SoylentNews is people
SoylentNews
It is often opined here on Soylent that Economics isn't real science. You would expect economists to disagree with that sentiment, but it turns out that there is a growing international movement within the world of economics itself, started by a group of students, that seeks to drastically overhaul the entire field. Some choice quotes from the article, which is in fact a review of a book "which formalises and expands the case" that economics is in need of reform:
In the autumn of 2011, as the world's financial system lurched from crash to crisis, the authors of this book began, as undergraduates, to study economics. While their lectures took place at the University of Manchester the eurozone was in flames. The students' first term would last longer than the Greek government. Banks across the west were still on life support. And David Cameron was imposing on Britons year on year of swingeing spending cuts.
Yet the bushfires those teenagers saw raging each night on the news got barely a mention in the seminars they sat through, they say: the biggest economic catastrophe of our times "wasn't mentioned in our lectures and what we were learning didn't seem to have any relevance to understanding it", they write in The Econocracy. "We were memorising and regurgitating abstract economic models for multiple-choice exams."
Part of this book describes what happened next: how the economic crisis turned into a crisis of economics. It deserves a good account, since the activities of these Manchester students rank among the most startling protest movements of the decade.
After a year of being force-fed irrelevancies, say the students, they formed the Post-Crash Economics Society, with a sympathetic lecturer giving them evening classes on the events and perspectives they weren't being taught. They lobbied teachers for new modules, and when that didn't work, they mobilised hundreds of undergraduates to express their disappointment in the influential National Student Survey. The economics department ended up with the lowest score of any at the university: the professors had been told by their pupils that they could do better.
(Score: 0) by Anonymous Coward on Sunday February 12 2017, @08:57PM
Study the philosophy of science before you presume to lecture others on what is, or is not a science.
Economics is, for the record, a science. It just doesn't happen to be a physical science. Nor is it an analytical science. It's a human science.
(Score: 0) by Anonymous Coward on Sunday February 12 2017, @09:43PM
Econ degree checking in here.
I would argue it *could* be a science. They are however right. Econ has a reproduction problem. With a good scientific test you should be able to come up with a theory test it and see if it is correct. Change the parameters and see if the hypothesis still holds true. The problem economics has is when you experiment and it goes wrong people go hungry. How do you build a model on something where you can not realistically change things without pissing everyone off? Economics has lots of models that they we are fairly sure work in particular cases. But there is no theory of economics. There is a lot of competing political dogma ideas. Even the models we have do not describe what happens if one of your assumptions turns out to be a variable and not a constant.
The teaching of econ is also *very* basic. You can easily understand any of the existing models with an algebra 1/2 education. Needless to say as someone with a CS degree (which is almost a math degree) I was bored. I had one teacher who figured out linear algebra and he applied it to everything. He was going along the right track but it probably needs real math people in there with diff equations. Better methodology and better ways to create/test models. Most models the lay person hears about are thinly veiled political dogma with poor assumptions on inputs/outputs and the process of change.
(Score: 0) by Anonymous Coward on Sunday February 12 2017, @10:39PM
Both econ and philosophy checking in here ...
Mathematics does not something scientific make, any more than a (mis)application of empirical techniques to fields that are not generally suited to such techniques.
In fields such as physics, the mathematics used is intended to provide a model for what is purportedly happening. The better the model, the better the predictions, and (handwaving some assumptions about determinism) there is no gap between the mathematical and observed behaviour that is not adequately explained by measurement errors. This is what makes physics an empirical science in which inductive reasoning is a feasible approach (handwaving metaphysical arguments about the nature of the universe and induction as a means of forming predictive models).
The assumption that somehow tacking mathematics onto an explicative narrative makes it more scientific is to misunderstand the nature of human sciences as opposed to empirical sciences. If you have a model that can strictly be described in mathematical terms, rock on. But the fact is that we don't, and inasmuch as it has been attempted, it has failed. The same could be said of psychology, sociology, anthropology and this should come as no surprise.
If you want to make economics more scientific (a questionable way of phrasing the goal) you would do well to start by proposing a more formal language with which to describe and analyse it. Applying formal logic to this is likely to be a doomed enterprise, but maybe one day you can approach psychohistory with it. I'm not holding my breath, here.
(Score: 2) by Phoenix666 on Monday February 13 2017, @02:39AM
I don't know about that. It often feels to me that people are grafting assumed certainties onto "real" sciences in order to discount other disciplines. If we apply the scientific method to an area of study, are we not doing real science? If we are basing our models on observed data, are we not acting empirically? If we are using mathematics to describe the phenomena we are studying, are we not using 'formal language?'
Or is experimentation essential to differentiating "real" sciences from "social" sciences? Because if that's the case, then what would we call astronomers and cosmologists, who have no mechanism for experimentation on the heavenly bodies they study?
Or is the defining characteristic that nothing that studies humans can be considered a "real" science? If that's so, then what do we call medicine? Is that as not-real as sociology? Pharmacology? Does that get thrown out into the realm of not-real because not everybody responds to the same drug in the same way? After all, we have no idea why that is, exactly. We can only rid ourselves of those discrepancies by a lot of hand-waving.
The thing is, the "soft" sciences, the "social" sciences, could probably do a lot better on the subject of reproducibility if the gloves were completely off. Want to test the validity of game theory predictions for market behavior? Great, randomly select 100 homeless and throw them into a featureless maze with the instructions that those who get it wrong will die. Want to firmly nail down what happens when we have a supply shock? Great, take 1000 Chinese and give enough rice to feed 100 to 10 of them and watch what happens. As it is, ethical constraints generally prevent that kind of experimentation. Designing experiments that get around that limitation, that achieve explanatory, reproducible results is damn difficult--order of magnitude more difficult than the simplicity of the physical sciences, whose molecules and forces can't have us thrown in jail or scream no matter how much we abuse them.
If anything I'd caution that our collective experience with climate science over the last 30 years ought to constrain the smugness of those who feel that their "science" is unassailable, immune to political narratives or social resistance, somehow more real and absolute than those who believe differently. It's not. (For the record, I side with those who champion climate science) And those who work in science can attest that there are far more political machinations and self-serving narratives involved than there ought to be in anything that calls itself science.
In short, there's a lot more hand-waving and gloss to differentiating between "real" science and "soft" science than many realize, and which shrinks or disappears outright under close questioning of those who practice science.
Washington DC delenda est.
(Score: 0) by Anonymous Coward on Monday February 13 2017, @07:04AM
Even in the social experiments you mention, you wouldn't necessarily be proving anything. There are just so many confounding variables in these events. I realize your examples were off the cuff and rhetorical, but they expose a real issue nonetheless. Your selection of Chinese, or homeless is going to be biasing. And the way they act in an experimental circumstance is going to be different than they may act 'in the wild.' Even the people who would volunteer for such study is a biasing. Even if you could coerce people into participation, that again imposes yet another bias. You'll be lucky to reproduce any results, let alone actually derive meaningful and certainly not actionable or predictive science from this. I mean I'm very curious to see how these experiments would turn out, but the results -despite their interest- would not be sufficient to base anything on.
Take astronomy as an example you gave, for a view 'from the other side.' Astronomy is the source for astronomic spectroscopy. That's a lengthy way of saying observing the electromagnetic emissions of various types of celestial bodies. We start with the observation that various bodies emitted various spectra. We created a hypotheses for the reason and then the key step is the testing and predictive power of said hypotheses. It ends up getting tied into quantum mechanics but essentially it was able to be proven with 0 doubt, that hot bodies emit spectra of a certain type that is a result of their elemental composition. Any and every experiment you could compose to test this theory would verify it 100% of the time. This allows you to take this as a granted and build upon it almost as a first principle. And yeah, the subject of this thread was a not coincidental play on words.
(Score: 0) by Anonymous Coward on Monday February 13 2017, @07:06AM
If you have a model that can strictly be described in mathematical terms, rock on.
Here is the thing you must have missed in your econ classes. They claim they do. Yet have managed to miss every recession and boom since the 1930s. The models they do have model very narrow fields of study. The rest is theory with little real good application. Then the political people enter in and try to bend it to say whatever they want. For example pretending giving everyone money for nothing in return will somehow magically stimulate the economy, or lowering taxes will somehow magically stimulate business investments.
you would do well to start by proposing a more formal language with which to describe and analyse it
They are trying to do that. But how do you define and graph happiness and utility? Which are real variables in economics as you know. But there is pretty much 0 way to measure either of them. Those are but two small examples of what our 'science' is. It is not even soft you can not even measure it. Lots of theory but not much to really back it up.
but maybe one day you can approach psychohistory with it
I gave up on economics because it is a waste of time. CS is much more lucrative and I can do real things with that. About all I got out of econ was a nose for BS business plans.
(Score: 4, Insightful) by driverless on Monday February 13 2017, @02:51AM
Friend of a sysadmin who works at a bank checking in here, he's a college drop-out, self-educated. In January 2008 when I was visiting him, he pulled up some data on mortgage defaults across the US. These are in default, these are (effectively in default but not called that), these are ninety days late, sixty days, etc. Now everything sixty days and over is in default, because if you've missed two mortgage payments you're not going to suddenly recover. This means the default rate is actually xx%, not y% like the official figures. Now that's for Minnesota. Let's look at somewhere big like California. Here's the figures. And that works out to $xxxB in effective default in just one state. The finance companies have this much, there's insurance cover but they can't handle this level of default, etc etc. We're fucked. Well, they're fucked, I work for a bank. Anyway, I'm done for the day, where do you want to go for dinner?
This was a random IT sysadmin with no formal education, who saw the subprime crisis and meltdown more than half a year before it arrived. And he wasn't some hidden genius, it was just a regular guy who got bored at some point, looked at the data, and saw what was obvious to anyone with the slightest bit of common sense. It's no wonder that people treat economists as little more than incompetent quacks.
(Score: 0) by Anonymous Coward on Monday February 13 2017, @12:55PM
Plenty of economists predicted the subprime crisis - as in, central bank policy creating investment bubbles especially in real estate - they just weren't published in mainstream media because no one wants to hear that.
(Score: 0) by Anonymous Coward on Monday February 13 2017, @02:19PM
What's that old adage:
"Economists have predicted 12 of the last 3 recessions"