The Dark Side

Off and on, we've been discussing the progress economics has made as a discipline. In those discussions, we often hear what a sorry state economics is in, that the theoretical models don't predict well, that economics isn't a real science, and so on. A comparison to physics often comes up.

One response to this onslaught against economics is to point out that physicists also have a theory that doesn't predict well, in particular the motion of galaxies and the matter within them is at odds with theoretical predictions. Visible matter cannot account for the gravitational "stickiness" of clusters of matter.

The solution physicists have adopted is not to question the theory, at least not for the most part - see below, but instead to make stuff up - dark matter, dark energy, I expect there will be a "dark light" soon as well - whatever is needed to make the equations work must exist.

When I respond to attacks on economics by pointing out that these theories seem every bit as controversial and uncertain as ours, in fact perhaps more so with this "making stuff up" step that assumes the theory is correct and then assumes the matter needed to make the theory work is present in just the right invisible amounts to get motions to agree with theory, I get told it's because I don't understand (and let's not get into the vacuousness of string theory - when I say the absence of testable predictions makes string theory philosophy, not science, I guess I don't understand either).

I'm going to get myself in trouble again by not understanding one more time. This is an astrophysicist from a blog in the NY Times called Across the Universe:

Searching for the Dark Side, by Chris Lintott, Commentary, NY Times: ...The ... kind of matter we – and everything we can see directly – are made of accounts for only a sixth of the total mass in the universe. In the absence of knowledge, we label the rest as “dark matter” and look to detect it via its effects on what we can see.

We have long known that there is more to our universe than we can see. The great eccentric of American astronomy, Fritz Zwicky, realised in 1930 that the galaxies within galaxy clusters were moving more rapidly than they should be. The only explanation was to assume that there was more matter present than scientists had thought... A similar problem exists on the scale of individual galaxies; if only the visible disk of a galaxy such as the Milky Way existed, the galaxy would fly apart in just a few million years. The solution is the same; if the Milky Way is embedded in a halo of dark matter, then all is well.

What is this dark matter? We know very little about it... [M]ost cosmologists now believe that dark matter is composed of slow-moving exotic particles that have yet to be identified.

Such a situation is unsatisfactory to say the least. Although particle physicists are working on the problem, ... it is embarrassing not to know what the main component of the universe is. The alternative is to assume that there is something wrong with our knowledge of gravity, and that the work of Einstein needs revising. Theories that attempt to do just that have been somewhat successful. ...

[It will be] a stringent test for theorists attempting to dispose of dark matter, but I for one hope they succeed. It would be tidier, somehow, to lose the enigmatic dark matter, and exciting to discover a successor to Einstein’s relativity. As George Bernard Shaw said in 1930, “Ptolomy invented a universe and it lasted 2000 years. Newton invented a universe and it lasted 200 years. Now Dr. Einstein has invented a new universe, and no one knows how long this one is going to last.”

"[I]t is embarrassing not to know what the main component of the universe is." That's why economists should hang their heads in shame when real scientists enter the room.

Comments

[jm]

What's always puzzled me about the "dark matter" hypothesis is, if dark matter is so gravitationally active as to be bound to galaxies, and there is so much more it than regular matter, why has it not collapsed along with the regular matter into the planets, making the measurable mass of the earth many times more than can be accounted for based on the regular matter present?

[b]

String theory is just games with math. Astrophysics comes perilously close to being just games with math.

If astrophysics were the only kind of physics, physicists would be held in the same disrepute as economists.

Economists of the world, please find some testable quantitative predictions, or at least historically testable quantitative predictions. And then educate us about them.

I want to take economics seriously. I want to think that economics is more than anthropology plus games with math. But no one's giving me the testable quantitative predictions that would allow me to think that way.

Maybe economics really is just like string theory: a bunch of mathematicians fooling themselves that they're scientists.

[piglet]

Mark, you don't understand. Sorry. You don't understand. You're getting yourself in trouble with that kind of comparison, that's what you are doing. For one thing, physicists at least agree on how to test their theories, they do test them, and when one of their theoretical predicitons is contradicted by observation, what happens is that physicists will announce to the world that they detected a contradiction. Which will then be discussed in the physical journals and reported in the general media. That is not what happens among economists. Or is it?

[math student]

Sorry, somehow I think physicists have less to apologize for: they invented mechanics, electromagnetic theory, optics and quantum theory (to name a few theoretical areas) all of which yield correct predictions of observable phenomena. Economists, by contrast, are unable to predict even the most basic economic events. (One glaring example is of course, a recession.) The hallmark of science is an ability to make correct predictions. It seems to me that part of the problem for economists is that they are studying something that may change over time, not some physical process that does not change in character. Also, it seems that economics has a measurement problem -- many relevant variables are not directly observable (as is, say, the velocity of a projectile), and the data that exist seem to be subject to considerable measurement error. Another problem with economic theory is that it often seems to involve simplifying assumptions that may be "too simple" to accurately characterize the object of analysis. It seems to me that economists are studying something that is inherently more difficult to study than physical systems that physicists work on, but they could benefit from more rigorous application of the empirical framework that physicists use: theorize, predict and test predictions; then reject theories that ultimately don't work out.

[Bruce Wilder]

I have little comprehension of what astrophysicists are talking about. I just barely understood what the special theory of relativity was about, and I honestly don't know what even that has to do with making things that go boom.

I kind of think that if one were to try to explain a philosophy of science and epistemology -- to explain what a theory is and how it relates to knowledge -- it would be best to take a simple theory about something familiar. Queuing theory would have been my candidate for illustration -- the four Little theorems are not difficult to understand, and people have personal experience with standing in line. I'd like to think that some people might even be amused to discover that there is a theory of standing in line, or that it can be studied, both analytically and empirically.

The thing is, the people, who study queues are pretty clear about how they judge if a model works well enough or not. They respect the necessary and sufficient conditions for the systems they are trying to understand. As far as I know, no queuing theorist takes money from right-wing foundations to bloviate about how the natural rate of standing in line, or makes speeches in front of the Chamber of Commerce declaring their passion for fighting the unnecessary shortening of lines.

Ultimately, all science is about method, critical method -- systematically checking to see if logic is valid and statements of facts are accurate, and building up shared meaning on that basis.

I don't know what astrophysicists are talking about when they refer to dark matter, or strings. But, I am pretty impressed that they have a tight enough logical structure, that they can figure out that there is something that they are not observing, that doesn't add up.

There are things in economics that don't appear to add up, and, often, only the non-economists appear to be clearly aware of that. I know that's not really the state of things; that, in fact, economic specialists are well aware of the limitations within their specialties. But, there's a lot of doctrinaire pronouncement, arrogance and esotericism in the public discourse. Economists might take a lesson from the physicists, and recognize that the heart of science is identifying what you don't know, what remains a mystery. That's a real scientific achievement.

[eightnine2718281828mu5]

---
String theory is just games with math
---

Mainly because this time around the math required by the theories didn't exist so, much like any good craftsman, they had to create their own tools.

Newton (and Leibniz) did the same by creating the mathematics (calculus) required for the expression of their physical theories.

[gaddeswarup]

Here is a long article by Partha dasgupta on the topic:
http://www.econ.cam.ac.uk/faculty/dasgupta/modecon.pdf
It originally started as review of some books of Heilbronner and the author says that he cannot do justice to the topic by making it shorter.

[calmo]

Interesting comments about the philosophy of science and how to keep both economics and physics in that box, science. Possibly string theory, but it doesn't look like History or Political Science or Psychology or...are even mentionable. We (recreational [speak for yourself]) economists have our limits...have Pride...and don't want to be associated with the likes of that ilk.
Seriously (just kidding) there are some good economists (Krugman) and some bad ones (Kudlow) just as there are some good physicists and bad ones. There is progress not only in physics, people but in economics --no matter the impression one gets from listening to Kudlow for 10 minutes (the lethal dose).
I refuse to be bullied by those physicists just because the Kudlows in their ranks are not so obvious.

[Mark Thoma]

Economists do put their theories to the data, test them, and update them when they don't work. etc. I don't acknowledge that physicists have a leg up on economists on this front. Nor will I acknowledge that we have made no progress. We have.

[I am not taking this position, but the fact that recessions are unpredictable can actually be viewed as a success. If policymakers use every available signal to smooth the economy, there will be no way to predict its future, only randomness will remain. There might be lots of noise remaining, and physicists and others looking at the data may see patterns where none exist that look predictable, but that doesn't imply that the models are wrong all by itself. It's also kind of funny when people use the fact we can't predict stock markets as some sort of evidence against us even though that's a prediction of our models -- it's actually a success in terms of confirming theoretical predictions.]

Out job is much harder than that of "scientists" - the things we model have brains and react to external stimuli in rational manners, and we are limited to (mostly) historical data where more than bivariate relationships must be taken account of - but that doesn't make it less rigorous. And even so, to take just one example, there is evidence that policy has contributed substantially to the Great Moderation - how'd that happen if the models don't work, if they don't predict, etc.? What is the problem that you have with such evidence? Do any of you really know what the evidence is, what it is based upon, the sophistication of the information extraction procedures used on historical data and the care that is taken with both the theoretical and statistical underpinnings? Take a look sometime - you'll be surprised. It has weaknesses, sure, and there are places a professional colleague could poke at and make you question the findings - that's what we do, try to poke holes in old stuff and build better stuff to replace it - but there's also a solid core to build upon. [I didn't choose this for any particular reason, but here's a February 2007 publication by Stock and Watson (prepub version) that is part of the current fuss over inflation dynamics - what is wrong with this in terms of scientific method?]

And do take a look at microeconomic models sometime. Everyone forgets about those, how well they work, how well they predict, etc. Instead, the focus is mainly on macro variables. But macro is only part of the profession, and perhaps not the most successful part when it comes to building economic models.

[Lafayette]

MT: "A comparison to physics often comes up."

Oh, why?

Comparing hard and soft sciences is infantile. Economics employs mathematics and so does quantum physics. So the sciences are comparable?

Short answer: No.

[Mark Thoma]

I don't think the charges against economics are fair so I plan to keep bringing this up as a way of trying to counterbalance that a little bit, though I doubt there's much I can do.

Yes, though I take a strong counter position, we have problems, economics is not perfect, etc., but it's not at all like how I hear it portrayed.

For the Bruce Wilders who have been confused by the hacks, we are doing our best to tell you who they are and dull their impact. I reran DeLong's post today for exactly that reason (as though the hacks in economics are somehow different the the ones in global warming, or the research on smoking, or, well, you get the point, the mere existence of hacks does not say much about whether it's a valid discipline within the academic community - if queues were going to raid someone's deep pockets, queuing theorists would show up saying all sorts of nutty things).

Bruce says "Ultimately, all science is about method, critical method -- systematically checking to see if logic is valid and statements of facts are accurate, and building up shared meaning on that basis." That's exactly what we strive to do.

[Lafayette]

MT: "the things we model have brains and react to external stimuli in rational manners"

Yes, well there is probably the first error of the science of economics.

It is comforting to presume that consumers behave rationally. That they are fully informed of the offer available and they chose that which suits them perfectly based upon the information available. All that reassures economists that their forecasts do not suffer from irrationality or unpredictability or, simply, chance.

Unfortunately, the above scenario, though comforting, does not reflect reality. Consumers make decisions based upon a great many factors that can be irrational. Impulse is just one of them. (Such as selling stock positions when an aging ex-Governor of the Fed speaks of impending recession. Or purchasing a suburban house that corresponds to a utopian "dream" that a couple may have of family living, but is clearly beyond their means.)

All irrational, but very real, especially the former that destroys equity value virtually overnight.

My point: People are irrational and the universe contains unforeseeable black holes, but we should not fault the technicians studying either because of their apparent failures.

If God had meant humans to be perfectly prescient in all they undertake, both economists and physicists would be showcased in museums. And, we know well that they aren't.

[Mark Thoma]

Our models don't require individual rationality, nor do the assume full information - far from it. And there's a whole literature on things like noise traders (tip of the iceberg), and the information literature is vast. Our models capture much of what you talk about, and a whole lot more.

[Lafayette]

Bruce: "systematically checking to see if logic is valid and statements of facts are accurate, and building up shared meaning on that basis."

And the "Flat Earth Society" may beg to differ.

A flat earth was a notion that reigned for decades upon decades as a "scientific truth". Until someone took a boat west and actually came back. Science predicted Columbus would fall of the edge of the earth.

It is easy to say "systematically checking" the logic against the facts. But, it is damn difficult to do. If not, oftentimes, near impossible.

One needs the means, first of all. Neither physicists nor economists have all the means necessary and "observed data" can be widely wrong. Secondly, the soft sciences are about human beings and their choices, which is a highly random variable ... unlike the planets.

And, because of this, economists often are predicting the past and not the future.

[calmo]

Mark writes Do any of you really know what the evidence is, what it is based upon, the sophistication of the information extraction procedures used on historical data and the care that is taken with both the theoretical and statistical underpinnings? Take a look sometime...shucks...no, I don't know.
and I feel horribly recreational...a mere gazer and not a serious looker, a dabbler with no appreciation for punishingly thorough foundations that make the place respectable...and not some VooDoo vortex for the astrologically warped.
Ok, I can mend my ways and become evidencial, but before I sit up and fly straight I would like to note that applied science (engineering and finance) appear to be our measuring tools for what is successful (Progress) in physics and economics.

[Bruce Webb]

"The solution physicists have adopted is not to question the theory, at least not for the most part - see below, but instead to make stuff up - dark matter, dark energy,"

Totally unlike economists who have taken to explaining 'yield curve inversion' by indeed "dark matter".

Physicists have to tinker at the margins. There really were things about Mercury's orbit that were inexplicable before Einstein. But at least they know and knew what they were measuring, Whereas we have 'productivity' and 'marginal efficiency'. As if.

Economics is fun, I enjoy the give and take, but it is way to infused with ideology to even be approaching the concept of "science".

[Bruce Webb]

"Economists do put their theories to the data, test them, and update them when they don't work. etc."

Mark have you ever met a Marxist? Or someone from the Chicago School? We have seen catastrophic failures from ecoomic applications from both the Left and Right, in some cases resulting in millions of deaths, and all I see is people tinkering with the model and adjusting reality to fit theory. I don't see a lot of people repudiating Uncle Miltie or Comrade Marx just because their theories didn't exactly work when the Praxis hit the Road.

[a]

Since economists have their own dark matter, to explain why the dollar hasn't collapsed, I'm not sure it's a great idea for an economist to bring this matter up...

Anyway, physics may not explain everything, but it has had incredible successes. It can predict where Mars will be next year. It can produce theory that leads to a hydrogen bomb.

And economics...? Well, the one thing economists like to tell us is that free trade is a great idea. They just leave out all the excepts...

[evagrius]

I feel a bit like Dostoyevski's Underground Man.

If economists decide that I make rational economic decisions, I will refute them by making irrational ones.

If economics is about human beings and their actions, then the supposition that rationality, ( a certain type of rationality), is a false one.

Better to ground economics in the study of necessity and how to transcend it. Necessity being food, clothing, shelter.

After that,ground economics in the study of desire beyond necessity and the fulfillment of that desire.

Then look at which desires are wholesome leading to some "happiness" and which ones are not.

Capitalism is based on transcending necessity and then the fulfillment of desire, some good, some not so. It's evident that consumerism is desire, some wholesome, some not, but most created by advertising. That desire is artificial and it's leading us to some undesirable consequences.

Adam Smith wrote on Moral Sentiments before writing on economics. That is, he looked at desires and their nature.

How come there hasn't been a modern or post-modern equivalent?

Is it because desire is fundamentally nonrational?

[anne]

"The solution physicists have adopted is not to question the theory, at least not for the most part...but instead to make stuff up - dark matter, dark energy, I expect there will be a 'dark light' soon as well - whatever is needed to make the equations work must exist."

Good grief.

[piglet]

Lafayette: "A flat earth was a notion that reigned for decades upon decades as a "scientific truth"."

Bad example. People thought for many millenia that the earth must be flat, but they didn't have even a concept of "scientific truth" at the time. Ever since the Greeks figured out the circumference of the earth in the third century BC, the shape of the earth was known to the scientists of the time. http://www.bede.org.uk/flatearth.htm

[piglet]

Do any of you really know what the evidence is, what it is based upon, the sophistication of the information extraction procedures used on historical data and the care that is taken with both the theoretical and statistical underpinnings? Take a look sometime...

Mark, you give us such a nice sample of economic thinking on this blog. Remember those debates about the "myth of rising inequality", the "productivity miracle", and how Medicare tripled health care spending? Disagreements about housing bubble yes or no, does the trade deficit matter etc.? Do you expect us to trust this kind of science to land us on the moon and bring us safely back to earth? Maybe it's unfair that macroeconomics gets all the attention, but look at your own blog.

[Bruce Wilder]

L: "A flat earth was a notion that reigned for decades upon decades as a "scientific truth". Until someone took a boat west and actually came back. Science predicted Columbus would fall of the edge of the earth."

As an historical matter, this is simply not true. Columbus knew the earth was spherical, because he knew the science of his day. Science had known the earth was spherical for hundreds of years. Eratosthenes calculated the circumference of the earth in 240 BCE, and his achievement was famous in the ancient world. Columbus's inspiration was not a novel insight, but a foolish miscalculation -- his estimate of the circumference of the earth was wildly shy of Eratosthenes' more accurate estimate.

And, it is hardly necessary to take a boat around the world, a la Magellan, to "prove" the spherical nature of the earth. Any mariner with a spyglass, looking to the horizon from even a modest elevation, can see the curve of the earth; just watch as the mast of a distant ship appears, and "grows". You can see the curved shadow of the earth cast across the surface of the moon, if you will only look up.

I agree, though, that really seeing the obvious is often hard work, and difficult to do. Drawing logical inference and making careful measurement requires discipline, and sometimes, great technical skill.

There's no doubt, that the "model" of a flat earth is the one, which most people used in their daily lives in Columbus' day, and use still. An architect doesn't need to take into account the curvature of the earth, in the roofline of a 2000 square foot house.

I think many, if not most, of the popular critiques of economics, are downright silly. And, I think economics is making progress.

But, I would still fault economics for falling far short as a "science" and the fault I would find, would be akin to not noticing the everyday evidence for the curvature of the earth. It seems to me that economists too often teach the equivalent of a flat earth model, while consigning all the contradictory evidence to more esoteric categories of research and writing.

Price theory, for example, leaves people to puzzle over the fixed prices of the supermarket. Classical equilibrium models hold an unholy fascination for macroeconomics. Professionals have all sorts of excuses for models that they know are too simple to account for the phenomena they are studying.

Trade experts know, say, about the implications of "external economies" and "increasing returns", but discard such esoterica to return to Ricardian comparative advantage to argue for "free trade" and against "protectionism". Non-economists are exasperated -- the categories of Ricardian comparative advantage don't allow for acknowlegment of what they commonly observe and are concerned about.

Too often economist's flat earth models are used as political clubs -- to use the pretige of scientific expertise to argue, for example that "entitlement spending" threatens the fiscal foundations of the national government, so social security has to be "reformed", or that abolishing the minimum wage would help the poor, or to further some other reactionary agenda.

[Carolyn Kay]

Are economists feeling like the much maligned cavemen in the Geico commercials? Of course economists are scientists, at least the ones are who work toward understanding the world as it really is, rather than as they would prefer it to be.

But I have to disagree that the theory of dark matter is just "made up". Every attempt at an explanation for observations that are not understood has to start somewhere, with an idea, a hypothesis, which the scientist then works to prove or disprove.

I'm not a physicist either, but am interested in the subject. Dark matter sounds to me like a very plausible hypothesis to be proved or disproved to explain the observations mentioned in the post. If string theory is correct, and especially if the strings have dimensions beyond the four we can observe, why would the only existing matter unfold only in those dimensions?

It's much more reasonable to assume that strings exist everywhere with various unfoldings, and those that we can't observe yet because they're beyond what we can easily understand are what scientists are calling dark matter. Who knows what characteristics strings of various unfoldings might have, and what kinds of interactions they might have with the strings forming observable matter? The answers to those questions may someday explain the behavior of galaxies and give us a better understanding of gravity--and even the clumpy nature of the universe.

Similarly, economists have to make assumptions to explain phenomena, then test the assumptions. Those who aren't ideological shills will adjust the assumptions according to where the data leads.

Carolyn Kay
MakeThemAccountable.com

[James Killus]

One of the problems that scientists have generally, even among themselves, is confusing frontier with established territory.

Physics isn't in trouble because it has difficulty with certain astronomical observations; that's a frontier. Dark matter, dark energy, etc. have little or nothing to do with, for example, the orbit of comets, (done quite well by Newton), or the photoelectric effect (Einstein on that one).

Similarly, whatever difficulty macroeconomics might have with predicting recessions (exactly how well do geophyicists do in predicting earthquakes, hmm?), economists do wuite well in predicting price responses to a supply crunch, sieving data to show that corporate boards are gimicking their stock options, and showing that tax cuts do not "pay for themselves."

The reason why economists come under attack in more areas is because more people have vested interests in money than they do in astronomical observations.

Physics has the great good fortune to (mostly) confine itself to easily observable and quantifiable phenomena that can be described (mostly) in a few shortish equations. So they get to sneer at, for example, chemistry and say "Chemistry is merely very compicated physics." Try removing the weasel word "merely" from that sentence.

But let a physicist get involved in something deemed important, like, oh, I don't know, the heat and mass transport of the atmosphere as modified by changes in trace gas composition, and suddenly, Lo! it becomes just another example of "junk science" if it fails to kowtow to someone's perceived self-interest.

So, as I've said before, the chemical, atmospheric, and biological sciences say, "Welcome to the club, Mark!"

The physicists (with a _lot_ of help from the chemical engineers) made a really big bomb over fifty years ago and they're still riding the gravy train that came afterwards. But their idea of a "theory of everything" is actually a theory of about thirty constants, not one of which will even help explain high temperature superconductivity. And they have to grub for grant money just like everybody else. So I'll develop physics envy when I find a physicist who is doing work that I envy, not just because some people know how to curl their upper lip.

[vorpal]

Physics has the great good fortune to (mostly) confine itself to easily observable and quantifiable phenomena that can be described (mostly) in a few shortish equations.

I am an astrophysicist. James, have you ever seen Einstein's Equation written out in full? I just don't see any merit to your statement.

Nor do I see anything 'easy' about colliding atoms at the speed of light and then measuring their sub-atomic emissions.

You all are talking about the 'Standard Model.' Physicists are the first ones to acknoledge its shortcomings. But don't let their modesty fool you, it is also one of the most successful theories ever conceived. It has predicted the _existence_ of particles.

The reason why economists come under attack in more areas is because more people have vested interests in money than they do in astronomical observations.

In other words, the profession is rife with corruption. Many economists are willing to take money in order to lend gravitas to the position of a vested interest with a preconceived agenda.

Its not all their fault. The media is more than happy to interview these people without giving the viewers proper warning on the conflict of interests.

Honest economists need to be more vocal. They need to be able to criticize their comrades when they do not acknowledge confilct of interest or when they say something that is unsubstantiated.

[Arne (not anne)]

"a few fringe writers who could be counted upon to support any point of view no matter how wild"

I was going to comment on Lafayette's misrepresentation of scientific knowledge in Columbus' time, but it has already been covered. The quote comes from piglet's link and refers to an 1896 book which popularized the tale.

I thought the quote was very pertinent to this discussion because at this point fringe writers about astrophysics do not get published in the same way as fringe writers in economics.

[James Killus]

Wow, vorpal, that's quite a leap. Is it your opinion that Creationism is evidence that evolutionary biology is "corrupt?" Or that global warming denialism is evidence that atmospheric science is "corrupt?" And how about them Big Bang deniers, eh?

Now the Einstein equation for the photoelectric effect (making severe allowances for the limitations of ASCII) is

E(sub)k = h^(v - w)

in which E(sub)k is the maximum kinetic energy of the ejected electron, h is Planck's constant, v is the frequency of the incident light, and w is the work function. It's true that it's incomplete, since it doesn't take the Fowler correction into account, but it's a pretty good approximation.

But I'll hazard a guess that this isn't the "Einstein's Equation" that you had in mind, although it was the only one that I referenced (I do hope that your emphatic way of expressing yourself did not lead anyone to believe that Einstein only had one equation, nor worse, that you think he only had one). In fact, I'll hazard a guess that you're thinking of General Relativity, wich is indeed a real Mother, which is why I put "(mostly)" twice in my description.

But, you know, I'll bet you've never seen the full input-output matrix of the U.S. Economy, nor the chemical kinetic equations that are used to simulate (in a greatly simplified fashion) photochemical smog formation. As I once said in another context, no matter what you do, you'll eventually find yourself inverting some damn matrix or another. And thank god for computers.

There can be enormous complexity deriving from simple equations. I imagine you already know that actually, but you've got your back up because you think that someone is disrespecting your field. Well, well, once again I get to say, "Welcome to the club." Only I'm not disrespecting physics at all; "merely" a certain attitude that can be found in certain physicists. I named no names; I leave it for people to self-identify.

[James Killus]

Oh, and another more general thought occurs to me. There are various entities in society, most especially the Federal Reserve Board, part of whose job it is to _prevent_ recessions. To this end they use economic models. Is it so hard to understand why this makes recessions hard to predict? Only when various factors overwhelm the predictability and controllability of the system do recessions occur.

But is there anyone here who believes that, based on the science of economics, the Fed could not _create_ a recession if it so chose?

[Mark Thoma]

I've been afraid to read this. I'm not getting hammered quite as much as I expected.

A couple of quick comments.

To a: We dispensed with dark matter pretty quickly. As a theory, it didn't fit the evidence. A brief flash, made some news (because of the name mostly - I always thought they were kind of poking fun at how physicists solved a similar problem), people took a look, and it was rejected.

To piglet: This blog is not, for the most part, intended to present academic papers and academic debates. Nobody would read it if I did. I feel fortunate if I can get the basics out and blunt the misinformation that appears here and there. Go read my last post on the inequality debate and see where I tried to take it - to the journals where it will get done right. Again and again I said blogs, editorials, etc. are not the best place to sort this out. So please don't mistake what goes on here for what I am talking about - the research appearing in professional journals. If you want to understand what we do, you have to read that work, you can't think that reading a blog and watching us try to blunt some of the hackery out there will allow you to judge our professional work. It doesn't. Sure, we have debates about what will happen next, what caused what, etc., and noisy data make it hard to sort things out immediately, but those disagreements are precisely what science is all about - competing to see who has the best model and moving our knowledge forward. And the competition in economics is fierce relative to other fields according to what I've been told.

To Bruce Webb: You don't have to look far to see Uncle Miltie's monetarism being thoroughly thrashed. There are still a few defenders, but very few. We don't target monetary aggregates and won't anytime soon. We even reject them as intermediate targets.

To anne: Physicists themselves are quite dissatisfied with having to "make stuff up." When they actually know what the stuff is beyond "slow-moving exotic particles that have yet to be identified," I'll use different terminology. But the real message is simple - physicists live in a glass house too.

To vorpal: I hope people will allow us to separate our field into different groups too. There's a whole load of established theory that has withstood the test of time. Neuroeconomics is putting a few dings in some of it here and there (though there are very, very good microeconmists who would disagree with that), but nothing that can't be incorporated into the existing structure as far as I can tell.

To James: Thanks. I think you see what I'm trying to say.

[calmo]

Charming characterization of the Fed, James: the recession prevention club.
Maybe just such a relief from the usual: maximizers of stable prices and employment, toiling away endlessly through the night keeping the ship so buoyant...such unsinkable efforts.
Yes, that B it: you have turned the confidence that is inherent in the body on its head.

[piglet]

James Killus: quite some disagreement. Where to start?

Is it your opinion that Creationism is evidence that evolutionary biology is "corrupt?" Or that global warming denialism is evidence that atmospheric science is "corrupt?"
Creationism is not not supported by any competent scientists so the point is mute. Global warming is more complex. There is debate going on, some of which is legitimate scientific disagreement and some is politically motivated pseudo-science, funded by corporations with a vested interest, as is well known. Still, despite the fact that this is an immensely complex question, the scientific community has figured out in quite an impressive way what is going on, and reached a solid consensus. In the process of this research effort, scientists have found multiple original ways of using proxy data to understand past climate patterns. On the physics side, the greenhouse effect is well understood and laboratory tested. Atmospheric scientists' computer models are well tested and some of their predictions have already been confirmed. On a purely scientific level, these are extraordinary accomplishments and there is no substantial scientific dispute about this.

Now let's turn to some of your economics success stories. "economists do quite well in predicting price responses to a supply crunch, sieving data to show that corporate boards are gimicking their stock options, and showing that tax cuts do not "pay for themselves."" First, I hope I will not appear arrogant but predicting that prices increase if supply breaks down doesn't earn you a nobel prize. People have known that long before there was an economic science.

And the other two points are equally in line with Common Sense. Common Sense predicts that people who have the oportunity to fix their own salaries will get a raise more often than the rest of us. I do of course welcome the work of economists who have actually gone over the books to prove that this is the case, but in my memory, the main contribution of economists has actually been the contrary: again and again, we were told that markets are effective and if corporate pay has sky-rocketed, that was just a healthy market response. And those tax cuts that "pay for themselves" were quite popular with some economists, weren't they? Apart from Krugman, who in the economics community called the bluff? This is exactly one excellent example to prove the opposite of your claim. Far from offering vindication, it is an example how some economists allow their science to be hi-jacked for political purposes.

[Mark Thoma]

Looks like dark light wasn't so far off after all. This is from the Scientific American posted online today:How is dark matter like the Hundred Years’ War? The war lasted 116 years, and dark matter may not be dark. Conventional wisdom holds that darkness is the whole point of dark matter. Something is pulling stars and gas clouds off course; when astronomers go to look for it, they see nothing that fits the bill, so whatever it is must not emit or absorb light. ...

Nevertheless, astronomers have speculated for years that dark matter could power certain unexplained sources of light in the cosmos. That interpretation has been controversial—not least because each of these unexplained sources would require a different set of dark particle properties. Now a pair of researchers has found a way to make the dark matter explanation more plausible. “It’s an idea that actually unifies these things and can explain everything with one new particle,” says astronomer Douglas Finkbeiner of the Harvard-Smithsonian Center for Astrophysics.

One way dark matter could produce light would be to meet dark antimatter. The two will annihilate and ultimately yield a burst of gamma rays. Ten years ago the Comp ton Gamma Ray Observatory (CGRO) detected twice as much gamma radiation as conventional processes should produce. Dark matter annihilation would explain the discrepancy...

Another hint dates to the early 1970s, when astronomers detected vast numbers of electrons annihilating with positrons. Lots of things produce positrons, from supernovae to neutron stars, but not in such profusion, and these sources tend to lie in the plane of the galaxy, whereas the emission came from an ellipsoid. In 2003 a team proposed that dark matter annihilation generated the positrons.

The trouble is that positrons are lightweight, so the dark particles would have to be, too—about 1/1,000 the mass of a proton. Apart from the inconsistency with the CGRO results, such particles would somehow have had to elude detection in particle laboratories.

Finkbeiner and particle physicist Neal Weiner of New York University can explain both sets of observations with the same heavy particle. ...

Although the need for a novel force may seem ad hoc, the researchers argue that new types of particles generally entail new forces. ... Their model might also explain other astrophysical mysteries; for instance, it could be the heat source that keeps intergalactic gas toasty. In any case, NASA’s Gamma-ray Large Area Telescope (GLAST) satellite, scheduled for launch this fall, should be able to test it. As scientists gradually uncover what dark matter is and what it does, the particles will take on real names, and the term “dark matter”—which is really just a placeholder for ignorance— will itself fade to blackness.

[anne]

No; I do not understand the essay. Physics has a philosophical base, as does biology, as does economics or anthropology. I find any of these subjects fascinating, each on distinct terms. The idea though of criticizing methodology in physics as though this is a defense of methodology in economic makes no sense to me. The analogy of methodology however or of what can be expected in the way of experimental precision between physics and economics is not compelling or necessary to me.

[piglet]

James: "But let a physicist get involved in something deemed important..."

There is hardly anything in our daily lives that works well that is not based on physical science - electricity, electronics, architecture, aviation, telecommunication, medical diagnostics, time measurement, you name it. None of these could work if our physical knowledge of the respective phenomena were not reliable. It is not fair, of course, to compare the scientific record of social sciences with the precision of physics-based engineering. No reasonable person has this kind of expectation. And yes, I agree and have said this before that the inability to predict recessions is not a valid argument against economics. But since Mark, and now James, keep insisting on comparing economics with physics, and keep complaining that physicists have no "theory of everything" so they are hardly more successful than economists, it must be stressed that this comparison is totally ludicrous and economists can only lose in the contest.

Mark: "Physicists themselves are quite dissatisfied with having to "make stuff up." When they actually know what the stuff is beyond "slow-moving exotic particles that have yet to be identified," I'll use different terminology. But the real message is simple - physicists live in a glass house too."

I don't want to get into too many details but you have to understand that the characteristic of physics in the last century has been the discovery that reality is "weird", "exotic" or however you want to call it. This isn't a failure of physicists to come up with the right equations. In fact, physicists have come up with extraordinarily accurate equations. But our minds have a hard time making sense of them. As hard as we may try, we cannot imagine the world other than as a three-dimensional flat space, and our minds cannot grasp the real behavior of subatomic particles. There is nothing in the macrophyiscal universe of our experience that could prepare us for the counterintuitive and bizarre quantum world. You have to bear this in mind when you sneer at physicists "making stuff up". Physicists are inventing hypotheses that may seem bizarre, but that is not because they are a loose bunch. It is because "bizarre" hypotheses have proved successful in the past.

Economists are free to do the same - and they do invent some bizarre hypotheses, at least that is what many people think - as long as they credibly test them, and as long as they do not try to pass their hypotheses as reality.

[Mark Thoma]

"criticizing methodology in physics as though this is a defense of methodology in economic makes no sense to me."

It makes no sense becasue that's not the point. I'm not saying their methodology (in general) is bad. They're saying mine is. I object to that.

[a bit more] One way to object is to point out they can't explain the motion of matter in the universe due to gravitational forces - their model's predictions don't work out for this basic exercise - unless this stuff, whatever it is, exists. So pointing fingers at economics and saying your models don't predict and therefore this and that doesn't really make the point. If the standard is models not predicting at some level (macro or micro), they have problems too. Big ones. Both disciplines have successes, both have failures, and we are working hard to move things forward.

I guess the fact that predictions about people's incomes that are wrong make the news before incorrect guesses about exotic particles, or that people get a whole lot angrier when we make mistakes that cost them a lot of money as compared to someone getting a theory about an exotic particle wrong, shouldn't surprise me. Economics predicts that. But they are all mistakes, theory that needs to be fixed, and the fact that other disciplines also have so many unanswered questions and problems with their models makes it harder to understand the stones that are cast our way.

[anne]

Mark Thoma:

"I guess the fact that predictions about people's incomes that are wrong make the news before incorrect guesses about exotic particles, or that people get a whole lot angrier when we make mistakes that cost them a lot of money as compared to someone getting a theory about an exotic particle wrong, shouldn't surprise me. Economics predicts that. But they are all mistakes, theory that needs to be fixed, and the fact that other disciplines also have so many unanswered questions and problems with their models makes it harder to understand the stones that are cast our way."

Understood finally, and agreed completely.

[Hal Varian]

I've weighed in on this debate in this essay:
http://www.sims.berkeley.edu/~hal/Papers/theory.pdf

My thesis is that economics should not be compared to physics but to engineering. Or, alternatively, not to biology but to medicine. That is, economics is inherently a "policy science" where the value of an economic theory should be judged according to its contribution to economic policy.

There are many who disagree with this view, but hey, let a thousand flowers bloom.

[anne]

Remember, as I have explained before, the nature of knowing biology is different than in physics, similarly with economics. Neither knowing however is superior as such. The nature of the problems to be solved differs subject to subject. Each living organism is distinct, so knowing for a population is knowing on a gradient or differences. Physical knowing can take crystals as crystals, but robins are not robins each being distinct.

[Mark Thoma]

Thanks Hal. That looks like something worth posting on the front page.

I also recommend this article of yours from the NY Times:

Avoiding the pitfalls when economics shifts from science to engineering, New York Times, Aug 29, 2002, Hal R. Varian

[Jim Harrison]

The difference between physics and economics can't be summed up by deciding which science makes better predictions because economics would be radically different than physics even if it were comparably successful in empirical tests. The quantities measured in physics are descriptive--length, duration, mass, force, energy, etc. The quantities used in economics are proscriptive. They are cultural entities that are created by human will. Economics is much more like linguistics than like physics. It formalizes a system built on the customs of the tribe. That doesn't mean that there's anything wrong with it, but simply that it is a category mistake to think it's doing the same sort of thing as a physical science--unless somebody would like to tell me what the GDP is in joules or watts.

In a formula whose explication is left as an exercise to the reader: Physics is about the mating, economics about the marriage.

[James Killus]

piglet, you will find quite a few people who _claim_ to be scientists who support Creationism or some variant of it suich as "Intelligent Design." In fact, some of these folks have published papers, even in reputable journals, just not on Creationism (e.g. some are physicians, etc.).

There are also legitimate scientists who continue to be global warming skeptics, and a greater number who function as "hired guns," accepting denialist money. This is an almost exact similarity of what vorpal was accusing economists.

My argument was a reducio ad absurdum of what vorpal said. I do not consider physics, atmospheric science, or economics of being corrupt; vorpal was making that accusation.

Lastly, I'm probably the last person to note misspellings, having recently had to convert to nine finger typing, but points are "moot," not "mute."

[Hot Air]

Just finished reading "The Trouble with Physics - The Rise of String Theory, The Fall of Science, and What Comes Next" by Lee Smolin a prominent physicist who seems to know or have worked with every prominent physicist, including string theorists, you've ever heard of or not (Hawking, Penrose, Ed Witten, Leonard Susskind, Stephen Weinberg etc., etc., etc.,) It's a brave book published just last year that makes a compelling case that the last 25 years of research in physics focused on string theory can't be called science. It makes no predictions testable in the real world that distinguish it from other competing theories. More importantly, the theory includes so many adjustable constants, that if expirimental results do not confirm a prediction the explanation given is that the constants need to be modified and the current expirimental apparatus is not able to measure the effect with the modified constants. When unexpected observations in the real world are made that contradict the theory, new constants and new dimensions are invented to explain the unexpected observation. The end result is that there really isn't one string theory, there are 10 to the power of 500 string theories. One of these, in theory, might describe our universe, but they don't know which one yet.

The book is a devasting indictment of the current state of the field of physics (no puns intended). The bright spot is that it explores of lot of other competing theoretical approaches that have been incubating over the last 30 years that do make testable predictions and explain observed anomalies like dark matter and dark energy. Physics appears to be on the verge of a course correction in the next decade. I graduated BA in economics about 20 years ago. Having monitored this blog and others for a few months it doesn't seem like much has changed in the field in that 20 year period. I'm not optimistic that much will have changed in economics 20 years from now since it seems to encompass all the same problems identified with string theory above.

[Bruce Webb]

"To Bruce Webb: You don't have to look far to see Uncle Miltie's monetarism being thoroughly thrashed."

Well it is remarkably robust in policy circles.

The fundamental problem of economics is that it has allowed itself to directly inject itself into policy debates in ways that are way out of proportion with the rigor of the underlying theory. People push policy choices based on cartoonish distortions of Adam Smith just as people pushed disasterous policy choices based on cartoonish distortions of Karl Marx. That every student in every graduate seminar on macro understands that these depictions are cartoonish doesn't seem to follow through when your colleagues produce cartoons for public policy discussion.

Take a personal example. A couple of years back a well known economist posited a 3.5% payroll gap in Social Security and used that as a point of departure to argue for private accounts. When I enquired "Whence 3.5%? The then current Report says 1.89%?" not only that economist countered on the blog but the even more well known host economist countered by e-mail that if you just consulted the numbers in table IV.B6 Unfunded OASDI Obligations for 1935 (Program Inception) Through the Infinite Horizon sure enough you would find 3.5%.

Sure enough if you looked hard enough you would find out that a entirely new measure had been slipped into the Report in 2002. Suddenly we were expected to make policy decisions not on a number (1.89%) projected over the standard 75 year window, instead we were expected to use one projected to heat death of the sun (3.5%). Both economists knew the context, neither thought it important to actually mention it to the reader. Guest poster Prof. A simply presented Infinite Future blindly and Host Prof. B blandly let it pass and then sort of chastised me for having missed the new data point.

Well Infinite Future Horizon projections were and are junk science. In my opinion they were injected into the 2002 Report for the first time because they were convenient hooks on which you could hang privatization schemes by using scary numbers. Yet some sort of perverse professional respect set in, junk science was allowed to pass as the basis for policy prescriptions.

Certainly there has been some push back, this site being a key example, but time after time smart people who know better like Bartlett, Samwick and Mankiw are given professional deference coupled with willful blindness about the ways they are carrying water for their political masters.

I admire Bruce Bartlett very much. He probably is the sharpest pencil in the Right Economic pencil box. He writes and argues so well that many people don't even grasp that he is fighting hard for the other side. But he is an advocate for a particular set of policy prescriptions revolving around Supply Side tax cuts.

There is nothing morally wrong about being a defense attorney, quite the opposite. But you shouldn't allow them to dress up in white lab-coats and pretend to be scientists. And that is the crux here. Certain economists have allowed themselves to be converted into paid shills, just as certain climate scientests have become shills for gas companies, and many health researchers have become shills for tobacco companies. Yet while most conventional scientists know junk science when they see it, a whole bunch of junk economics simply gets a free pass on the basis that all concerned are just concerned about truth, justice and the American way.

Well no. There is an ideological battle being waged here. Its not new, in fact it goes back throughout history. My own field is not immune, it is said that "The victors write the history". Then again most non-Marxist historians have never claimed equivalence to physics.

[piglet]

James: "piglet, you will find quite a few people who _claim_ to be scientists who support Creationism or some variant of it suich as "Intelligent Design.""

I happen to have read William Dembski's "Intelligent Design" book and I was surprised how little effort the author makes of appearing scientific. In fact, most of the book discusses theological questions, and "God" is mentioned numerous times. But this is really beside the point.

But I would like to know how you answer my main objections: that your alleged economics success stories are more than questionable, and that you wrongly identify physics with esoteric cosmological speculation while overlooking that physics-based engineering is proof of physical science's spectacular success.

[Randall]

I'm perhaps too late to this discussion to be noticed but I must point out that

1) Advances in physics, basic sciences and even economic theory arise mostly by the process of been confronted with "reality", strugling to make sense of it and proposing sensible ideas or theories to be "falsified" once more. To call this process "making stuff up" is not the best way to describe the basic process of creation of knowledge by us humans.

2) This "making stuff up" activity is what many call doing science, proposing and discovering new things. In this sense, that something in nature seems to clash with the current understanding of it is perhaps the best thing that can happen to a physicist! Either we discover that the equations do not apply or must be changed in certain conditions or something "new" is PREDICTED (or proposed) and then discovered (or shown to be false, in which case the process starts over). I believe all these is (or should) be true for those "doing" economic theory also.

3) The prediction of the existence of dark matter (rather than the need to correct the equations) seems to be a sensible one since last year:

http://chandra.harvard.edu/press/06_releases/press_082106.html
http://cosmicvariance.com/2006/08/21/dark-matter-exists/

[James Killus]

Ah, I forgot to re-check this thread so I'm a bit tardy in reminding piglet that I do not "wrongly identify physics with esoteric cosmological speculation while overlooking that physics-based engineering is proof of physical science's spectacular success." My example was atmospheric science and global warming, not cosmology.

And I was thinking more of folks like Behe and Ayoub, not Dembski.

[Meh]

Mark, I think you need to take a good, hard, long look at your profession. The problem isn't that there are hacks on the opinion pages of the WSJ. The problem is that there are hacks publishing in peer-reviewed journals, getting papers published in _economics_ supporting things like Uncle Miltie's monetarism etc.

Many of them go on to get tenure.

The basic distinction between Economics and other disciplines isn't about the scientific virtues (as you note, analysing a moving, living system is always going to be harder going, with more mistakes than predicting projectile trajectories.)

No, the problem is that you have next to no accountability in your field and next to no ability to actually sort the wheat from the chaff, as evidence by a lot of the papers published.

It's almost entirely a professional, rather than a methodological problem. You can't get too far in Physics by denying the evidence. You can do ok for a while, but things generally catch up with you, then you're in trouble.

In Economics by contrast, you can get tenure denying the evidence, hell, it can give you a leg up, as you well know.

And that is the problem.

[piglet]

James, you are still ducking the hard questions (Mar 13, 2007 3:48:41 PM), here as well as on the other thread. Which is quite embarassing and won't earn you much respect, I have to say. I agree that the one who brought cosmological speculation up was Mark, not you. But you wrote this piece of tortured argument: "But let a physicist get involved in something deemed important, like, oh, I don't know, the heat and mass transport of the atmosphere as modified by changes in trace gas composition, and suddenly, Lo! it becomes just another example of "junk science" if it fails to kowtow to someone's perceived self-interest." Utter rubbish, sorry. And totally oblivious to the role physics is playing in our techno society.