Tuesday, September 27, 2011


Sorry to bother you with more of this drivel, but you might as well know. Here is Krugman, spouting off again about how useless we all are.

I looked up a definition of "science," and came up with this:
The intellectual and practical activity encompassing the systematic study of the structure and behavior of the physical and natural world through observation and experiment.
So, as economists we are systematic, we study the structure of the economy and the behavior of individuals in it, and we observe and experiment. Thus, apparently, we are a science. But Krugman does not think so:
I’ve never liked the notion of talking about economic “science” — it’s much too raw and imperfect a discipline to be paired casually with things like chemistry or biology, and in general when someone talks about economics as a science I immediately suspect that I’m hearing someone who doesn’t know that models are only models.
Of course the chemists and biologists also have models, and those models are only models as well. What is it we do, if not science?

Here's what Krugman closes with:
...liberal economists by and large do seem to be genuinely wrestling with what has happened, but conservative economists don’t.
Apparently Krugman has not been to a macro seminar or serious economics conference for a long time. But maybe all the people writing those papers on the financial crisis are all liberals? I didn't ask. Maybe Krugman has in mind a particular definition of "wrestling," i.e. he/she who actually wrestles uses an IS-LM model.


  1. Really? You looked up a definition of "science" in a book and decided that macroeconomics fits the definition provided, ergo it's really really a science!

  2. So when can we expect our host to provide an argument responding to the reference Krugman post?

  3. OK. Here's something I wrote which is forthcoming in the AER:


    It's in part about the financial crisis, right? But maybe I'm not conservative enough for Krugman, so I don't count.

    Also, there's this:


    That's intended for John Quiggin, but it fits Krugman pretty well too, as they have much the same ideas.

    Read and enjoy.

  4. why is it that mathematicians, chemist or physicist never seem to have the urge to call themselves "scientists". my father is a scientist -- in the traditional sense -- but i never recall him referring to himselves as such.

    the only people i have ever heard referring to themselves as "scientists" are people who do political science, christian science, or freshwater macro types. i have never heard a game theorist refer to himself as a "scientist"

    go figure.

  5. Three out of the first four posters are worthless. I leave it to the audience to figure out which three.

  6. I have to say the reason I read Krugman's blog is that many of his predictions (not all) become reality. For example, I regret that I did not purchase more of the long term treasury bond. If I have done so, I would have made quite a bit of money.

    Personally, I believe the true definition of science is useful predictions. So how would your macro theory predict:
    1. Would Euro breaks up?
    2. What is the likely 10 year treasury bond yield 5 year from now?
    3. When would housing market normalize?

    If a theory cannot make predictions that can be tested, then I don't think it is science, at least not in the most rigorous sense.

  7. Here's another definition (fresh out of Wikipedia): "Science is a systematic enterprise that builds and organizes knowledge in the form of testable explanations and predictions about the universe." What does it mean if an economic scientist maintains that it doesn't make sense to "test" economic models because we know "all models are false"? We await more deep insight from 'Hooka Cat.'

  8. i like where this thread is going which is essentially towards the intellectual bankrupsy of the pretend scientist of the williamson ilk. the problem of the socalled scientist of this sort is that the only place their predictioms hold true, is in their own mind. then they create some artificial definitions of what constitutes science, and what not. for example we read a lot of the supposed problems in this blog about the supposed failures of old fasiom is-lm or modern new keynesiam type models on this page -- not because of testable implication that proved false -- but instead because it does not conform to some arbitrary protocol of how to specify assumptions in "scientific" (read: religious) ways. now i have a lot of problem with basic is-lm or nk models, but none have anything to do with the stuff this page spells out (such as that it is
    not specified in terms of preferences/technology/endowments).

    one of the wonderful aspects of this blog -- which i think isnfantaxtic -- it the insight it gives into the cult type mentality of certains corners of the profession.

  9. Steve,

    Do you have the drop on Nasar's new book? I've been having wet dreams about this sort of tome for a couple years now, but if it's a history of economics according to Krugman & DeLong, I'd tayjer just reread Keynes and be done with it. Nasar seems promising if for no other reason than her familiarity with academic economics and economists.


  10. It's a science but it does not mean we've been doing a good job.

  11. Just out of curiosity, Steve, you say "We observe and experiment." So before Daniel Kahneman and Vernon Smith, economics was not a "science" by this definition?

    Actually, that seems almost right to me. I'd say that before the development of instrumental variables techniques for using natural experiments, economics wasn't a science by this definition. But now it is.

  12. No, you can be a scientist and not do experiments. Further, there is plenty of heated debate about the usefulness of experiments. Some economists go so far as to say that natural and controlled experiments should be the only way we gather evidence. Others seem to think economic experiments are completely useless.

  13. Anonymous 7:41 is so smart, he can't spell bankruptcy properly. Go back to your hole, Luddite, and leave economics to serious people.

  14. Hooka Cat,

    Don't get carried away. We scientists are a polite and gentle bunch.


    No I have not read Nasar's new book. I thought A Beautiful Mind was excellent. There are some very interesting details in there for economists.

    "why is it that mathematicians, chemist or physicist never seem to have the urge to call themselves "scientists". my father is a scientist -- in the traditional sense -- but i never recall him referring to himselves as such."

    Everything is endogenous. If other people keep insisting you're not a scientist, you have to keep reminding them that you are.

  15. Steve:

    OK. So in the definition of "science" you posted, you interpret the phrase "observation and experiment" to mean "observation and/or experiment"? So you think experiment is not a necessary requirement for a field of study to be a science?

    I'm just wondering, since I always thought experiments were the essential heart of the scientific method (which is why I did an experiment for my dissertation instead of a theory paper).

  16. As a chemist observing a) 95% of economists failed to anticipate the biggest financial disaster in 50 years, and b) leading economists still don't even agree on the causes/cures of the previous or current disaster, it's hard for me to regard economics as a science in any overall sense akin to the physical sciences. I'd say you're at the alchemy stage - pre-scientific, but moving in the right direction.

    Jeff B.

  17. Scott Sumner has an excellent post on the question of whether economics constitutes a science or not:

    "Is economics a science? Should we care?"

  18. "Everything is endogenous."

    Except commodity prices.

  19. Noah, identify for me please all of the experiments in Charles Darwin's _The Origin of Species_.

    All of these stipulative definitions of "science" derived from a poverty of examples -- mostly taken from a junior high science class -- are not worth the bytes they written with.

  20. With regard to the predictive power evolution, consider:

    The discovery of various fossil records and various missing links between different species, the "evolution" of bacteria and the appearance of super-bugs.

    In fact, you may test the idea of evolution by doing experiments using bacteria in a lab. You can build a evolutionary model to predict and simulate what will happen if we have several species co-evolve together. You may also observe as well as quantify that once a natural predator is removed (i.e., wolf), the deer population will explode. All there are experiments that can be carried out and quantified.

    And in order to disprove the theory of evolution, you may:
    1. Try finding a specie that does not depend on DNA or RNA, or a specie that does not have any mechanism for evolution.
    2. Try finding a fossil record that does not have anything do to with any specie that comes before it, i.e., something that is "intelligent ally designed". For example, a silicon-based life form that suddenly appeared at some point in time, out of nowhere, and has nothing to do with carbon-based life form.

  21. Do astronomers regularly conduct experiments on planetary orbits?

  22. Observational sciences: not actually science at all, apparently.

  23. With regard to astronomers:

    How about the speculation on "dark matter" or "dark energy" in astronomy?

    How about the theory of big bang and how it was proved via measurements on cosmic background radiation?

    How about the experimental evidence supporting an expanding universe?

    How about NASA colliding a satellite into a comet to obtain samples for investigating the models of our solar system?

    How about the fact that the general theory of relativity was validated by measuring the bending of light? It is very similar to what you mentioned with regard to planetary orbits. No?

    Experiments do not mean you have to change something (i.e., create a star in a lab). It does, however, mean that one has to make a non-trivial prediction that can be falsified by observation or through other empirical means.

    Coming back to macro, is there any falsifiable prediction that anyone can make with regard to the current crisis? If so, what are the predictions? For example:
    1. If current policy remains roughly what it is, would 10 year bond rate go up or go down?
    2. Will we be in a recession in 6 month period?
    3. What steps we need to take to reduce the current unemployment rate to <7% in a two year period?
    4. What steps we need to take to resolve the Euro crisis?

    Note all these predictions can be falsified, at least in principle.

  24. Noah: Did you read, e.g., Angus Deaton and James Heckman on the usefulness of 'natural experiments' and 'field experiments' in economics? Perhaps doing an experiment for your dissertation wasn't such a hot idea (esp. if you want to communicate outside the echo chamber).

  25. "The demand for goods is not a demand for Labor"

    -- an insight well known to all economists prior to Keynes.

  26. This is all kind of silly, but regarding a couple points above, if "experiment" means "measurement" then economists have been doing experiments for a long time.

  27. I suspect "experiment" means "measurement" unless we're talking about economists.

  28. The dictionaries give the "definition" of whatever you are looking for. If you look up the "definition" of science, it hence gives you the "definition". This does not necessarily mean that though it gives you "what science should be", and/or "how should we define a science". These are I think different things.

    To call or not call economics as a "science" is to me beyond having a look at the dictionary. We might need to discuss epistomology (for example) through which we can think about. It may be even more fruitful if we examine the scientific methodology, the logic behind it. Some may think empirics is the core of a science, and for some maybe experiments are already enough.

    I think economics was not a science. Nor is it now. The reason we use similar methods and techniques does not make a study/dicipline a science. We use Newton's rules and calcules a lot. We use particular continuous-time models in financial economics es well. Game theory is by just itself a branch of applied mathematics. We use it quite often and even design mechanisms. They help a lot to solve economic problems.

    I think whether economics is science or not is irrelevant. We do not need to define it. Methodology and the way doing/using it is indeed crucial though. The dictionaries tell us what the definition of science is. They cannot answer the question of whether economics is science or not. As economists, we should find the answer, if needed.

  29. "They cannot answer the question of whether economics is science or not."

    Apparently you can't either.

  30. Here is the financial modelers' manifesto:


  31. Speaking from the humanities, I really don't see how modern "serious" macro is different from comparative literature. Replace Kristeva and Derrida with Lucas and Prescott and you've got the same general structure of self-referential emptiness. Very few people in macro use data as a lamp rather than a crutch.

  32. "Speaking from the humanities, I really don't see how modern "serious" macro is different from comparative literature. Replace Kristeva and Derrida with Lucas and Prescott and you've got the same general structure of self-referential emptiness. Very few people in macro use data as a lamp rather than a crutch."

    This comment just proves how little you know about what Prescott and Lucas do. There isn't a person on the planet who knows more about the data in the National Income Accounts than Ed Prescott.