless bounty of conditions because we've been using it and testing it for 150 years. Aluminum is a little trickier because it's a more reactive metal and has no yield strength (you can't bend it back and forth without fucking it up) so initially, aluminum components failed more unexpectedly than steel components. Things got really bad when everybody shifted to carbon fiber and other composites because the makeup of composites is subject to much greater variation and can be tested in far fewer ways. So your model of steel reveals few surprises, your model of aluminum reveals a few more, and your model of carbon fiber reveals more. But then, the more of these systems you have in the world, the fewer uncertainties you have in your system and the more likely it is that your model accurately reflects the problem. The issue with economics is that you can introduce things like mark to market accounting and suddenly every assumption is challenged. An investment bank can come up with the idea of collateralized debt obligations and have them on the fucking market within a few months. We're here in the land of Zero Interest Rate Policy which is, in economic terms, the equivalent of saying "what if gravity were positive rather than negative?" So in part, they're blameless because modeling a plane (or a mouse, or a supernova) has pretty good boundaries where you need to model and a good sense of where you don't while modeling an economy is a giant loosey-goosey unhappy no-fun zone. But in part, they're fully to blame because economic "science" is founded on the notion that you can model some aspect of that giant loosey-goosey unhappy no-fun zone in order to learn something valuable. And they take that assumption and extrapolate it out to the entire goddamn human race. An aircraft manufacturer will get an idea that maybe jet canopies could be made of a stronger material and, along the way, invent crazy glue. An economist will fail to notice that they're making aircraft out of crazy glue. Or are the inputs and outputs so multitudinously abundant that it's impossible to measure?
Elaborating: All science is, to some extent, modeling and selection of boundary conditions. For a biological model, it's fairly settled and fairly defensible that the model extends all the way to the mouse. For a behavioral model, the model may extend all the way to the mouse's cage. For a chemical model, it may include mouse pee. For an economic model, the model doesn't include what the mouse had for breakfast, but it counts lunch twice, and only addresses twitches of the mouse's whiskers when the mouse meets a girl mouse but not a boy mouse because SCIENCE, MUTHERFUCKER.