Economists as engineers and humbler, better scientists

Here’s a paragraph I wrote about fifteen years ago.

The culture of economic expertise places inadequate weight on integrating insights from multiple perspectives, that it frequently places an unreasonably high ‘burden of proof’ on heterodox views, and that it has a penchant for spelling out the normative implications of its analysis by way of ‘peremptory rules’ which narrow the scope for dialogue.

This is just a sketch of a post, because I don’t have time for more, but for some time I’ve wanted to put down a marker about one of the ways in which results from Kaggle are instructive. The winners of Kaggle competitions often win not by trying to build the One True Model of the phenomena they’re trying to model, but rather by building a large number of models all of which have some explanatory power and which are independent of each other and then aggregating their insights. Random forests approaches often perform very well. One might call it a ‘wisdom of algorithmic crowds’ approach.

In any event, this is the antithesis of much economic modelling which involves reaching for the One True Model and then parameterising it.

I was reminded of this by David Colander’s latest piece urging humility on economists (pdf). His proposal? That instead of thinking of themselves as analogous to dentists (which Keynes suggested) economists should think of themselves as engineers:

An engineer’s approach to modeling such a complex system as the macro economy would likely focus much more on statistical models and methods of pulling patterns out of the data. It would explore a wide variety of formal models to gain analytic insight, and then would integrate the many variety of models with the statistical models to interpret the patterns. That would involve a fundamentally different way of doing macro and of thinking about macro problems. Similarly, with micro. Economists focus much of their applied micro policy discussion on Pareto optimal solutions even though we know that all actual policies will violate Pareto optimality. We can contort our micro policy models designed to provide Pareto optimal solutions to provide insight into non-Pareto optimal solutions, but, generally, that contortion comes at a cost. It means that we spend less time discussing other models that betting fit not Pareto optimal solutions, but “reasonable person solutions” that more closely reflect society’s value judgments. An engineering applied microeconomics would likely have an entire branch devoted to measuring society’s value judgments and integrating those judgments into applied policy instruments. Our scientific applied micro leaves the topic almost totally undiscussed.

Quite.

Other required reading on the point is this paper entitled “Identification Problems in the Social Sciences and Life”.

I made the latter point in a discussion of discursive collapse a while back.

 

This entry was posted in Economics and public policy, Methodology, Philosophy. Bookmark the permalink.

35 Responses to Economists as engineers and humbler, better scientists

  1. Julie Thomas says:

    “measuring society’s value judgments”.

    Why would one assume that value judgements (even by reasonable persons) are consistent across all the ‘tribes’ in a society? Surely there is too much complexity in contemporary society for this to be a useful way of thinking. There are at least two different tribes; the right and left, that value different things.

    And people change their values; it seems to me that marketing both exploits this human ability to change or adapt and contributes significantly to the change in values that has come about in our society.

    To take the idea of ‘thrift’ which has been a fundamental value of western society since the inception of capitalism. This ‘value’ has been trashed by all the messages projected onto the public by the marketing arm of capitalism.

    Banks encourage us to spend and borrow; financial advice is still for people to borrow to get ahead. It seems to me that marketing actually encourages people to ‘value’ all the 7 deadly sins, particularly, sloth, gluttony, pride, envy, greed, and lust.

  2. Paul Frijters says:

    Nick,

    you and I agree on most of this, but the one bone of contention I keep coming back to is the fact that economics is on a victory march with its current approach. Why would it change when it is so ostentatiously winning?

    Trying lots of different models has a long ancestry in macro. The early macro models by people like Tinbergen (who started out as a physicist) were all of the engineering variety, just looking for abstractions and data patterns. We still in fact do this, but we sweep the fact that we do this under the carpet for cosmetic reasons: if you dont give the outside a difficult structural headline model that they dont understand, you dont get taken seriously. Young academics and critical outsiders simply demand such models, even if the policy process doesnt really take them seriously. You really dont have a choice. The magic of the magician is in not being understood by anyone but other magicians. That dynamic is far more powerful than commonly allowed for because people do not want to admit to themselves that tricks work on them.

    If one of your kaggle people is better able than existing institutions at predicting macro-economic trends, they can make a lot of money. Far more than a couple of thousand dollars. More in the realm of millions. Even more if they manage to link it to stock markets.

  3. JC says:

    Farmers rational? I don’t think so, but there are many types of ‘farmers’ these days. It seems to me though, that ones out here are the descendants of the original ‘squatter’s’ and they retain all the features of that worldview; which is quite charming really, if one isn’t judged to be of the ‘small farmer’ or worker class.

    You lost me, Julie, especially with the non-nonsensical comment about Da class.

    Sure farmers are behaving rationally. They’re pissed at the possibility of getting thrown off the land at market value, while the government gloms the reward. If you don’t think that’s perfectly rational response, you have absolutely no idea what you’re talking about.

    There isn’t really any unity between farmers and greens, one could see that in Barnaby Joyce’s response to Bob Browns suggestion that they ‘join’ up. LOL it was like Bob had put the hard word on him, his response was so emotional but he does speak from the heart.

    Yea, I can easily sympathize with Joyce’s recoil at the thought of joining up with the Greens. Again , a perfectly normal response. And why would there be any unity with the farmers and the greens seeing the farmers priorities are rational?

    It’s a great thing IMO that the farmers are ‘forced’ to share what isn’t theirs anyway. Nobody owns the land. Their land was ‘stolen’, appropriated from the people who were here first and who used it without denying other people access.

    I presume you’re renting, Julie. If you owned a home or land you wouldn’t be peddling such obvious nonsense. If you aren’t renting I would presume your navel gazing opinion about ownership would take on a far different response if an indig showed up to the front door and demanded the keys and your immediate exit without compensation.

    It’s a shit thing that the government is far too influenced by the ‘zombie’ economics that encourages them to behave in a mendacious, grasping and utimately dysfunctional manner; that is where the government has been going wrong over the past decades.

    Yep, sure. Two decades of solid economic reforms from the Hawke and Howard governments that allowed us greater economic flexibility and now as a result we’re the richest nation on earth per cap shows just how we took the wrong fork in the road… . Lol Are you into planetary travel?

    JC you like to comment on everything, I have noticed, even if you are quite ignorant of the topic.

    Thanks for paying attention, Julie. I normally skip through your bone headed cliché riddled stuff. And I don’t comment on everything.

    You like to see yourself in print?

    Pixels aren’t print. You need to try harder with the put downs as that doesn’t work and sound kinda dumb.

    If you are referring to GFC mining, the fact is that the mining companies are offering ‘rent’ to the farmers and will be paying that rent each year that the ‘wells’ are being used.

    But not the royalties. As I said, mining firms only pay once. How it’s divvied up they don’t care and it’s grossly unfair on the farmers.

  4. Dan says:

    JC – your malware-like commenting style seems to be manifesting even more than usual today.

  5. Julie Thomas says:

    He’s – has to be a ‘he’ – commented on both threads and yet he says he thinks my stuff is bone headed and cliché riddled. What’s he want?

  6. Nicholas Gruen says:

    I think you might be on the wrong thread JC

  7. JC says:

    Thanks Nic… I actually posted it was posted in error on this thread and to delete. Dunno where the oops comment ended up.

  8. desipis says:

    I’m not sure I’d insult engineers or scientists by applying either label to economists. It seems that people miss that they’re studying people by getting distracted by the numbers, believing that they can make a hard science from the later by ignoring the former. Quality not quantity.

    If anything I’d see economics as similar to psychology. There are the base level elements we have some basic understanding of, these form subsystems which we also have some basic understanding of, and the subsystems combine together to build the whole system. We might understand the basic properties and traits of each level. However, we lack the in depth understanding required to approach the system from an engineering perspective, and any scientific investigation will be too abstracted from the whole to provide useful prescriptive solutions. For the immediate future, it’s the (ideologically driven) visions and narratives that will provide the best hope for practical solutions the worlds problems. One thing the multi-model approach allows is for the model’s constructor to build in their world view through the way they weight or combine the models.

    Economics in inherently political; any attempt to frame it as a politically neutral, technocratic problem is simply one that advocates an economic model that supports its ideological or political goals.

    If one of your kaggle people is better able than existing institutions at predicting macro-economic trends, they can make a lot of money. Far more than a couple of thousand dollars. More in the realm of millions. Even more if they manage to link it to stock markets.

    I think this is missing the point. The point isn’t to best predict the system as it currently works. It’s to work out the best way of running the system, and that’s something you can’t determine or implement simply by becoming the best cog in the machine.

  9. Richard Tsukamasa Green says:

    The engineer analogy intrigues me for other reasons, especially if I look at engineers who have done economics (like Walras, or Phillips). What I note is this blindspot that comes from an ill acquaintance with complex systems. By complex I mean systems that adapt or learn, and behavior changes because of changes in behavior and that exhibit fallacies of composition. Futhermore look at how many of the quants on Wall Street whose modesls the hedge funds and banks were using were building their models not in mistaken beliefs of a “one true model” and imitation of high physics, but along the lines of the engineer methodology suggested above. They were using approaches that seemed to be working. They failed in a large part because of these blindspots, or at least a failure to keep these issues in mind. Not surprisingly, the quants were very often engineering grads lured by wall street salaries.

    But as engineering is to physics, what is the equivalent to a complex system like biology? Horticulture? Gardening? We need a “one true analogy”!

    I am serious about the comparisons with biology though. It reminds me of this old speech by Krugman addressing Evolutionary Economists. It’s a very good piece. If I taught advanced undergraduate economics I’d almost certainly force it on my students. It examinines both what economics can learn from biology, but also the way that biology is actually studied and by extension what models are actually for.

  10. conrad says:

    “If anything I’d see economics as similar to psychology”

    The certainly are similarities, and there are also ideas and problems which are essentially identical but are treated entirely differently (notably things that occupy the overlap between social psychology and economics and some areas like gambling and addictions).

    I think one of the main differences here is probably due to history, in that historically, psychology has generally allowed crazy ideas. As far as I can tell, this means that psychology has more diverse types of theories than economics, and it is possible to propose these without any great formal analyses like economists seem to like. The upshot of this is that you really do have multiple models for everything that people are willing to consider (and indeed methodologies for trying to describe problems) and so if there is some better solution to something you won’t be caught in a phlogiston style trap, where the current theories and methods people like become too embedded to change easily. The downside of this is that you have vastly more crap to sift through and you have to argue with people that believe in this crap if you propose something, and many of them seem to have very little appreciation of formal logic, theory testing and the like (and some of this crap the public loves also).

    A good question is why psychology has remained at one end of the crap/theory-profiliferation spectrum and economics appears to be at the other — perhaps the answer here is just that the type of people attracted to one or the other is different, so this sort of mentality is self-reinforcing. It seems to me that economics is very blokey and you get the Nobel prize for showing you can do maths better than someone else. Psychology, alternatively, attracts all sorts, many of whom will propose theories without understanding the first thing to do with even simple mathematics (these people are the majority in some areas like developmental psychology), which means they won’t even care about your very well specified formal theory. In addition, if these people are the majority, your very well specified formal theory will lose out to the one that isn’t but everyone seems to be able to appreciate (cf., understand), because no one will understand your theory unless you present it very carefully.

  11. Dan says:

    Conrad – re. crazy theories, three words: Real Business Cycle.

  12. Tel says:

    http://en.wikipedia.org/wiki/Talk:Dimensional_analysis

    WikiProject Measurement — High-importance
    WikiProject Physics — High-importance
    WikiProject Economics — Low-importance

    That tells you everything you need to know about Physics envy and the newer wave of Engineering envy. Economists are happy to wear the trappings, but uninterested in doing the groundwork.

  13. Mike Pepperday says:

    So, Conrad, you reckon that there’s less crap in economics because it’s more blokey? Okay, okay, slip o’ the pen, just kidding.

    Dan I don’t know anything about the craziness of Real Business Cycle theory but it obviously has the power to move you. Psych hasn’t even got a theory you could call crazy. Just loads of crap as Conrad says.

    We have had this conversation before. If you want to do genuine science theory you must deal in idealisations – “ideal-types.”

    Economics does this: homo economicus does not exist. Nor does perfect competition, or market clearing, or anything else in economic theory (I gather). It is with idealised theoretical abstractions that you get to develop science theory.

    Psychology never does it. I’ll give you an example: self-identity theory. As you may know, there are two theories: you get your identity from what you do (I’m a carpenter), or else from what you are (I’m a mother). The first sees himself as different from others and complementary to them and what he IS is irrelevant. The second sees herself as the same as others and in harmony with them and what she DOES is immaterial.

    They are pretty clear concepts and apparently the literature goes back to God knows when but do you ever see anyone take the concepts to extremes as economics does? Instead it’s all waffle and there are two camps basically not on speaking terms. If you asked those who do see the two theories as equally valid whether people might get their self-identity both ways, they will think the question fatuous and you will be snowed with another blast of waffle.

    But if they took those two theories as extremes, as pure types, they might get some focus. It is standard procedure in economics and the natural sciences. It is how to do genuine scientific theorising. It is NORMAL. It is just that in the social sciences (outside of economics) nobody does it. Why not? I don’t know. Given the success of economics and natural science it ought to be the first thing to do.

    They’ll never be scientific (and blokey) till they do.

  14. conrad says:

    Mike, many people in psychology do try and find what you call idealized forms. A really good and relatively short discussion that argues in favor of what you talk about from one of the exceptionally smart grandaddies of this approach in cognitive sciencec is here: http://waldron.stanford.edu/~jlm/papers/McClelland09PlaceOfModelingtopiCS.pdf

    I also think the search for idealized forms and solutions to problems are basically orthogonal incidentally (in economics also) and the reason is obvious — many problems are so complex it’s likely that no-one will ever find idealized forms, but that doesn’t mean trying to undestand them is useless (far from it). This is why many of the great mathematical models of cognitive science (which presumably fit your definition of idealized forms) tend to deal with lower level mental processing where you really can abstract away quite simple mathematical forms that lead to complex behavior (e.g., perception, memory) versus things which are undoubtedly vastly more complex and indeed difficulty to define like personal identity (c.f. e.g., speech perception).

  15. KB Keynes says:

    There is too much ego in Psychology I am afreud to say

  16. Dan says:

    [email protected]: as an hons graduate in psych I am the first to admit it doesn’t usually lend itself particularly well to empiricism at a fundamental, building-blocks level. However, there are many important social psychological phenomena that are, I guess, descriptive in orientation that are amenable to study in a lab setting – see, for instance, Kip Williams’ work on ostracism, or social loafing, the Dunning-Kruger effect (which is in any event of great amusement value), etc.

    Even at the building blocks level, there are exceptions. For instance, I did my hons thesis on Terror Management Theory which makes very basic, indeed reductionistic claims about the fundamental reason that human cultures and values exist, yet lends itself to experimental work very well, thank you. It’s almost too good to be true.

    Ultimately I think like (I think) Rutherford said, the only real science is physics, the rest is just stamp collecting. But sometimes said stamp collections can be useful/valuable.

  17. Mike Pepperday says:

    Conrad, I looked at that paper. I don’t think it is anything to do with it. It is about “models” which are “are explorations of ideas about the nature of cognitive processes.” A scientific theory is not an exploration. The abstract says “simplification” is essential. That’s not idealisation. When Newton connected tides and pendulums and a falling apple he didn’t do it by simplifying.

    You say many psychologists try to find idealised forms. “try to find” – that’s a funny way to put it. Did Newton try to find mass and distance? Anyway I don’t see any idealisations in psychology. None. In economics I can readily name some though I know nothing of economics. In psych, which I do know a bit, I can name none. If that paper idealises something, tell me what it is for I can’t see anything. This is a bit like your claiming the other day that psychology has axioms but then you couldn’t quote any. Instead of directing me to a very abstruse paper, what about giving me some examples of psych concepts which people have idealised?

    I’m a bit mystified by the sentence “…the search … are basically orthogonal…” but I will say that whatever the “problems” are, their solution will be an idealised relationship between ideal types.

    “many problems are so complex it’s likely that no-one will ever find idealized forms, but that doesn’t mean trying to understand them is useless”

    In my view, if they don’t find them they won’t build genuine theory and “trying to understand” will (continue to) be useless.

    “vastly more complex and indeed difficult to define like personal identity.”

    Well, they ARE talking about identity. So why not idealise it? Where do they get the idea they can differ from standard science? Economics doesn’t so why does psych (and other social sciences)?

    As for difficult to define – science cannot depend on defining. Definitions are good for teaching school children established knowledge, not more. A definition is words and each word would also need defining so it’s an infinite regress. A science based on definitions is therefore impossible. Newton’s “mass” has no agreed definition. Science theory depends on ideal-types which cannot be defined.

  18. Mike Pepperday says:

    Dan – you are misunderstanding something. Empiricism is the exact opposite of theory building. Theory has nothing to do with a lab setting. Read my post again maybe. The building blocks of a scientific theory are things which do not exist. That’s the way of science theory. Consider: the shortest distance between two points is a straight line. There are no such things as points or straight or a line. They don’t exist yet from this sentence and one or two more, Euclidean geometry – which is empirically indispensible – can be deduced.

    I had never heard of TMT (thought it was going to be about airport security) so thanks for that. I read Wikipedia which is basically a puff piece. I was amused to see potty training discussed. I’d be of the opposite opinion from TMT: I’d have said we’ve banished death from our society and so don’t fear it at all. That is, you retire from society before dying. If you chat to the “grey nomads” you will find they are indeed conscious that death is approaching. They just want to enjoy life before the end. They are frank about it. Seek meaning for their lives? Forget it.

    But if you have a theory and you go looking for support you will find it. TMT has never been tested – all the mentioned papers are out to prove it. They prove nothing to non-supporters and a paper that fails to support will make no impression on the supporters (probably won’t be published). That’s not science. Genuine science does a test which actually tests – i.e., which might show the theory to be wrong. There is no potential way to prove TMT wrong. On the other hand TMT will not be a basis to build knowledge on. It will go on supporting academic careers for a generation or two and then fade as its supporters run out of things to say. It was ever thus in social science.

    I don’t know enough about it but I also have a non-scientific objection: to me TMT is missing quotidian resonance. Compare with those identity theories which simply make sense. Or compare with Seligman’s Helplessness theory which you can actually experience for yourself and which has heaps of anecdotes. But subconscious terror of death? Nah.

  19. Dan says:

    Hey Mike,

    I think you’ve misunderstood a bit – no TMT theorist would claim that we haven’t attempted to banish death, they’d just say that because we are ultimately cognisant of it, the effectiveness of such attempts is limited.

    As for theory vs empiricism – well, Becker and Rank developed the theory (based on observation; of course, a theory is not much good if it has no bearing to observed reality, and I think Newton would be the first to agree here) which has subsequently been operationalised in an experimental setting by social psychologists.

    What would ‘non-believers’ in TMT claim supports the result, replicated over and over, that people become more defensive of their worldview when their mortality is made salient?

  20. Tel says:

    I did a bit of reading on Terror Management Theory, seems like just a restatement of Darwin, only with a lot less emphasis on reproduction.

    Consider: the shortest distance between two points is a straight line. There are no such things as points or straight or a line.

    The entity you are considering is not a scientific theory.

  21. desipis says:

    If they’re claiming there’s no such thing as straight, it’s clearly a part of the homosexual agenda, not science.

  22. Tel says:

    My point (feels odd that I should have to explain it) was that geometry is pure mathematics, whereas science and engineering are applied mathematics. You can spend a long time doing geometry, but you won’t discover the speed of light or charge on an electron by that method.

    Since you bring up the topic of sailors… when at sea, if you want to travel from one point on a map to another point, drawing a straight line between the points is not the shortest distance. Navigation is also applied mathematics by the way.

  23. Dan says:

    Tel: I don’t think that needs elucidation. What’s its relevance though?

  24. Mike Pepperday says:

    Dan, you would like to persuade me that TMT is useful. It’s in your heart. Where’s the other side? You would leave that to me and no matter what I said, I could never persuade you TMT is wrong. That’s not so much a personal criticism as saying it isn’t science. A genuine scientific theory states its falsification criteria – the tests which would show it to be wrong if it failed to pass them. That makes it scientific.

    “…developed the theory (based on observation; of course…”

    Nooooo!! No scientific theory is based on observation. It is based on things that do not exist and cannot be observed. I feel like I am talking to a wall. Honestly, what is the problem here? Why do social scientists insist on something at odds with natural science theory? Newton’s first law says a body will go on in a straight line at constant velocity forever. Observe it? There is not a single example in the whole universe. Not one. Galileo was saying what I am saying 500 years ago. I wonder what the hang up is.

    You did your honours thesis and it was accepted. Maybe you can build a career on TMT. I expect many people will. They’ll go to conferences and refer to each other in their papers and build careers. But they won’t contribute one jot more to understanding human psychology than countless thousands of other social scientists of 20C, some of them famous in their day. All those brilliant minds! Speak to an older academic and you will hear names and theories that were the hot thing when he or she was young and are now forgotten. TMT is going to fade into those mists. Why not take a breather and read some Popper?

    “Operationalised” is social science jargon. It gives significance (or majesty) to a concept unknown in those branches of knowledge that actually work and upon which further knowledge can be built.

  25. Mike Pepperday says:

    Tel – whether you call geometry “science” or not is immaterial. It is theoretical knowledge, knowledge without which you would not, to use your example, be able to measure the speed of light. Still, let me say: the components of what you would call science theory are things which do not exist. And truly: Galileo was saying this 500 years ago and with that began the construction of the world we know. It worked for the natural sciences, it works for economics so why not take the same approach in psychology etc?

    I did not mention sailors but I did specifically mention the geometry was Euclidean. You are presumably thinking of a Mercator map. I used to teach map projections.

  26. Dan says:

    Mike: please take my word for it when I say I couldn’t give a fig whether you are convinced by TMT or not. Quite to the contrary, I am interested in alternative explanations for the experimental findings that have occurred under the auspices of TMT.

    I did my hons thesis some years ago and my present interests and career aspirations are entirely different (in fact I’m doing a masters in political economy at USyd right now).

    As for theory and observation: I hear your despairing cry but still cannot concede that theory comes out of a vacuum, or that if it did, it would be useful.

  27. Mike Pepperday says:

    “…cannot concede that theory comes out of a vacuum, or that if it did, it would be useful.”

    So how do you cope with Newton’s first law? It certainly wasn’t observed. You wouldn’t deny it is useful so what do you do? Do you relegate this to a corner of your mind labelled Contrary To Preconceptions And Not To Be Thought About?

    As a budding political economist, how do you cope with homo economicus, market clearing, etc, etc, etc, none of which exist to be observed and which are extremely useful?

  28. Dan says:

    They are abstractions of reality. Newton’s genius was to spot the mechanics of what was actually going on when, say, an apple fell out of a tree.

    Abstractions are of course useful when used appropriately, but as you say, not real, and as I say, not from a vacuum.

    I think we are talking past each others here.

    Let me take a different tack: most science these days is probabilistic rather than one-to-one.

    That doesn’t mean it isn’t related to abstract theory, nor that it doesn’t tell us something about the world.

    Isn’t science about being able to make predictions? As such, it seems your definition of science is really narrow. For instance, as I understand it, medicine wouldn’t meet your criteria for what constitutes science.

  29. Mike Pepperday says:

    Yes, Newton’s genius. That’s his brain, not a vacuum at all.

    Spotting what in reality? Millions of people have seen apples fall.

    It was despite reality. Philosophers from Comte to Popper – everyone who has thought about it – says there is no observation without a theory. No one asserts the contrary. Observing without a prior theory is impossible. So you cannot “abstract” from reality; theory comes first. Only after the theory can you see how the apple falls.

    Your TMT is a case – invent the theory then go observing. The problem is that social scientists don’t observe in order to test the theory; they go looking to “prove” it. Every philosopher (even ones who haven’t thought about it) knows you cannot prove a theory. Can’t be done. Yet whenever social science has a theory it tries to do exactly that.

    Just like religion and astrology – you have theory and glory halleluiah you see it confirmed.

    “most science these days is probabilistic rather than one-to-one.”

    Nope. That’s just social “science” and it fails. Your “these days” speaks volumes.

    “That doesn’t mean… …that it doesn’t tell us something about the world.”

    Quite right. Administrators since long before Herod have wanted the stats. And, it certainly enables useful predictions; that’s why they do it. But it isn’t science; it’s bureaucracy. As someone said, correlations aren’t causes.

    Certainly medicine meets the criteria. The virus enters like this and the enzyme chops it up like that and it reassembles itself like so… Straightforward science. But lots of drug testing is bureaucracy. Note that you could never find out how a drug actually works by conducting public trials and running computer regressions on them. No medical researcher would be so foolish as to imagine they could.

    Yet that is just what social science does. For example, over the last couple of decades the game theorists (blokey economists) have been doing it in large numbers. What’s come out of it apart from successful careers? Nothing scientific and at most a few factoids. One factoid is that women test to be a whole standard deviation less rational than men. It’s quite predictable! This is what passes for science in social “science.”

  30. Dan says:

    Gnah, sorry, you’re just being contrary now. A lot of physics is tested probablistically nowadays – there’s so many variables at play. And there is heaps of medicine where treatments are seen to work but we don’t (yet) know why.

  31. Jacques Chester says:

    Ah, discipline envy. Beware the Lord High Executioner:

    Then the idiot who praises, with enthusiastic tone,
    All centuries but this, and every country but his ow.n

  32. Jacques Chester says:

    And beware also stray punctuation, it would seem.

  33. Dan says:

    T… t… t… timing – the secret of good comedy.

  34. Chris Lloyd says:

    “In any event, this is the antithesis of much economic modelling which involves reaching for the One True Model and then parameterising it.”

    I think this is probably an out of date comment. There are any number of wisdom type methods in the econometrics and statistics literature. Try googling “model averaging” and “ensemble forecasting”. Most of the economitricians under the age of 40 that I know are on top of this kind of stuff.

  35. Nicholas Gruen says:

    If you look at the quote in this tweet, consider how a battleship is designed – as a process of bringing all the given knowledge to bear on the problem in a structured way with all matters that can be, being brought to the table and considered in turn and related to each other.

    Economics is nothing like this. It involves grabbing hold of some model and then running it. As for all the complicating factors, well, let’s not worry about them.

Comments are closed.