Economics: science or not?

[bascule]

http://www.scienceforums.net/forum/showthread.php?p=511093#post511093

 

I also would disagree with you on Economics not being a real science, but that is another debate for another time, though I think your view reflects the general view often found in scientific debate concerning science as pretty much what Karl Popper and the Logical Positivists say it is.

 

So what do you think, are economics a scientific field or not?

 

I'm personally going to go with "god no". Economics is not empirical and relies too much on unpredictable human behaviors to ever have testable theories in the way you can in the hard sciences. I think the financial crisis proved very thoroughly that when you try to take scientific modeling and apply it to economics, what you end up with are not reliable predictions about the future.

[swansont]

Since I made the original statement in that thread, I'll reaffirm my "no." There is no rigorous testing, one of the requirements of science. Too many variables are in play, with no control over holding them constant in order to test hypotheses.

[Sisyphus]

I'll add a third no, generally speaking. Most economic arguments I've read don't even focus on actual empirical data, but rather what "makes sense." Hence the persistence of (::turns head away from worm can lid: stuff like classical economics that assumes rational actors, or certain left-wing ideologies that assume benevolent ones, when actually looking at the world shows these things to be nonsense. And when they do, of course its always historical analysis, not predictive experiments.

 

Just to be devil's advocate, I could say that, for example, astrophysics doesn't really have controlled experiments either. Then again, they have astronomical amounts of data...

[abskebabs]

Interesting, I'm glad this has merited discussion, and I must say I'm happy to finally be aware of the Other sciences section, I know where to place related threads in the future(please forgive me I can be a bit slow!).

 

I must confess, if I was asked the question over 2 years ago, my answer probably would have been no. My views have been influenced by my readings, especially by the work of Ludwig Von Mises and other members of the Austrian School of Economics. I think, on balance, you may find that we actually have more agreement than necessarily meets the eye;).

 

First we should be aware of our definitions. If you do not consider logic or mathematics as a science, then I think I could agree with the statement that Economics is not a science. However I would disagree with such a doctrine on the basis that we make use of these sciences at many points in our reasoning between results, theories and theorems that involve the natural sciences. To throw out the theoretical apparatus we continually employ as unscientific, I feel would be therefore unscientific, even what is referred to as employing inductive reasoning, often uses deductive reasoning to gain its conclusions, relying on the laws of logic and mathematics.

 

Economics is the study of Human Action (known as Praxeology in Misesian circles), chiefly concerned with market phenomena in an industrial economy necessitated by the existence of private property and indirect exchange by the use of a medium of exchange. For the study of this at present imaginary construction, it does make use of others like Robinson-Crusoe and a world ncompassing Socialist commonwealth.

 

I would say it meets the conditions of a theoretical science, when elaborated in the Misesian framework starting with the action axiom much like Mathematics and Logic. Additional understanding and anticipation of unknown and unknowable normative factors like values, volitions and hence where prices are tending is required however to make predictions; that are therefore not scientific in the natural scientific sense, but entrepreneurial.

 

The action axiom states simply that human beings act purposefully, henceforth aiming at an end and applying a means with which to achieve it. To argue otherwise would be a self contradictory act, as it would employ a means to achieve the end of disproving the original assertion. This makes it meet the criterion of a Kantian a priori statement that may at first strike the reader as odd, as one may imagine that Kant was aiming to produce laws of thought and reasoning, hence trying to establish a repetition axiom as the basis of Mathematics(a feat later accomplished by Paul Lorenzen and other members of the Erlangen School). However the action axiom is more general, as thought itself is another form of action. The inter-relations between the 2 are discussed in more detail in the work of another Austrian school theorist; Hans Hermann Hoppe.

 

One may object, like Friedrich Hayek, to the backing of the action axiom, when applied to , multiple individuals but oneself, for which one can show the truth is a priori. There are other counter-arguments, but a basic one would be that the uniformity of the logical structure of the human mind has always been a basic scientific assumption. These views were competently argued for by logicians like Gotlobb Frege and Wittgenstein, fighting the psychologism and polylogism prevalent in their times.

 

Now starting from the action axiom, one can elaborate various notions that must follow from it, like action being necessarily an exchange phenomena, even for an isolated individual exchanging various alternative states of perceived future affairs, and hence requiring the category of ordinal preference between these. When we enlarge our discussion to interpersonal exchange the only thing that differs is that we must begin with the knowledge posessed by both actors as to how certain actions of either could benefit the other. From this, analysis of these categories one can begin to see how the notions of opportunity cost, and psychic profit ad loss are involved and associated with action, and these are completely different to monetary profit and loss. Misesian Economics does not rely on "purely rational"(a contrary use of the word rational, as action apart from nervous reflexes is by necessity rational and purposeful) actors aiming only at monetary profit like classical and much of neoclassical economics.

 

Now, what I've mentioned so far does not change the characteristic that Economics can never be like a natural science, but what I have aimed to explain is that it attempts to theoretically and deductively the necessary categories implied in human action, especially market exchange to produce a body of necessary truths that must be fulfilled et ceteris paribus(all other things being equal), e.g. the law of demand or the Ricardian law of association. However, as the diligent ones among you have noticed, it is very difficult to ascertain the way factors are and to comb through the complex cause and effect relationships affecting the economy. What a study of Theoretical Economics can give you is an understanding, but not predictive power in the sense of the natural sciences. For those interested, you may want to look into the nature of Case probability(sometimes called Knightian uncertainty).

 

Hence if the minimium wage is raised to $10k per day, we can be fairly sure there will be some unemployment due to our understanding of market phenomena based on analysis of the fundamental catgories underlying human action like exchange; but not be able to predict how much unemployment will result due to the knowledge problem associated with et ceteris paribus conditions. Indeed, even being fairly sure qualitatively is held to uncertainty, due to the fact we may not know of mitigating central bank monetary inflation or domestic cartels that strongly mitigate the effects of the legislation(Ok, I admit that probably wouldn't happen with a $10k minmium wage).

 

Now the Austrians, never, argue and this is accurate in my view, that Economics could be like a natural science. However, it is a valid theoretical science informing us of the categories necessarily implied by the actions taken on the market given the conditions of our world, and the logical structure of the human mind. The closest we get to the kind of knowledge about actual normative factors that affected Economic exchange is through Economic History, that must analyse ex-post, using the apodictic information gathered from praxeology as well as interpretive understanding to track the cause and effect relations that dictated the past.

 

In this respect, i'd say the relation between Economics and Economic history, is like that between mathematics and physics. No physicist would deny the laws of mathematics or engage to test them when studying natural phenomena as they already underly the backbone of his theoretical apparatus used to discover normative relations in the real world. Hence, the Austrian Economist uses Economics as a theoretical superstructure with which to interpret relations in the past and present.

 

Hence, a large number of Austrian school theorists predicted the 2007-2009 housing crisis, the 2001 bubble, among others, while Mises did predict the Wall Street crash 4 years in advance. However, they did not make these predictions "scientifically" but entreprenurially, using the correct theoretical knowledge they'd already harnessed deductively.

 

Now having elaborated on the Austrian school, I must agree with you as far as you're statements refer to every other school of economics. They question basic laws in their methodology, adopt totally unrealistic ideals in Friedman's methodology, attempting to use historical statistical data as way to "test" their theories, advice they never really adopt. Hence they mostly don't practice what they preach and tend to copy the norms of Physics, while not making much use of them apart from perhaps mental masturbation through applied mathematics. I would say these other schools, especially the dominant neoclassical synthesis richly deserve the title "pseudo-science".

 

Phew! For more, I recommend a visit to mises.org

[iNow]

Looking forward? No, it's not science... In fact, it's hardly better than reading tea leaves or figuring out the color of someone's chakra.

 

Looking backward, however, yes. We can, in retrospect, apply the principles of science to understand the system and determine causes of various activities, look strongly at correlations, and accurately describe how specific inputs influenced specific outputs in the economic machinery.

 

 

So, while it's view of the past is scientific, economics view of the future is pretty much on par with numerology.

[Mokele]

If you do not consider logic or mathematics as a science, then I think I could agree with the statement that Economics is not a science.

 

Bingo.

 

Mathematics, logic, etc aren't a science, though they are tools of science. If simply applying logic to something makes it scientific, then Theology is a science. If simply using math makes it scientific, then Astrology is a science.

 

Your definition of science is so broad as to be useless.

 

Science is testing hypotheses. Nothing less. Yes, there are descriptive elements (species descriptions, determining physical constants), but those only exist in service of hypothesis testing.

---

Merged post follows:

Consecutive posts merged

And, prememptively, "it's too big/complex/historical to do experiments" doesn't fly - see astrophysics, ecology, and paleontology, respectively.

[Mr Skeptic]

I'd say economics is a science, but with a caveat: there is pathetically little data, and experiments are absurdly hard to do. Without decent data and rigorous testing, science gives you stuff like the phlogiston and caloric theories, that "make sense" more so than make clear predictions. You can do science with very little data but it won't be very solid.

[Sisyphus]

Now having elaborated on the Austrian school, I must agree with you as far as you're statements refer to every other school of economics.

 

My theology is a priori truth, while of course all others are wildly unrealistic.

 

(For everyone's sake, I think it's time to broaden your reading material.)

[Mokele]

I'd say economics is a science, but with a caveat: there is pathetically little data, and experiments are absurdly hard to do.

 

Not really. Sure, at the macro scale things are a bit difficult, but it's very easy to set up smaller-scale experiments which, if designed right, can give valuable insight.

 

As I pointed out earlier, look at ecology - it's vastly more complex than economics, but we've managed to gain quite a level of understanding by finding model systems, using long-term manipulations, etc.

 

Any economist who claims "we can't do experiments" needs to go take ecology. For a lot who harp on about 'hard work', they sure are lazy.

[Mr Skeptic]

Another problem with economics is that people use economics theories to guide some of their economic actions. For example, a new scientist article shows how some scientists can predict when people are buying stock because the market is rising rather than due to intrinsic value. But almost immediately, the theory will no longer be valid -- people will use that theory to predict when to sell stock, and so undermine the theory. As a comparison, the three body problem in physics is very simple and well defined but unsolved because the equation for each body depends on the equations for the others.

---

Merged post follows:

Consecutive posts merged

Sure, at the macro scale things are a bit difficult, but it's very easy to set up smaller-scale experiments which, if designed right, can give valuable insight.

 

And such experiments are done, but the insights gained aren't all that valuable.

 

As I pointed out earlier, look at ecology - it's vastly more complex than economics, but we've managed to gain quite a level of understanding by finding model systems, using long-term manipulations, etc.

 

I bet ecology would be much harder if organisms studied ecology and used it on themselves. I don't think ecology is more complex, there is just a lot of it.

[Mokele]

And such experiments are done, but the insights gained aren't all that valuable.

 

Neither are the insights from any single experiment in other sciences, including physics. It's putting them all together that makes it possible to gain larger insights. You can't just abandon the scientific method when it doesn't solve everything on the first try.

 

I bet ecology would be much harder if organisms studied ecology and used it on themselves. I don't think ecology is more complex, there is just a lot of it.

 

Actually, IME, anything with humans is easier because you can give humans instructions, and we document everything anyway.

 

Anyhow, we could quibble over which is more complex, but the point is that both are tremendously complex systems, yet one has embraced the scientific method and became a legitimate field, while the other has avoided actual experiments and became pseudoscience.

[Sisyphus]

Another problem with economics is that people use economics theories to guide some of their economic actions. For example, a new scientist article shows how some scientists can predict when people are buying stock because the market is rising rather than due to intrinsic value.

 

Here's a fun statistic: in 2009, 73% of all stock trading volume hasn't involved human decisions, but automatic, high-frequency algorithmic trading. So not only is it totally divorced from what they're actually buying and selling, but it's not even human beings looking at the numbers. Human beings get the profit, though, pretty much directly at the expense of long term investors. So, just as technology has lowered the barrier to entry with online investing, it raises it back up again (big firms can spend literally hundreds of millions of dollars on these algorithms). And it's not even pretending to be related to intrinsic value. Who wants to make a prediction about what's going to happen next?

[The Bear's Key]

I'll add a flat-out no as well

 

I would say it meets the conditions of a theoretical science, when elaborated in the Misesian framework starting with the action axiom much like Mathematics and Logic. Additional understanding and anticipation of unknown and unknowable normative factors like values, volitions bluffing and hence where prices bets are tending is required however to make predictions; that are therefore not scientific in the natural scientific sense, but entrepreneurial.

Wouldn't that mean poker is science too?

 

 

...of the Austrian School of Economics.

........

...of another Austrian school theorist; Hans Hermann Hoppe.

........

Now the Austrians, never, argue and this is accurate in my view, that Economics could be...

........

Hence, the Austrian Economist uses Economics as a...

........

...having elaborated on the Austrian school

I hear so much on the Austrian economics school of thought. And it has me wondering.

 

Do the Austrian economists match up with the Austrian economy? It has strong labor unions, who help ensure a healthy wage nationally. The government also is measurably Social Democratic (politically centre-left), and it's forbidden nuclear power in favor of hydropower and other renewables (at least 60% of total electricity, the rest gas and oil).

 

 

...a large number of Austrian school theorists predicted the 2007-2009 housing crisis, the 2001 bubble, among others, while Mises did predict the Wall Street crash 4 years in advance.

Would you mind verifying those?

[CaptainPanic]

Economics is in the same league as History. You can study it, get a Msc or a PhD in it, in but it will never give any guarantees for the future.

[The Bear's Key]

...a large number of Austrian school theorists predicted the 2007-2009 housing crisis, the 2001 bubble, among others, while Mises did predict the Wall Street crash 4 years in advance.[/quote']Would you mind verifying those?

Actually, for the housing crisis it wouldn't be surprising. It's not as much Austrian but common-sense economics.

 

What goes up quickly, usually comes down quickly. I'm not economist, but I had detected some fairly obvious patterns (and its connection to stock patterns) when a guy teaching me about Futures revealed some pretty impressive tendencies which you could use to increase your "betting" chances by laying aside emotions and remembering a few simple but critical rule-of-thumbs while maintaining vigilance.

 

He'd printed out the last few weeks of the graphed market, then the last few years, and then the last 50 years. The principles worked regardless the scale you looked at it (on the web you can pick out any group of years).

 

Of course I didn't pursue it, he just liked teaching me a potential for cash that was still fairly risky but minimized it in some areas if you were extremely careful.

 

He probably hadn't intended for me to draw a larger lesson (including about stocks in general).

 

It's amazing how the market behaves, and I almost could visualize it as a live beast -- one that you have little chance of knowing exactly what it's going to do, but still you can make fairly educated assumptions -- just never 100%, and rarely even close to that. In an economics class later I gained more knowledge and deduced further things.

 

But in the end, it's all probabilites mixed in with behavioral tendencies and periodically snap judgements, nothing but a fancy betting scheme. It most certainly isn't science.

Edited August 31, 2009 by The Bear's Key

[abskebabs]

Bingo.

 

Mathematics, logic, etc aren't a science, though they are tools of science. If simply applying logic to something makes it scientific, then Theology is a science. If simply using math makes it scientific, then Astrology is a science.

 

Your definition of science is so broad as to be useless.

 

Science is testing hypotheses. Nothing less. Yes, there are descriptive elements (species descriptions, determining physical constants), but those only exist in service of hypothesis testing.

---

Merged post follows:

Consecutive posts merged

And, prememptively, "it's too big/complex/historical to do experiments" doesn't fly - see astrophysics, ecology, and paleontology, respectively.

 

Hmm, yes I suppose Theology does make use of Logic, but I wouldn't compare it to Economics because even though it is aprioristic in a certain sense, it does not rest on synthetic a priori foundations of the Kantian form, whereby the fundamental statements are such that even an attempt at their denial relies on implying the existence of the original statement.

 

The view that Mathematics and Logic are not sciences (but "tools of thought") is essentially the modern day view of science that has come from the influence of the school of Philosophy known as the "Logical Positivists" (Popper and Karnapp were prominent members I believe).

 

Their foundational statement is that:

 

1. All statements are either analytical or hypothetical.

 

Analytical statements are essentially defined tautologies, that do not have any relation to reality, but can be part of a self contained deductive truth.

 

Hypothetical statements are statements of pure hypotheses that require "testing" for falsification, and cannot be proved right but can be ruled out. These are the only statements that can have any relation to the real world.

 

According to this view there can be no a priori truths that can refer to reality.

 

The problem lies with the foundational statement of this philosophy of science. Statement 1 can only be analytical or hypothetical by its own requirement. If it is analytical, then it is itself an anlytical statement that we cannot take to have any relation to the real world! If we consider it hypothetical then we must test as a hypothesis, whether all statements can be hypothetical or analytical only! Hopefully you realise how ridiculous this, and how empty a defnition of science it provides.

 

I think you have a good point about complexity. It is extremely hard to separate cause and effect in Economic phenomena, the goal of historical research in this area is therefore to ascertain the inflouence of different factors already known from theoretical economics( e.g. what influence did the price restriction have, what extent was the role of moral hazard? etc). There may be a role for "experimental economics" in terms of discovering the extent of the influence of theoretically known factors, in certain scenarios, but it cannot replace economic theory, in much the same way that you can't use physics to test maths, it relies on it already to find the extent of normative factors within the physical universe.

 

The atheoretical approach to Economic phenomena was exactly the approach of the German historical school and the American institutionalists. The problem is that as soon as one stops recording raw data, and attempt the use of statistics or regressions, you are already applying a priori assumptions and theory, only this time it is absolutely arbitrary. The root of the problem is that Economic phenomena are subjective, and what I mean by this is that the value that is placed on products is ultimately based on the subjective preferences of individual consumers. The prices of higher order goods and capital are dependent on the prices of products, as their price depends on the profit anticipations of entrepreneurs and speculators.

 

Concepts like price and opportunity cost will never be "empirical" in the sense we define our terms in the natural sciences. You don't "test" them, just like you don't gather apples to find if 2+2=4. They are already implied in the act of exchange, whether interpersonal or autistic, and deductions based on these categories is sound, but the extent of the influence of the factors cannot be known beforehand in reality, and this does require historical research and interpretation. Entrepreneurs constantly use this knowledge combined with historical understanding to curb their actions to the expected prices of the future.

 

Another problem for an alternative approach to Economics is the lack of constants, whereby all causative factors are variables. For example, if you "measure" the change in price by +5% after a 10% drop in supply one has not "measured" elasticity of demand but recorded a historical effect. Indeed supply and demand curves are problematic in this regard because they assume simultaneous determination of the price by the change in supply, when the actual influence is causal, and takes time to affect the change in price. These are just 2 other problems of the Mathematical approach to Economics.

 

I think you guys are thinking this is saying a lot more than it actually is:

In the Misesian scheme, it is acknowledged that the scope of Economics is far smaller and modest with regard to markets than is assumed in the rest of Economics. Quantitative predictions are explicitly recognised as impossible due to problems associated with knowledge of et ceteris paribus conditions. For a better attempt to explain some of the foundations of this approach to Economics, I'd recommend the following paper:

http://mises.org/journals/scholar/long.pdf

[Sisyphus]

Hmm, yes I suppose Theology does make use of Logic, but I wouldn't compare it to Economics because even though it is aprioristic in a certain sense, it does not rest on synthetic a priori foundations of the Kantian form, whereby the fundamental statements are such that even an attempt at their denial relies on implying the existence of the original statement.

 

Actually, being in the sad position of having read a great deal of theology, I assure you that many (most?) theologians would claim to be doing exactly that.

 

Also, this:

 

The action axiom states simply that human beings act purposefully, henceforth aiming at an end and applying a means with which to achieve it. To argue otherwise would be a self contradictory act, as it would employ a means to achieve the end of disproving the original assertion.

 

is, in fact, very much faulty logic. To argue against it implies that human beings are capable of acting towards an end, not that they must, and willfullness /= towards an end, let alone by coherent or rational means. This is Aristotle stuff.

Edited August 31, 2009 by Sisyphus

[abskebabs]

Actually, being in the sad position of having read a great deal of theology, I assure you that many (most?) theologians would claim to be doing exactly that.

 

Ok, fair enough, would you be able to read the first few pages of Human Action, and the above paper, and tell me what you're impressions are then? I'm just curious if it is similiar to the kind of stuff you've read in Theology.

[Mr Skeptic]

Here's a fun statistic: in 2009, 73% of all stock trading volume hasn't involved human decisions, but automatic, high-frequency algorithmic trading. So not only is it totally divorced from what they're actually buying and selling, but it's not even human beings looking at the numbers.

 

So basically, the majority of stock trading volume is based on economic theory, rather than anything that could stay constant. Therefore there is no constant strategy to make profits, but rather the best strategy depends on the strategy that others are using (ie, if you know what another will buy/sell, you want to buy it first so that you get it while it's cheap, and sell it first so you sell it while it's most expensive).

 

This is basically what I was saying, economics is self-referential which makes it much harder than stuff like ecology.

[Sisyphus]

It should be noted that evolution is self-referential, as well. Every organism is part of its own environment, and there is no constant ideal survival strategy.

[J.C.MacSwell]

I will say yes, definitely.

 

Is Pharmacology a science?

 

Psychology?

 

How about the study of the Origins of the Universe?

 

All of the above have limits to how rigorously they can examined, making it hard to separate the wheat from the chafe, good science from bad, but are they not all science?

[bascule]

Those of you defending economics as science:

 

Can you name one falsifiable prediction made by "economic science"?

[abskebabs]

Those of you defending economics as science:

 

Can you name one falsifiable prediction made by "economic science"?

I've spent most of this thread trying to relay the difficulties inherent within the subject for this to be the method of its advancement, whether it be the necessarily subjective and already theoretical and logical aspect of its subject matter, or the trouble to do with the fact that alll its positive statements are made fully aware of the fact they are made assuming et ceteris paribus(all other things being equal, knowing these conditions do not exist in the real world. Another problem is the causal, not simultaneous influence of the interrelated data. One example is the statement et ceteris paribus, that increasing the money supply will produce price inflation or lower price deflation due to increased productivity.

 

I've spoken with a friend from another forum about this, and I've realised I have misrepresented Mises' statements about Economic methodology with Rothbard's. For example Mises never spoke of an "action axiom", so what I'm trying to say is don't take me as authoritative in my attempt at an account here, I think I've made a few mistakes.

 

I've tried to explain why the nature of prediction is an altogether different thing in Economics, but just for fun, how's this:

 

Let's raise the minimium wage to $1k an hour and see if we don't increase unemployment and force economic activity underground.(assuming the federal reserve doesn't inflate EVEN crazier than already)

[iNow]

Right. So, uh... still no falsifiable predictions, I see.

[Mr Skeptic]

Those of you defending economics as science:

 

Can you name one falsifiable prediction made by "economic science"?

 

Since I already mentioned one, how about this:

For example, a new scientist article shows how some scientists can predict when people are buying stock because the market is rising rather than due to intrinsic value. But almost immediately, the theory will no longer be valid -- people will use that theory to predict when to sell stock, and so undermine the theory. As a comparison, the three body problem in physics is very simple and well defined but unsolved because the equation for each body depends on the equations for the others.