May 2016

Sun Mon Tue Wed Thu Fri Sat
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30 31        
Blog powered by Typepad

« A 420 Character Letter to the President | Main | The Poor are Not Getting Poorer »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Perhaps more important is to have statistics taught as epistemology - but couched in mathematical language. Otherwise people are effectively running regressions without knowing what they mean.

These links might clarify what I mean:



I guess a proper dichotomy could be having economics taught as two components: (i) philosophy (perhaps in the style of PPE programs) (ii) modelling - where economic philosophy is translated into mathematical language.

Models at the end of the day are only approximations of reality (as a famous statistician once said, "All models are false, but some are useful"). And the conclusions of your models can be heavily dependent on your modelling assumptions - in some sense, the conclusions of the model are already 'assumed' from the outset.

For instance, the neoclassical model - which concludes that tax cuts don't matter, Ricardo equivalence holds - such conclusions are only arrived at due to assuming that consumers have infinitely long lives and have perfect foresight.

Modelling can be a useful exercise (how else would people make forecasts?) - but its limitations should be recognized, and can never be a substitute for a broad 'philosophical' framework. Utility maximisation using Lagrange multipliers cannot match the breadth and depth of works like Adam Smith's and Wilhelm Ropke's.

Anon - the conclusions of more philosophical approaches are every bit as dependent on assumptions as the conclusions of mathematical models. These are two styles, but they hardly amount to a dichotomy - they're both integral and work quite well together.

Peter - fantastic post. I'm a little curious about how this phrase "engineer" can be misleading, though. Sen himself, for example, is not afraid of deliberate social action informed by social science, and there is nothing conflicting between deliberate social action and a philosophical disposition. I think the heart of the issue is that we don't want to train technicians who just learn how to push buttons in some social machine which we imagine more control over than we actually have. If the discipline churns out such technicians, we are not living up to our full potential. Keynes, for example, is a very philosophically disposed economist with substantial shelf-life. Is he an "engineer"? In the sense that he sees the prospect for deliberate social action, yes. In the sense that he is a technician, clearly no.

The word "philosophical" doesn't recommend itself.

The principle use of the word "philosophy" is for intellectual work producing intractable puzzles, due to dependence on some deeply embedded conceptual confusion.

Another use of "philosophy" is for efforts to "get the fly out of the bottle" -- identifying conceptual pathologies and moving beyond them.

I make an argument, from a Misesian utilitarian perspective, regarding the relationship between economics and ethics in the discussion thread "Why Capitalism?", especially in the following 4 posts:



Ed Glaeser has a post up that is very much in the same vein as this one:

I never liked Sen's work. The only way I have ever learned anything from him is by thinking through his errors and confusions.

Ed Glaeser's "The moral heart of economics" (Economix) is interesting. However, I think he needs to define freedom. As Hayek, Mises and many others have written, freedom is not the absence of rules; that would be chaos and the rule of brute force. Freedom is the rule of law through principles as opposed to the arbitrary rule of fickle men.

Under the rule of law, the law is applied equally to all people; the state doesn't favor one group of citizens over another. Applying law equally would forbid redistribution of wealth by the state and progressive taxation.

And freedom is freedom from something. People are free from fraud, theft and violence by fellow citizens, for certain, but also from the state. That means that the state should have limits as to what it can do to citizens. For example, there should be limits on how much the state can tax the people. The state does not have a right to 100% of people's incomes.

But fixing a percentage would be arbitrary. Why would 50% be better than 60%, other than that it is lower? Because 50% is less than 60% doesn't mean that the state has a right to 50% of income. There must be a principle that fixes the limits of the state in order to have the rule of law and not the arbitrary rule of men. The traditional classic liberal principle was that the people have the moral right to their property, as determined by God, and that the state should limit itself to taxation for the God-given purposes of protecting the life, liberty and property of its citizens.

And what principles do we apply? In the area of economics, the principles apply to property. Property is a moral good and taking it someone's property unjustly is immoral and called theft. That applies to the state as well as to other citizens. The state can commit theft, even if the state is a democracy.

So the moral foundation of economics is the right to property. Property equals control. Without control over one's property, one doesn't have property regardless of who owns the paper title. Freedom is the freedom to use and dispose of one's property, to control it, as one sees fit without coercion from others or the state. The state has a right to tax citizens in order to carry out its legitimate functions, but those are limited to protecting life, liberty and property.


You seem to have a very narrow view of philosophy. Historically, it has been an interest in answering the question of what makes for a good human life. To answer this, we need to know a few basic things: what is the nature of the universe? what is the nature of humans? what is knowledge? what is truth? what is beauty? what is virtue? Philosophy is love of wisdom -- in fact, if you break down sophos to "so" and "phos", it means love of the inner light. Mises starts with a prioristic knowledge -- paying attention to what the inner light tells us.

At the same time, shelf life doesn't necessarily tell you much. Rousseau, Marx, and Keynes all have shelf life. They do because their philosophies are great for those who wish to rule over others. Longevity doesn't mean either truth or virtue. One must still have judgment.

Troy -

You really think that's why Rousseau, Marx, and Keynes have shelf life? Think of which of Marx's ideas have shelf life, for one thing. Can his economics really be said to have "shelf life"? Is anyone really returning to that. Essentially, no. There are always true believers, of course, but no. Which of his ideas do seem to have "shelf life"? His philosophy? His social theory? I don't know enough Marx to say exactly but that's my impression. Exactly what has shelf life and what doesn't in Marx is actually an excellent demonstration of the value of the idea of shelf-life. It's not entirely sufficient on its own of course, but there's a reason why Marx's economics haven't had an especially impressive shelf life while his social thought has.

It does raise questions, though, about why conflicting ideas persist. This question seems to be bothering A LOT of people lately. I can't take your assessment of Keynes seriously, but you do raise a good point. Keynes has shelf-life as does Hayek. This appears to be a conundrum for some people, but it doesn't have to be. We need to get to a point where we recognize that our insights and claims are always partial, contingent, and imperfect but that they are useful. I can't make everyone like Keynes - but nobody should be surprised when they come across a person that sees value in Keynes, Hayek, Hume, Rousseau, Smith, Malthus, Buchanan, and Samuelson. I won't explicitly embrace Marx here, but I'm sure he's good for something.

I probably should have explained this more plainly: because our ideas are at best contingent, partial, and imperfect, then two ideas which seem to contradict when we forget that they are contingent, partial, and imperfect may actually coexist just fine when we remember they are contingent, partial, and imperfect.


Economics (PPE) philosophy is very different from the philosophy behind data interpretation, though both can be grouped under the same "philosophy" umbrella. The difficulty in grasping the philosophy of data interpreation/statistical modelling (epistemology) is that it can be very technical and thus made less accessible.

How many students really think about what "hypothesis testing" really means? (btw statisticians generally discourage people from using hypothesis testing, but these methods are still prevalent in other fields)

'Modern' economics is mainly driven by mathematical modelling, and claims that just because it uses empirical data it is superior to having a proper economic philosophical framework. But I believe this is mainly due to general ignorance about the philosophy of data interpretation/statistical modelling.

Peter's point has been brought up several times - I thought I'd chip in my idea, as no one has really suggested what I have said. And I believe it is a more pressing need, and ties in well with the fact that most econometric models have been proved to be failures pertaining to the current financial crisis.

anon -
I wasn't refering to epistemology. I was refering to PPE. My point is that there's no reason to assume the difference involves a dichotomy. Both PPE and modeling requires simplifying assumptions that can be problematic for conclusions that are drawn. That's not a distinction between them. Both are used in good economics. If all you're saying is that they're different ways of approaching problems, I'll obviously agree with you on that. I assumed there was some implicit juxtaposition - I may have been wrong to assume that. If you're not implying such a thing, then fantastic.

re: "How many students really think about what "hypothesis testing" really means?"

Right. Many people do economics badly. No argument from me on that point.

re: "ties in well with the fact that most econometric models have been proved to be failures pertaining to the current financial crisis"

Do you man "most forecasting models"? Let's not lump econometrics in with forecasting indiscriminately, please.

I guess my point was in favour of economics being taught in two components - both integral, and connected:

1. The first would be having economics taught as philosophy - Smith, Hayek... and economics is intertwined with politics, history, philosophy and so forth. Which is probably why economics used to be known as 'political economy'.

2. The second would be on mathematical modelling - the building of models of the economy. This would be contingent on your economic beliefs, so the first component would guide you in your modelling endeavours.

Like all mathematical models, these models are only approximations of reality - and having better assumptions can actually lead to worse modelling outcomes!

Models should be recognized for what they are - because they can only be approximations of the real thing, their applicability is limited. To what extent are the models useful - over what time frames, and how precise are the forecasts - these are usually very difficult questions.

In much of statistics, the honest answer in many situations is that we just cannot make any definite conclusions from the data. However, such conclusions do not allow you to publish papers...

Modelling can be a useful exercise - and to interpret empirical data, you cannot refrain from modelling activity. But it is an 'art' - which is not well communicated to economics students. The situation is better in Maths & Stats departments - but these usually come from hands-on experience... at my old university, there was only course dedicated to such aspects of applied statistics.

In most econometric models, mechanisms like price adjustments are not well-represented... and in-built in these models is a heavy inflationary-bias. These apply to the empirical data too, for instance, in GDP computation.

Granted, it is quite difficult to amend these problems but at least these in-built assumptions should be recognized.

I have searched for books on the philosophy of statistics (data interpretation) for a long time - there are only a few, and very technical. Much of it is hidden in specific case studies in journals and working papers.

It would be wonderful if an applied statistician could pen such a book one day, especially catering to a general audience (though it would be difficult to discuss these issues, without mathematical terminology).

But for now, I hope the links which I had posted earlier shed some light on this issue.

"both integral, and connected"

Definitely. Completely agree with what you have here.

I find people are often assumed not to have philosophical foundations when (1.) they don't primarily do the foundation-laying work, or (2.) simply when their foundations differ from the accusers foundations.

Most economics has a solid philosophical foundation and doesn't feel the need to tweak it much, going on with modeling and empirical work. There are other economists who spend their careers in the "philosophical foundation" stage and assume others simply don't have a philosophical foundation. I think it's a poor inference. Still - everyone could obviously benefit from coming back to their philosophical foundations every once in a while, and everyone could obviously benefit from testing their ideas empirically or expressing them more formally every once in a while.

You add "history" to your philosophy list... that's interesting... I would have thought that goes more in the second category. Boundaries can be fluid, of course.

Yes, the literature on the philosophy and history of statistics and probability sucks -- there is no good general survey.

What Mario said. Sen has achieved popularity as the caring persons economist but we all care, the problem is to come up with policies that deliver on a sustainable basis. Hausman and McPherson "Economic Analysis, Moral Philosophy and Public Policy" 1996/2006 are on the same track, heavily influenced by Sen but apart from introducing the problems and the literature they don't advance the discussion.

Someone made me read the new Adam Smith biography by Nicholas Phillipson, and I don't know what it added but it certainly demonstrated that he was obsessed with both the moral and institutional frameworks of human action, including property rights (and the laws of inheritance). Strange that "the hidden hand" became the signature idea for his work. He was also obsessed with Newton and he may have harboured the ambition to produce a body of moral and political economy that would be as comprehensive as Newtonian mechanics, however he wore out with the project only half done (distracted by duties as Commissioner of Customs) and he insisted that almost all his unpublished drafts had to be burned (an exception was his essay on Newton).


Even a stopped clock is right twice a day. But I'll be even more generous than that and say that even the worst thinkers sometimes stumble upon a good idea or two despite themselves. There are also those who get more out of a writer than others -- and better things than others. Rousseau's list of infamy includes providing the underlying world view of the French Revolution and all the other Romantic revolutions that followed (including all communist revolutions), ruining several generations of anthropologists with his "noble savage" nonsense, and the development of the blank slate from a Lockean arugment against monarchy to a justification for social engineering and dictatorship (aside from it being just plain wrong). Yet Goethe found something in his works to influence him, and Maria Montessori found something in his works to develop the Montessori method of early childhood education, so a generous interpretation by great minds (and better people than Rousseau) can find something there.

Marx's real-world effects have been mostly horrific. It's hard to identify any aspect of his philosophy that is correct. Yet there have been some uses of his philosophy in literary studies that has generated some interesting literary theory and literary interpretations. Of course, I'm much more inclined to Austrian economics interpretions:

Keynes was a mercantilist, and I don't understand how anyone can take him all that seriously as an economist. Yet, I will not deny his influence on some good economists. (Much to Mr. Krugman's horror, I'm sure, I will be using in an upcoming paper Dr. Krugman's work on economic geography to argue that the Austrian disequilibrium view is the correct one -- I can do this because he is, after all, oftentimes a good economist, when he's Dr. Krugman.) Still, Keynes' popularity comes mostly from the fact that he justifies government intervention. Governments love him, and those who want to influence government therefore love him. You have to tell the king what he wants to hear to get his ear.

So even the absolute worst philosophies can end up having positive influences, if they influence people better than the original philosophers. (This statement is still very much up for debate with Marx.) I think it is better, though, to start with those who have good philosophies and work from them. A metaphysics and ontology that matches the real world; an epistemology that describes real thinkers; an ethics for real people; an economics that builds on these things.

Troy - I do follow your blog, btw. I'm not especially literary so I don't have much to say but I do keep up with it.

Forgive me for not getting to hung up on a mid-eighteenth century European that refers to "noble savages". That may not be PC of me, but I'm simply not going to loose sleep over it.

It's impossible to even talk with you about it when you take this sort of view of Keynes. No matter what anybody says you can always respond "well they secretly just want to please the rulers". How does one rebut that, Troy? And how do you expect people to be inclined to talk with you about it when they know you think that of them?

As for the mercantilists - they get a bad rep, unfortunately. I recommend William Grapp's "The Liberal Elements in English Mercantilism", QJE 1952. Joseph Dorfman is also good on this point at least with respect to American mercantilism. I know this is a losing battle, though. When most people here "mercantilist" they think "protectionist", and it's usually best to just nod your head and say "yes, protectionism is bad". But in case you're interested in why Keynes wasn't bashful about embracing mercantilism, you should check those sources out.

re: "I think it is better, though, to start with those who have good philosophies and work from them. A metaphysics and ontology that matches the real world; an epistemology that describes real thinkers; an ethics for real people; an economics that builds on these things."

Well I'd certainly agree with his sentiment.

Rousseau matters because his ideas still affect us. His ideas about the "noble savage" derailed anthropology until about the mid-20th century. And people still believe it -- otherwise Steven Pinker and Matt Ridley wouldn't have to point out that the death rate for war is higher among tribal people, and decreases the more complex societies become. Rousseau argued that civilization decivilizes. Pinker and Ridley have had to argue, and are still arguing, that civilization civilizes -- and this is entirely due to Rousseau's philosophy. Further, Pinker's The Blank Slate is a direct attack on Rousseauianism, because his ideas on how the mind works still dominate. They dominate education (and not in the positive direction taken by Montessori), and they lay the justification for political correctness, on the theory that if we just change the way people speak, we will change their minds. Which gets it backwards. So if you care to understand what is going on in the world, why we have particular ideas which unfortunately dominate, then you need to understand Rousseau and his ideas and the genealogy of those ideas. Otherwise, you're just taking the world as given and not questioning why some people believe what they do. Some of the very worst ideas we have in the West are traceable directly to Rousseau. To say you don't care about Rousseau is to say you don't care about ideas or their origins, or the effects of those ideas.

I didn't say that all Keynesians secretly want to please their rulers. I said that if you want to please your (no-Marxist) rulers (who pretend to support markets), be a Keynesian. This doesn't deny the fact that there are honest people who sincerely think Keynes is right. You can even get the Montessori method out of Rousseau, even if more typically you get modern American pedagogy out of his ideas.

If you don't like "mercantilism," then crony capitalism. Practically the same thing. Mercantilism was pro-business -- more, it was corporatist. I'm pro-market.

For me, mercantilism had its time and place. It was a transition out of the Medieval economy and into the market economy. But it's much like alchemy, which preceded the sciences of physics and chemistry. I would take a modern-day alchemist about as seriously as I take Keynes. All of his magical thinking -- external shocks,"animal spirits," people for no apparent reason cutting back on spending -- amounts to no explanation of anything at all. His attempt to invert Say's law is absurd -- and demonstrates that he understood absolutely nothing about the driving force of the economy, which is the creation of new things for people to buy, and not the production of things they mostly already have. You need the latter to maintain what you have -- but it doesn't create more wealth.

If we take the topic of my blog Austrian Economics and Literature, the reason Austrian economics works so well with literature is that it considers the creators of wealth (new works of art) to be the visionary entrepreneurs. What would a Keynesian sociology of artistic production look like? It certainly wouldn't look anything like the real thing. How would one boost the aggregate demand for literature? What would that even mean? And if we did boost aggregate demand for literature, that would result in more people going into writing, resulting in a misallocation of human resources into the arts, where they find they are able to make a good living now. The result would be a general reduction in the quality of literature, as it would end up being dominated by those looking to simply make money rather than being driven to create art. The result would be a literature bubble, with an inevitable bust (people can only read so much). So a Keynesian approach to literary production would result in a proliferation of bad writers producing bad literature, resulting in a literature bubble and subsequent crash, where the reputation of the literary arts would be harmed, having been exposed as something people are doing not because they love the art, but because they thought they could make a quick buck. A Keynesian approach to literature would do nothing but harm the art. It may help a few good artists during the boom, but it's not likely that the great artists would be the ones to survive the recession. And it would take years to recover that reputation. Of course, in the meantime, Keynesian policies would try to reinflate the bubble, making the problem even worse over time.

Ya - I read your most recent post on Keynes and literature. I didn't follow a word of it. Now I have even less of a sense of what you thought Keynes thought. Do you think there is no place for the entrepreneur in Keynesian outlook or something?

When Austrian economics is applied to understanding the sociology of artistic production, it looks remarkably like the real thing. When Keynesian economics is applies to understanding the sociology of artistic production, it looks bizarre. What should that tell you?

The comments to this entry are closed.