" And the reason why I made that conversion is important. Libertarian insights in a lot of ways are basic, Econ 101 insights about the efficiency of free contracting writ large and converted into norms or political values. In other words, I think one of the most essential libertarian fallacies is building a politico-ethical system around positive social science findings (and, I want to stress, basic and introductory social science findings at that). It's kind of an odd way of going about formulating a politico-ethical system. We don't adopt Nietzschean super-man ethics because of evolutionary biology, and we shouldn't simply adopt libertarianism because of these insights. I want to be clear - my point is not that you have to mix up normative and positive findings to come to libertarianism. My point is only that it's possible to get everything there is to get out of libertarianism simply by improving people's knowledge of social science. This is only to say that it's not entirely clear to me what should be important here: teaching people more social science, or sharing libertarianism. But even that isn't entirely satisfying - after all, the reason why I abandoned libertarianism was because I kept learning social science. Yes, the market is efficient and the price mechanism leverages decentralized knowledge. But if institutions don't or can't internalize costs and benefits social scientific insights start to militate against the efficiency of markets. Uncertainty and imperfections ensure that market forces, as fantastic as they are, are going to remain sub-optimal. I haven't abandoned any of the introductory insights in adopting these views - the complement the introductory insights that I still use. I still have a relatively contractarian view of human relations. I still take a fairly atomized, individualist view of things. I still come down on Hayek and Mises's side of the socialist calculation debate. But I can't call myself a libertarian. So, if what we really want is to get people to take the implications of social science more seriously, then its not clear that that would move people towards libertarianism either." -Daniel Kuehn
I think this is an interesting insight, though I find issue with it for this reason: Did Rothbard not "keep learning social science"? What about Walter Block, or other career Austro/Libertarians? Is there an implication that in order to maintain Libertarianism, we must stop at elementary observations? I'm sure DK can clarify if he wishes, and I hope he does. Anyway, thoughts?
Good job Student. You took the answer right out of my mouth :p Thanks DK btw for taking the time to post here. Bloggers like you, Selgin, and Friedman are the main reason I come here.
You don't worry about a priorism because we're not doing logic - we're doing science and we're trying to get a workable explanation of the world - not a solution to a logic problem.
So, by "workable explanation" you mean a model that is not relatable to anything else other than the model itself. For instance, I have a model of the solar system. It doesn't need to fit the laws of physics. It just needs to be an animation of planets spinning around the Sun.
Daniel Kuehn:The problem is I have trouble bridging that disagreement if I'm not clear on why you disagree or why you raised those issues in the first place.
I'm not clear myself on why I disagree, because I don't fully understand your position. Until I'm convinced that matters such as quantification and inter-subjective utility comparisons are extraneous, based on your explanation, I'm going to be skeptical of that notion. That only makes sense, doesn't it? With all due respect, I'm not going to just take your word on it as a matter of faith.
Daniel Kuehn:You all post A LOT, and each post has several questions in it, some of which I don't know how to answer because I don't understand what you're driving at ("i.e. - 'but are there any constraints to the optimization'? I don't know what to say other than 'of course' because I have no sense of why you would think there wouldn't be). So please, don't assume I'm ignoring it and if you think it's absolutely essential to answer, just ask again. I do care - I like interacting with Austrians. But I can't reargue first principles every time I do. If it's necessary for clarification, sure.
Sorry, I wasn't assuming you were ignoring it - I used the word "seem", indicating that my initial impression could well be wrong. Anyways, I'll be happy to post again or just give you links to the posts which you missed.
Have you already argued first principles somewhere else? I'll be happy to read that material also.
Regarding confusion over words like "optimization", I admit that I'm not a trained economist - I'm a software engineer. So as I indicated before, my experience with words like "optimization", both professionally and non-professionally, is that they're typically treated in an objective sense. For example, "X is the optimal way to do Y" is intended as a completely objective statement (whether it really is is another story).
The keyboard is mightier than the gun.
Non parit potestas ipsius auctoritatem.
Voluntaryism Forum
Caley -
If you have a law no model should contradict it. I don't really have a problem with your action axiom. There are a range of accounting identities that no good model ought to contradict. But there's a lot that we don't have solid deterministic laws for. This is the nature of any complex science - it's true of biology too, and that's a better thing to compare it to than physics.
So what do we do with this lack of knowledge and lack of laws? Should we just reject a formalization of anything and work deductively from the few laws we do know? You all talk about scientism often. Well that sounds like "philosophism" to me. You're trying to mimic what philosophers do because you know we don't have what physics has to work with.
An alternative is to model reasonable theories consistent with the laws we do know and test how well they match the data. This will not allow you to pierce that veil into incontrovertible truth. It will give you very strong theories to understand how they economy works.
autolykos -
Nobody is asking you to take anything on faith. I'm saying that if you're skeptical of something and expect me to answer to that skepticism you ought to be able to articulate exactly what it is that I'm answering to. You're the one that brought all these issues up, after all.
re: "Have you already argued first principles somewhere else?"
Nothing I know of to grab for immediately, but word searches on my blog are your best bet. I think the skeleton of my point on first principles has been laid out in this very comment section and a lot can be inferred from what I've said here. What exactly is still unclear?
As for economist lingo - essentially no modern economist isn't a subjectivist and any reference to optimization is a reference to constrained optimization. I really can't think of anyone that maintains anything other than a subjective value theory and constrained optimization. So you're pretty safe assuming both.
Unconstrained optimization is essentially meaningless in economics and I have never seen it discussed or done by anyone in the decade that I've been either an economics student or economic analyst.
Well, I wouldn't go that far. There are plenty of applications where unconstrained optimization is appropriate. For example, profit maximization is a straight up unconstrained optimization problem (take first derivative of profit function set equal to zero to get first order equations, take second derivative to check second order constraint). http://publish.uwo.ca/~imedovik/OptimizationInEconomics.pdf You can set up as a constrained optimization problem, but the since the constraints always hold, it simplifies to an unconstrained optimization problem (explained very briefly here: http://www.econ.brown.edu/~rvohra/econ113/profit.pdf).
Another example that comes to mind is in econometrics. You get the normal equations for an OLS regression using unconstrained optimization: http://www.yongyoon.net/econometrics/ols_derivation_1.pdf
But I dig where you are comming from. In the context of this discussion (maximizing individual utility or social welfare), you are right there is really no imaginable situation where you would need to worry about unconstrained optimzation.
Ambition is a dream with a V8 engine - Elvis Presley
re: "For example, profit maximization is a straight up unconstrained optimization problem (take first derivative of profit function set equal to zero to get first order equations, take second derivative to check second order constraint)."
The constraints, I would have thought, would have been the cost function. We can talk about anything as an unconstrained optimization if you approach it like this, can't we? For example, "maximize utility subject to your budget constraint" could become "maximize net utility (or whatever the utility-equivalent of profits would be)".
The math may or may not be constrained depending on whether you stuff your constraints in your objective function or not, as you point out. Perhaps it's better to put it this way: there is no optimization done in economics that could not be specified as constrained optimization by separating out legitimate economic constraints like "costs" or "budgets" from the objective function. I think that should work :)
I disagree with the general perception, which seems to be shared by Austrians and otherwise, that physics is more complete than economics. It seems that economists are biased in favour of "natural science" because they are less familiar with it, like a case of the grass is greener on the other side or something.
Daniel Kuehn: Caley - If you have a law no model should contradict it. I don't really have a problem with your action axiom. There are a range of accounting identities that no good model ought to contradict.
If you have a law no model should contradict it. I don't really have a problem with your action axiom. There are a range of accounting identities that no good model ought to contradict.
As one can glean from Bob Murphy's recent MMT posts, economic laws, as Austrians conceive them, aren't about accounting identities.
Daniel Kuehn: But there's a lot that we don't have solid deterministic laws for. This is the nature of any complex science - it's true of biology too, and that's a better thing to compare it to than physics.
But there's a lot that we don't have solid deterministic laws for. This is the nature of any complex science - it's true of biology too, and that's a better thing to compare it to than physics.
The section of my article that mentions weather patterns deals with this faulty comparison.
Daniel Kuehn: So what do we do with this lack of knowledge and lack of laws?
So what do we do with this lack of knowledge and lack of laws?
Lack of laws? Economic theory, as Austrians conceive it, is entirely comprised of laws.
Daniel Kuehn: Should we just reject a formalization of anything and work deductively from the few laws we do know?
Should we just reject a formalization of anything and work deductively from the few laws we do know?
Working deductively from a few laws is how we discover further laws. Do you think a systematic exposition of what Hayek called "the logic of choice", spelling out the necessary implications of what we mean when we posit certain choice-laden phenomena is impossible?
Daniel Kuehn: You all talk about scientism often. Well that sounds like "philosophism" to me. You're trying to mimic what philosophers do because you know we don't have what physics has to work with.
You all talk about scientism often. Well that sounds like "philosophism" to me. You're trying to mimic what philosophers do because you know we don't have what physics has to work with.
William James defined philosophy as "an unusually stubborn attempt to think clearly." Don't you think that is a worthy goal for both the natural and the social sciences? And if Mises was right that taking certain approaches to the social sciences is to simply not think clearly about the nature of human affairs, wouldn't it make sense to avoid such approaches, and adopt more fruitful ones? Isn't it interesting that almost all non-controversial economic laws (even among mainstreamers) are the ones discovered primarily through discursive reasoning, and not through hypothesis and experimentation?
Daniel Kuehn: An alternative is to model reasonable theories consistent with the laws we do know and test how well they match the data. This will not allow you to pierce that veil into incontrovertible truth. It will give you very strong theories to understand how they economy works. With human action, any theory can match any data, if the interpreter simply posits countervailing factors. "the obligation to justice is founded entirely on the interests of society, which require mutual abstinence from property" -David Hume | Post Points: 20
Daniel Kuehn:An alternative is to model reasonable theories consistent with the laws we do know and test how well they match the data.
And the "laws we do know" are determined how exactly?
Daniel Kuehn:re: "In other words, you're refusing to make explicit examples of what you mean by those terms. How nice. It's too bad you couldn't even be explicit about this refusal." Jesus! Fine. An example of "objective value" would be to argue that diamonds have a higher price than bread because they take more labor and capital to produce. An example of "subjective value" would be to argue that while bread is more valuable to a person an additional unit of bread is less valuable than an additional unit of diamonds because bread is plentiful, and that the price is ultimately determined by the marginal value of these subjective preferences. An example of constrained optimization is to think of a case where someone has different preferences but a limited budget and must make purchasing and saving choices that maximize their utility subject to the budget they have available. An example of unconstrained optimization is to think of a case where someone has an infinite amount of money and sets out to satisfy all his desires with it. Unconstrained optimization is essentially meaningless in economics and I have never seen it discussed or done by anyone in the decade that I've been either an economics student or economic analyst. There are your examples. Now do you see why I said it's easier to just say "we agree"? No, because we don't (completely) agree. But thanks for your examples just the same. I see no issue with your examples of constrained vs. unconstrained optimization. My disagreement lies with objective vs. subjective value. It seems to me that your example of subjective value still contains two notions of objective value. First, there's the notion that the lower marginal value of bread w.r.t. that of diamonds is necessarily due to bread being more plentiful. Second, there's the notion that the price (of bread?) is necessarily determined ultimately by the marginal values of these (two?) subjective preferences. The keyboard is mightier than the gun. Non parit potestas ipsius auctoritatem. Voluntaryism Forum | Post Points: 5
Jesus! Fine.
An example of "objective value" would be to argue that diamonds have a higher price than bread because they take more labor and capital to produce.
An example of "subjective value" would be to argue that while bread is more valuable to a person an additional unit of bread is less valuable than an additional unit of diamonds because bread is plentiful, and that the price is ultimately determined by the marginal value of these subjective preferences.
An example of constrained optimization is to think of a case where someone has different preferences but a limited budget and must make purchasing and saving choices that maximize their utility subject to the budget they have available.
An example of unconstrained optimization is to think of a case where someone has an infinite amount of money and sets out to satisfy all his desires with it. Unconstrained optimization is essentially meaningless in economics and I have never seen it discussed or done by anyone in the decade that I've been either an economics student or economic analyst.
There are your examples.
Now do you see why I said it's easier to just say "we agree"?
No, because we don't (completely) agree. But thanks for your examples just the same.
I see no issue with your examples of constrained vs. unconstrained optimization. My disagreement lies with objective vs. subjective value. It seems to me that your example of subjective value still contains two notions of objective value. First, there's the notion that the lower marginal value of bread w.r.t. that of diamonds is necessarily due to bread being more plentiful. Second, there's the notion that the price (of bread?) is necessarily determined ultimately by the marginal values of these (two?) subjective preferences.
Daniel -
re: "As one can glean from Bob Murphy's recent MMT posts, economic laws, as Austrians conceive them, aren't about accounting identities."
Absolutely - and I commended Bob on that post. My point to Caley is simply that because I think we can formulate theory without deriving it from laws, that is not to imply that the theory we formulate can contradict laws. Of course it can't. Caley seemed to confuse my point that it doesn't all need to be derivation from laws with some assertion that I'm saying that "it doesn't need to fit" those laws. That was poor logic on Caley's part.
re: "Lack of laws? Economic theory, as Austrians conceive it, is entirely comprised of laws."
Relative lack of laws, I ought to say. We don't have laws for everything we are interested in having laws for, and no derivation from the laws we do have can get us there.
re: " Do you think a systematic exposition of what Hayek called "the logic of choice", spelling out the necessary implications of what we mean when we posit certain choice-laden phenomena is impossible?"
Not impossible. Misleading, potentially. We know very few hard and fast biological laws. Imagine if biology were restricted to what we could derive from those laws, and then imagine an institutional hostility to explanations that were not derived directly from those laws. We would have a very odd set of biological conclusions. If you want to just do a priorism all day, that's your choice. But when you approach people who don't do a priorism - who instead use the scientific method as it is traditionally understood - your stance toward that work ought to be no more severe than epistemological agnosticism. You don't know if what we're doing is "true" but it could be. I'm willing to go one step further and say <i>"it could be true and it seems to work pretty well so I'll use it to structure the way I understand the world"</i>. You may not be willing to take that step. But you can't discard these conclusions because they aren't derived from your axioms. That would be epistemologically inappropriate.
re: "Don't you think that is a worthy goal for both the natural and the social sciences? "
Of course! But science is also a stubborn attempt to think clearly. To agree that philosophy is valuable is not to get to the bottom of the question of whether economics is or is not science.
DD5:And the "laws we do know" are determined how exactly?
I brought this up earlier but Daniel has been ignoring my substantive posts under the guise of being forewarned about my online presence or something. I can't wait to pull that one in debate myself.
Uhm, I would love answer you but a faerie angel told me that you once stole penny candy from a blind midget, therefore I will have to ignore all of the points you raise that indicate my argument is completely without merit.
It's all just a series of non-sequiturs.
DK what about this point?
And if Mises was right that taking certain approaches to the social sciences is to simply not think clearly about the nature of human affairs, wouldn't it make sense to avoid such approaches, and adopt more fruitful ones? Isn't it interesting that almost all non-controversial economic laws (even among mainstreamers) are the ones discovered primarily through discursive reasoning, and not through hypothesis and experimentation?
My point to Caley is simply that because I think we can formulate theory without deriving it from laws, that is not to imply that the theory we formulate can contradict laws.
That is not what I meant. Contradicting and not fitting are not the same. You can theorize in a bubble without trying to connect to anything else. Contradiction is not necessarily apparent. They way scientific method is taught in school easily allows for that.