Economic Theory versus Economic History
In our ongoing Methodenstreit here at Free Advice, I thought the following comment from Kevin Duncan was a good springboard to amplify my position:
Questions such as what has driven real wage growth among certain sectors of the economy are routinely asked within labor economics, and fairly robust empirical methods give reasonable and consistent answers. The same thing has occurred with human capital attainment. Things like the “mincer equation,” have been incredibly robust benchmarks for individual wages across various countries and time periods, almost stupidly so in explaining about the same variance, and having routinely approximated coefficient estimates. Yes, I’m aware of the Austrian critiques against such extrapolation, but used descriptively they do offer us a better insight into how things like education attainment compared to raw skill growth tends to affect life time earnings of workers, and other extrapolated knowledge.
A lot of tax policy work has slowly come to a larger consensus as empirical methods have improved in how to approach discontinuities.. Same thing goes among estimating how various health care policies can impact inter-hospital level of care. The main factor here is that these studies focus on smaller areas than are typically mentioned in the WSJ or reported by groups such as Macroeconomic Advisors.
This is a perfect illustration of the difference between economic theory and economic history.
Let’s say we figure out “what has driven real wage growth” using empirical methods. After estimating our coefficients with backfitting, we can forecast future wage growth using our calibrated model. Bam! It predicts extraordinarily well, given the crude quality of our data collection.
OK great. Now: Does anybody think we have discovered a true economic law in the same way that physicists think that there are actual laws governing the behavior of matter?
I hope not. No matter how much “experience has confirmed” an empirical regularity in the social sciences, people still have free will (at least operationally, if we’re doing a social science rather than looking at them as collections of atoms) and so those forecasting models could go out the window tomorrow.
In contrast, if I logically deduce that, “The purchasing power of money will be lower, other things equal, when the stock of money increases,” then that is a genuine law of economics. If you allow me to define the terms etc. in such a way that that proposition is true today, then it will necessarily be true for all time and all cultures.
Last point: Ultimately, we really don’t even have any good reason to expect that Nature herself lacks free will, and might upset what we took to be her “laws” tomorrow. This is of course Hume’s famous problem of induction. But the physicists and chemists can shrug at the philosophers and say, “You go ahead and argue about why it works, but it clearly does.”
In contrast, the economists–certainly the macroeconomists–have no business whatsoever telling the rest of society, “You go ahead and argue about why we can predict so well in our field, but clearly we can.”
Bob, just off the cuff here but there are laws that exist in economics. I actually think there are laws that govern ALL things. Ergo if there are laws that govern human behavior, which there are, then there MUST be laws which exist within economics. Here are a few that I will express in not so elegant language.
no work equates to no production
No innovation equates to no productivity gains
A lowering population equates to less opportunity of trade.
A Central Bank Impacts monetary policy. – This one is just fun I wanted to say negatively but then I am just being snippy.
Now does this mean you can PREDICT future economic behavior? Not so much. Heck even now when attempting to form a crystalline structure physicists and chemists can CAUSE the structure to be created. They can use it for its properties, but even now to date I do not think they can predict the EXACT way the atoms will bond together, add in natural influence outside of a lab and it is right mess. This does not mean that you do not UNDERSTAND the behavior or the laws governing it but rather you cannot predict variables outside the scope of the lab.
This is not to say we ‘know’ anything either I suppose… Rather that through perception and evidence we build up a concept only to tear it down when we find a flaw in it. We must also contend with an ever changing policy structure. Please tell me the last 10 year period where we did not innovate or create laws effecting economics in some way. Hard to predict the future when there is no consistent past to go off of. So instead you have to create generic constructs that ignore the impacts of those dealings or seek to explain them in context.
Bottom line. There are economic laws. But we are too fickle to figure them out through real experimentation.
He doesn’t deny that there are economic laws. He even gave an example of an economic law in this very post.
Dan wrote:
He doesn’t deny that there are economic laws. He even gave an example of an economic law in this very post.
Thanks Dan. Using induction, it is a law that I just need to stop writing on this topic. I can’t believe how people are misinterpreting me.
(I’m not saying maliciously; it’s just becoming clear how odd the Misesian position is to people.)
Ahh… My Mistake in reading then.
“I actually think there are laws that govern ALL things.”
What are these laws? I don’t mean give me examples, I mean, in what sense do you think they are real? Are they commands from God? Platonic forms? Necessary facts about any universe whatsoever?
What is, for instance, the “law of gravity,” and how does it exert its governing force over concrete objects? (Yes, I understand “inverse square,” but what I want to know is how you think this law “governs” a physical object?)
A law is the inherent nature of entitites, and hence the inherent nature of relationships between entities.
I don’t know where they come from, but I think I know where they don’t come from.
One more thing: was there a “law” governing your making this post? If so, was the law responsible for the post and not you? After all, if it was *governing* you, I would tend to blame the law!
Maybe it has something to do with his nature….
No innovation equates to no productivity gains? That’s ridiculous. There are productivity gains even though the equipment is depreciating. People naturally become more efficient at performing a task and you have productivity gains.
Of course businesses can predict economic behavior. How else can they order inventory?
Lol, with efficiency at 100% there can be no improvement without innovation… How about that I foolishly assumed 100% efficiency.
“I hope not. No matter how much “experience has confirmed” an empirical regularity in the social sciences, people still have free will (at least operationally, if we’re doing a social science rather than looking at them as collections of atoms) and so those forecasting models could go out the window tomorrow.”
Yes. Successful statistical studies in the social sciences are descriptions of the “laws” holding, GIVEN certain institutional, or customary, or habitual practices. Or as Terry Nardin put it more simply, they are “disguised descriptions of a practice.”
And such descriptions may be very useful. But they are not like the law of gravity.
“if I logically deduce that, “The purchasing power of money will be lower, other things equal, when the stock of money increases,” then that is a genuine law of economics”
What is the epistemological status of this proposition? Analytic a priori? If the latter, it tells you nothing epistemologically *necessary* of the real world.
E.g., even with your ceteris paribus assumption, an economy with administered fixprices might show no increase in prices even if the money supply increases.
“E.g., even with your ceteris paribus assumption, an economy with administered fixprices might show no increase in prices even if the money supply increases.”
The purchasing power of the money is not only relevant to prices. Why do you insist that fixprices (even if they never were renegotiated or modified) invalidate this reasoning? They don’t. If every single price is fixed, but there is simply less production, or an increased amount of money with the same production, the result would STILL be a lower purchasing power of money – people would simply end up with money they cannot buy anything with, as there is nothing to buy.
It’s synthetic a priori, since it derives from the nature of action.
Is action defined to be purposeful behavior, or is it given an independent definition, or is it an undefined term that humans are supposed to understand via intuition?
Independent definition? How would it be connected to action if it were independent of action?
By “independent” I meant independent of the action axiom. My question was, is “human action is purposeful behavior” a definition of action, or does action have its own definition apart from that statement?
For instance, “a square is a rectangle with equal sides and right angles” is a definition of the word “square”, whereas “a square is a rectangle with a diagonal that is sqrt(2) times its side” is a statement about squares that is not simply a definition of the word “square” (although it may logically follow from the definition).
“If the latter, it tells you nothing epistemologically *necessary* of the real world.”
What is the epistemological status of this proposition? Analytic a priori?
The wage will be higher, holding other things equal, when marginal productivity is higher.
Two thoughts –
First, “holding other things equal” is pretty weak fare which is precisely why we need empirics to see what is and isn’t held equal in the real world and even more crucially how our independent variables (stock of money, marginal productivity) impact those things that aren’t held equal in reality.
Second, even if these were the sort of laws you claim they are, we still need to parameterize them just like physicists do. Our paramaters change over time. You seem to make a big deal out of this, and I’m not sure why.
When Einstein figured that some parameters of physics changed in different circumstances nobody questioned the scientific bona fides of physics – they tried to understand when and how they changed.
When Einstein figured that some parameters of physics changed in different circumstances nobody questioned the scientific bona fides of physics
What do you mean? That the mass of an object increases as it approaches the speed of light? That sort of thing?
Mass, time, yes – these parameters we thought were constant before aren’t. Big deal. We can try to understand how they change.
Daniel, I’ll risk to respond on behalf of Bob that he is making such a big deal out of this because some of the claims made by mainstream economists logically contradict such general economic laws and must thus be abandoned without trying to look at them empirically. As an example, I’d give Krugman’s idea of liquidity trap.
You’ll have to share what general economic law that violates, I think.
The idea of a liquidity trap violates the principle that money is not something that is valued for its own sake.
Well that’s a little vague and I’m not sure it’s an economic law whatever you mean by it. Do you think the store of value properties of money implies valuing money for its own sake?
The whole point of the liquidity trap is that at zero interest rates people become indifferent between two different assets. Surely that indifference doesn’t violate any economic law – I’d think it follows from one in fact.
This supposed indifference only comes from an implicit assumption that money is valued per se. Then, indeed it may be deduced that buying bonds from people isn’t going to get them to spend the money they get.
If, however, one recognizes that people hoard money to buy something in the future then the whole liquidity trap evaporates.
re: “If, however, one recognizes that people hoard money to buy something in the future then the whole liquidity trap evaporates.”
OK now I am thoroughly confused. It sounds like you are just describing one of the principal reasons why we worry about liquidity traps.
Do you think that, say, Krugman would disagree with you on the first half of that sentence?
“When Einstein figured that some parameters of physics changed in different circumstances nobody questioned the scientific bona fides of physics – they tried to understand when and how they changed.”
Not a very good analogy. Einstein discovered that the timeless, changeless laws of physics are not quite what we thought they were, not that they change over time!
I feel like you’re saying the exact same thing.
If you have a full understanding of parameters that are different in different contexts how is that different from saying that there is one set of fixed parameters but that they are different in different contexts.
I think you’re just trying to disagree here – we’re saying the same thing. The point is, when Einstein noted (as in economics) that some things are not as invariant as you might think they are, that was not cause to jettison the idea that what we were dealing with was science.
Another way of putting this is that Einstein’s contribution was to endogenize certain parameters that were previously considered to be exogenous.
Daniel,
I agree with Gene. Relativity and quantum theory are exquisitely verified (not “proved”) with experimental observation and astronomical observation. There is nothing even in the same ballpark in economics.
Physicists think there is a simple set of laws governing the behavior of matter. It’s true, they occasionally refine their understanding of what the laws are. But once they think they’ve got them, the laws themselves are stable.
There is nothing like that in economics. If you find some pattern in stock price data, you have no reason to expect it to continue. The Pareto distribution of wealth or income doesn’t actually hold for different countries exactly; it’s got huge error bars.
Bob,
I’ve been following the methodology discussion here with interest, as it is closely related to some of the key arguments in Uncontrolled.
I think the more productive challenge to put to proposed economic laws is “Do they make useful, reliable and nonobvious predictions for the effects of proposed interventions?”
All the best,
Jim
Thanks Jim. For what it’s worth, I think you understand the problems with these issues better than anybody I’ve heard on it. I’m hoping that if you learn anything from my discussions, it’s that we do have a source of knowledge in economics that we can get by introspection; it’s not just that we have to be super duper careful in setting up our control groups etc. in empirical studies.
Am I right to say, from reading your post, that all of these claims are methodological ones and not philosophical ones? Like human action may very well be deterministic – but that’s not what the economist is interested in – because it very well seems as if we have choices and if we’re going to make abstractions about human action, we should take that as an assumption. Likewise, as you say for the natural sciences, maybe nature isn’t deterministic, but the scientist is uninterested in that because it very well seems as if it isn’t, and that is the better assumption to make.
Why would you turn to what physics does as the yardstick for science anyway? I don’t understand that decision.
To get more concrete – is this woman a scientist?: http://www.nature.com/scitable/knowledge/library/primate-sociality-and-social-systems-58068905
If not, why not. And if she is, why isn’t economics?
Daniel, are you really this far into our debate and you still think my position is, “Economics isn’t a science”?
I am saying that the criterion for a “good theory” in economics should not be, “Accuracy of falsifiable predictions.” I am completely repudiating Friedman’s famous essay on this topic.
But, I still think economics is a science. I claim that the logical, deductive approach is appropriate for this particular science of economics.
My definition of science and the rest of the world’s definition of science.
OK, if you want to use your terms then change my last comment to: “is this woman doing appropriate empirical work, if not why not, and if she is why aren’t economists doing appropriate empirical work”.
Or simply – would you level the same complaints against her that you do against economists, and if not, why not.
Or a little more playfully: why is there no chimp praxeology.
“Or a little more playfully: why is there no chimp praxeology.”
There certainly is! Ethologists make praxeological assumptions about the beings they study all the time. (E.g., “The crow bent the wire *in order to extract the food from the bottle*.”)
“but the assumptions underlying [QM] literally defy comprehension.”
No they don’t. I comprehend them. Whitehead comprehended them.
Gene, are you just being a smartass? I’m sure you’ve heard this famous quote from Feynman?
“There was a time when the newspapers said that only twelve men understood the theory of relativity. I do not believe there ever was such a time. There might have been a time when only one man did, because he was the only guy who caught on, before he wrote his paper. But after people read the paper a lot of people understood the theory of relativity in some way or other, certainly more than twelve. On the other hand, I think I can safely say that nobody understands quantum mechanics.”
Too bad he died before meeting you, Gene!
Feynman never read Thomas Aquinas; that was his problem.
If Feynman never read Aquinas, that was indeed a problem, but that probably wouldn’t have helped him understand quantum theory.
Who has been defending Friedman?
Who has been defending Friedman?
Maybe we need to just walk away before someone gets hurt, Daniel. In my mind, you have been defending mainstream economics in exactly the way Friedman described it in his classic essay.
Well, which elements of his claim.
I quite obviously haven’t been saying that realism of assumptions is not important and predictive capacity is.
I post all the time about the dangers of the view that predictive accuracy is the test of good science – that’s been exactly my position throughout this discussion.
Prediction is, at best, a nice byproduct of science. But it’s not guaranteed and in a science of a complex system it’s even less guaranteed.
If in normal times we can get a decent sense of how things will go a quarter or two out, then that’s nice.
DK wrote:
I post all the time about the dangers of the view that predictive accuracy is the test of good science – that’s been exactly my position throughout this discussion.
Prediction is, at best, a nice byproduct of science. But it’s not guaranteed and in a science of a complex system it’s even less guaranteed.
Too late, someone has been hurt–me. I just banged my head on the keyboard.
You can’t seriously think this is something I’m making up right now. Opposition to Friedmanite views on prediction is the bedrock of all of my arguments on the relationship between prediction and science.
Am I taking crazy pills here?
People know this right? Other commenters here that are familiar with my blog?
DK okay, but if you think the realism of assumptions is more important than predictive power, I’ll flip it back on you: You don’t think Heisenberg was a very good scientist, right? I mean, sure quantum theory can predict experimental observations better than any theory yet discovered by mankind, but the assumptions underlying it literally defy comprehension.
Huh? The fact that I don’t like Friedman does not mean that I endorse the exact opposite.
Science is producing claims that help us understand the world that are disciplined by empiricism (and in most cases peer review – although I suppose that’s not strictly necessary).
Useful understandings of the world require accuracy in assumptions and accuracy in the conclusions of the model. Those conclusions can be either explanations or they can be predictions, depending on the science.
Any theory that is more accurate on both assumptions and conclusions is obviously preferred to one that is less accurate. But I have no priors on where we should focus most of our attention. Presumably it changes over time because there are diminishing returns. If we have really great, accurate conclusions then we should probably focus some attention on improving the accuracy of our assumptions.
Now – Friedman’s mistake is twofold: first the strong preference for accuracy in conclusions. Second, the weird preference for conclusions that come in the form of predictions over conclusions that come in the form of explanations.
What I talk about repeatedly when these issues of scientific legitimacy come up is that in sciences of complex systems you’re almost always going to have more luck with explanation than prediction. And, contra Friedman, that’s just fine.
There really isn’t anything weird about Friedman’s preference though Daniel. Let’s talk Planck not Heisneberg. He came up with a way of making predicitions that worked but that confuted all his (and anyone else’s) understanding at the time. This drove towards a new understanding. That new understanding would never have arisen had everyone said, “bah, no explanation, unrealistic, stick to the old well oied theory with the wrong predictions.”
I don’t think Friedman ever argued “you get a good prediction you’re done”, only that good predictions have to be the arbiter of last resort.
Another example. Lamarckism is more explanatory and simpler than natural selection. It was a good clever candidate theory. Just wrong. Or one of my favorites”preformationism” which is neither foolish nor absurd but actaully a brilliant imaginative leap. But one quickly refuted.
Or stick to economics: I can give an explanation of what happened from 2008 – 2013, and so can Paul Krugman. How do we referee between these rival theories? I know Krugman’s answer. Is he not “part of the rest of the world,” Daniel, when it comes to understanding how empiricism works in economic science?
Ken B – I am less concerned with the preference. Different people worry about different things. When these people get together the result is better for it.
The problem I have is a methodology that defines prediction as the be-all-end-all. Friedman’s claim is fine if you’re someone that is personally bugged by that. It’s less defensible as a totalizing definition of science.
I should reread Friedman before making too strong a claim. Maybe he couched his language, I forget. I’m more reacting here to this sense of Friedman’s positivism that everyone has.
Bob –
On refereeing theories of course there’s a question of how well each explanation fits the 2008-2013 data. They may both be able to explain the period but what does that mean – potentially one idea explains a lot more of what’s going on than the other. If they both explain different things that are going on, maybe it’s not a question of arbitration at all – they could both be right.
Then of course you can ask how realistic each of your assumptions are.
Then you can test out of sample – does one idea ONLY explain this incident while the other seems to characterize many incidents? The latter is probably, although not necessarily preferable if we think there are structures and patterns to the universe.
Presumably there are a lot of other things to invoke for arbitration, but these come to mind.
And then of course there’s prediction – forecasting. I put much less weight on this since we are dealing with such a complex system, but if one can forecast a lot better than the other that’s still telling I think.
Daniel, I realize I keep saying I’m done, but your further posts make we want to clarify something: What you just described is the OPPOSITE of what Krugman says. He is proud of the little IS-LM model that could. Why is he proud? Because it’s very simple–with widely unrealistic assumptions–and yet in terms of predictions, it came through this crisis with “flying colors.” That’s the criterion by which he says he has been right, and the austerians are dead wrong.
So, you need to stop saying that you and the rest of the world are saying one thing, while the Misesians are in the corner. Your own view is not how most economists would describe their work, including Paul Krugman and Brad DeLong.
@DK – fair enough. I need to reread Friedman to be sure too. And I’m partly reacting to the sense of anti-positivism so common on FA.
Bob –
I don’t know if Krugman agrees with me on science or not.
BUT – I might need to reread but I don’t think he was proud of unrealistic assumptions. In fact I’d guess he thinks the assumptions are quite realistic.
Just thinking predictions are a nice way of demonstrating the strength of a theory (something I accept) is NOT what Friedman is saying.
I don’t know why you are acting like there are these two options here. You can think predictive power matters without taking Friedman’s position.
IS-LM is also simple. If we can parsimoniously isolate the factors that are most important AND it predicts well AND explains well (explains situations like the Great Depression, for example which happened BEFORE IS-LM was developed), why shouldn’t we get excited?
New Keynesian IS curves add some bells and whistles that are probably good to add but it speaks well of the theory that it performs well in a simplified version too.
“Lamarckism is more explanatory and simpler than natural selection. It was a good clever candidate theory. Just wrong.”
No, it turns out that Lamarckian evolution does occur. Got work to keep up with the sciences, Ken.
It’s disputed Gene, and it’s mostly on the fringe even there.
It could occur if you see an association of different genomes forming a new unit. (hologenes).
Actually the one best possible example might be the biggest: eukaryotes arising from a microbial invasion. Even there you can explain it with natural selection.
I have seen discussion of possible challenges to the Central Dogma but I don’t think they are widely accpeted.
Daniel Kuehn,
Last thing and I have to walk away: In this very thread, you have said
(1) You and the “rest of the world” define science as making falsifiable predictions from a theory, and then looking at the actual observations to see how well the theory did.
(2) You are shocked that I think you believe the accuracy of falsifiable predictions is how we should judge a scientific theory.
Can you at least understand why I have no incentive to continue this discussion?
1. “You and the “rest of the world” define science as making falsifiable predictions from a theory” I’ve never defined it that way. Show me where I have if you think I have.
2. See 1.
Daniel, you are being needlessly nit-picky. You said you define science the way the rest of the world does. The rest of the world defines science, basically, as “making falsifiable predictions from a theory.” From Wikipedia:
“The strength of a scientific theory is related to the diversity of phenomena it can explain, which is measured by its ability to make falsifiable predictions with respect to those phenomena.”
Ergo, you say you define science as such in one breath and then deny it in another, and are frustrating Bob with your cognitive dissonance.
When I look at Wikipedia’s science page I don’t see that sort of weighting towards prediction that you have in this quote (was this from further down? I didn’t read the whole thing).
It’s the strong weight placed on prediction that I have a problem with, that is not nit-picky in a discussion like this – it’s a major point, and I am happy to disagree with Wikipedia over that specific point.
I am 100% fine with the initial definition Wikipedia has under “science”. It doesn’t make nearly as much Popperian assumptions as you make here (and it shouldn’t).
Obviously if you’re with Popper or if you’re with Friedman those more specific idea can fit into this definition of science. But you cannot argue that those more specific ideas present an uncontested definition.
http://en.wikipedia.org/wiki/Scientific_theory
‘The rest of the world defines science, basically, as “making falsifiable predictions from a theory.”’
No. This idea was dead in the philosophy of science 40 or 50 years ago. And scientists never worked that way, even if they may have given the view lip service.
Just because a Popperian got a hold of the page does NOT mean that is a good understanding of what science is!
And it’s probably generous to even think that a Popperian got a hold of the page. Some of these claims are just “in the air” because of Popper even for people with no real dedication to Popper.
I’m sure you could find tons of scientists who do science in very non-Popperian ways that intinctively make Popperian claims about “what science is”.
Science is what scientists do.
That said all the scientists I know of accept that predictions are essential.
Clearly he is learning a lot from your Krugman posts.
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
prediction ≠ explanation
I’m going to declare Daniel a Misesian and move on with my life. Winning!
Is Mises joining the empirical science camp or am I leaving it?
I only ask because I’m not doing the latter and Mises is dead, which leaves me a little confused.
2DK,
Can’t reply to you in the thread, so will reply here.
=It sounds like you are just describing one of the principal reasons why we worry about liquidity traps.=
I think Krugman isn’t just saying this by liquidity trap. He also means that this won’t change unless consumers become more optimistic. Thus, since deflation causes some decline in production they can only become less pessimistic which strengthens the trap, ad inf. But if people hoard money to buy something in the future and see instead shrinking production they’ll start spending. The supposed trap doesn’t exist.
I would just like to say that you just BLEW MY MIND when you proposed the possibility that nature has free will.
I would just like to say that you just BLEW MY MIND when you proposed the possibility that nature has free will.
You’re welcome. Let’s push it further: What it would mean for Nature to suddenly deviate from the “laws” of the material universe, especially if it happened in a way that seemed to exhibit purpose, would be called “a miracle.”
And so, every sentient being is in that sense a miracle. There are billions of miracles all around us, but we’re used to it so it seems like nothing.
In some sense I disagree with Bob. I think prediction is the final arbiter of a good theory. But he makes a good critique here and in the other post. Here’s what I see as his key point:
” if you can’t predict very well, then using the empirical method in constructing your counterfactual is very dubious.”
Indeed. And empiricaly mainstream economics has a pretty mixed record.
I think one can accept the force of this argument, which sound Hayek like to me, without abandoning the notion that science must be empirical. All you need do is accept that economics hasn’t got there yet.
This is my take on the earlier thin thread.
Most people either don’t think much about what makes science science or they implicitly agree with something like my view (they may parrot some Popperian buzzwords that might make you think otherwise, but if you looked at what they did or pressed them on it I think that would fall apart pretty quickly).
This view is that there is no clean formulaic definition of what a scientific theory must be. It’s more loosey-goosey. Science must be careful and methodical in explaining phenomena, it must be empirical, and those two things must work together iteratively to constantly improve theory. A scientific theory can therefore be assessed by: (1.) realism of assumptions, (2.) realism of proposed theoretical mechanisms, (3.) ability to explain past phenomena, (4.) ability to predict future phenomena. All of these are great. The fourth one is truly impressive if it is to be had because it cleans out a lot of the cognitive biases that might plague the third one. But there’s no clear way in which one trumps the another. Rather, these are all points that scientists care about and when scientists try to persuade other scientists these are the things they invoke.
So that’s my view and I think implicitly that’s most people’s view.
Friedman’s positivism (or I should say what everyone takes to be Friedman’s positivism – I’d have to reread to see if this is truly fair) says that #4 is what matters, the rest be damned. I guess Popper kind of falls under this, but he’s more of a very strict adherence to either #3 or #4 that doesn’t allow for any wiggle room or recognition of the problems with falsifiability in the real world.
Mises is on the other extreme and says that only #1 and #2 matter, that if you take care of them #3 and #4 take care of themselves and if #3 or #4 contradict #1 or #2 either you’ve done your logic wrong or you’ve done your empirics wrong. He can’t conceive of empirics trumping “good” theory.
If you want to dispute my claim that most people are implicitly in agreement with me, you’re welcome to – although I fully acknowledge that Popperian buzzwords slip into conversation. The fact is, though, when you look at how science is actually done it’s done my way.
There is nothing new at all about this for me. I’ve always wailed about people giving #4 primacy over #3. You might not have seen me do it, but I am positive you’ve never seen me argue that #4 beats the others. And if you think you have, you need to quote me on it.
What I said about Friedman goes for Popper and Mises too btw. I am working off of the sound-bite version of what they said. I don’t pretend to have looked into them in any great depth. If the sound-bite isn’t right I’m more than happy to learn something about what they actually said. What I care about is the argument. If they are actually more in agreement with the position I’ve staked out here, fantastic – I’m a happy camper.
And instead of “buzzwords” I should probably say “buzzphrases”
Are 1) and 2) empirical questions?
Regarding the dispute over realism of assumptions, versus accuracy of predictions, this quotation is relevant:
[T]he purpose of a model is not to be realistic. After all, we already possess a model that is completely realistic—the world itself. The problem with that “model” is that it is too complicated to understand. A model’s purpose is to provide insights about particular features of the world. If a simplifying assumption causes a model to give incorrect answers to the questions it is being used to address, then that lack of realism may be a defect. . . . If the simplification does not cause the model to provide incorrect answers to the questions it is being used to address, however, then the lack of realism is a virtue: by isolating the effect of interest more clearly, the simplification makes it easier to understand.
–David Romer, Advanced Macroeconomics (New York: McGraw-Hill, 1996), pp. 11–12.
Exactly how moon landings were calculated. It’s wildly unrealistic to treat the capsule as a point particle.
Exactly how DNA sequencing is done. It is wildly unrealistic to see complex molecules in a sea of random motions as a sequence of letters.
Exactly how epidemiology is done. It’s wildly unrealistic to think of people as simple parametrized units that can carry a contagion.
>>This is of course Hume’s famous problem of induction. But the physicists and chemists can shrug at the philosophers and say, “You go ahead and argue about why it works, but it clearly does.”
We learn when we find ourselves in error. We don’t learn *anything* from induction. That’s a myth.
The better we can expose our theories to potential error, the better we can find out where we are wrong. This is also the ethically correct thing to do, because better exposing our theories to error means better submitting them to another’s independent judgement. We want people to decide the truth for themselves not to rely on our authority.
To put things another way, no one can justify a theory for you. Whether you regard a theory as true our not is aways a judgement call on your part – no one can force your hand in that matter. Science should not be about *justifying* one’s opinion, but instead about trying to put our theories out into the open where they can be scrutinized.
Induction was merely a myth that attempted to explain a new phenomena. People didn’t want to have to respond to the authority of the king, so instead they argued they were responding to the authority of nature. Induction was supposedly the process where by this took place. However, who is to say when we begin to believe in a theory? Before or after we test it? This is merely a personality issue, and each person is quite different. How many theories have we believed in the past that turned out to be incorrect?
We don’t need induction. In the end induction is for socialists – because they have all the answers, and they know they are right. Socialist don’t understand why there is value in having a free “market place” of ideas.
It’s funny this topic is getting so much attention. I’ve recently dedicated a lot of time to looking at methodology and arguing the postive case for the “Dualist” methodology when it comes to science (Physical/Natural world versus the Social world of Human Action). It just so happens that this coincides with the recent David Friedman vs. Bob Murphy debate on this topic.
Up until the 1950s, the methodology of economics recognized this dualism and was abandoned with the help of Friedman’s essay (1953) and others in the logical empiricist camp. Looking at the state of affairs of mainstream economics and it’s utter failure to explain or predict the recent troubles in our economy, methodology is (imo) at the very root of the problem.
If anyone is interested, below is a link to my case for the dualism in the sciences and for the Misesian methodology in economcis and the social sciences.
http://www.libertyforlaymen.com/2013/09/austrian-school-methodology.html