Here is today’s extended version of our QOTD”

“[George Bernard Shaw] once said that as he grew older, he became less and less interested in theory, more and more interested in information… Nothing is so hard to come by as a new and interesting fact. Nothing is so easy on the feet as a generalization. I now pick up magazines and leaf through them looking for articles that are rich with facts; I don’t much care what they are. Evocative and deeply percipient theory I avoid. It leaves me cold unless I am the author of it myself.

“In the case of economics there are no important propositions that cannot, in fact, be stated in plain language… Complexity and obscurity have great professional value; they are the academic equivalents of apprenticeship rules in the building trades… They exclude outsiders, keep down the competition, preserve the image of a privileged or priestly class. The man who makes things clear is a scab. He is criticized less for this clarity than for his treachery.

“Additionally, and especially in the social sciences, much unclear writing is based on unclear or incomplete thought. It is possible with safety to be technically obscure about something you haven’t thought out. It is impossible to be wholly clear on something you don’t understand; clarity exposes flaws in the thought. The person who undertakes to make difficult matters clear is infringing on the sovereign right of numerous economists, sociologists and political scientists to make bad writing the disguise for sloppy, imprecise or incomplete thought.”

- John K. Galbraith in Writing, Typing, Alcohol and Coca Cola, from a speech at Harvard as it appeared in the Nieman Reports. (1978, The Atlantic)

via The Quotation Of The Day Mailing List

Category: Data Analysis, Philosophy

Please use the comments to demonstrate your own ignorance, unfamiliarity with empirical data and lack of respect for scientific knowledge. Be sure to create straw men and argue against things I have neither said nor implied. If you could repeat previously discredited memes or steer the conversation into irrelevant, off topic discussions, it would be appreciated. Lastly, kindly forgo all civility in your discourse . . . you are, after all, anonymous.

24 Responses to “QOTD: Data vs Theory”

  1. Lyle says:

    Actually the proposition about plain language is true for physics as well. It is a matter of really knowing the subject. Take Richard Feynmann because he trully understood he could communicate in plain language. One can explain the ideas without math, but it takes understanding.

  2. rtalcott says:

    Great find…the entire speech is worth reading….

    rt

  3. NoKidding says:

    Truth.

  4. postman says:

    Maybe in Shaw’s day, no important propositions in economics could not be stated in plain language. But today in practice, most economists are micro-applied econometricians, a pursuit requiring considerable math and stat training. A very small proportion of economists and a lot of people called economists–including business people and journalists–are macroforecasters. The imprecision of their forecasts should not impugn the entire profession of economists.

  5. RandyClayton says:

    I think this perhaps best illustrates the difference between scientific method and how the social sciences operate.

    I doubt a physicist like Feynman would buy into this. Theory — in Feynman’s universe — allows for prediction. Theory can be modified as a result of experiment and from this comes knowledge.

    Data? Information? Perhaps interesting or useful to fill columns or airtime on CNBC. It is not knowledge and it does not allow one to predict.

    “If I could explain it to the average person, I wouldn’t have been worth the Nobel Prize.”
    Richard Feynman quote

  6. postman,

    show us an example, from: “…micro-applied econometricians, a pursuit requiring considerable math and stat training…”, that will prove your supposition–it’s important, yet, too, complex to explain..

    (btw, GL w/that~)
    ~~

    this: “…Complexity and obscurity have great professional value; they are the academic equivalents of apprenticeship rules in the building trades… They exclude outsiders, keep down the competition, preserve the image of a privileged or priestly class. The man who makes things clear is a scab. He is criticized less for this clarity than for his treachery…”

    is the, simple, Fact of the Matter.

    and, in reality, if someone understands, what they’re talking about, they can explain it to anybody..
    ~~

    People, if they have any interest in understanding that JKG understood, quite, what he was talking about, should leaf through http://www.amazon.com/Industrial-Madison-Library-American-Politics/dp/0691131414

    there’s a real why our, current, ‘Economics’ pseudo-Science is orchestrated around JMK, as opposed to JKG..

  7. Bruman says:

    I propose that you fail to appreciate the dialectical nature of paradigm shifts, in that a well-vestimented emperor may as yet find himself percieved denuded by the pedestrian pedestrian.

    On the other hand, I think you have a good point there.

  8. Robert M says:

    BR
    I don’t know if this is my computer. I bookmarked your QOTD. When I took it to the article to do so I noticed the stories preceding and following it were on Wallison and trichet comments on commodity prices. The stories can be accessed by clicking on them.When I returned to the main page they aren’t there.
    Any ideas? Is it possibly a virus that blocs pages based on your advertisers?

  9. postman says:

    @ Mark E Hoffer

    There are numerous theorems of statistics/econometrics which, if applied incorrectly, would show that a utility’s law-abiding behavior is inconsistent with its regulatory requirements, that a lender is overcharging African Americans when it is in fact scrupulously fair, etc. etc. Correct application and understanding, on the other hand, will yield the correct conclusions.

    By definition, microeconomists work on one firm/industry at a time, and any one instance of applying these general econometric techniques is of great importance for their clients, but not generally and presumably, not to you. But the aggregation of all these applications covers a tremendous amount of ground and is important in that sense. And the theorems and their proper applications can’t be easily explained in plain language to someone without the requisite background.

  10. RandyS. says:

    “I wouldn’t give a fig for simplicity this side of complexity but I’d give my right arm for simplicity on the other side of complexity.” – Oliver Wendell Holmes

  11. Julia Chestnut says:

    JKG was a great man. It’s a smaller world without him. I do respect his son in his own right, though.

  12. postman,

    w/this: “…And the theorems and their proper applications can’t be easily explained in plain language to someone without the requisite background…”

    you’re kidding, right?

    or, that’s, exactly, the kind weasel-dom that’s being called out, in the QOTD (above)..

    seriously, would you give an ex.? then, we’ll see, or not, yes?

  13. formerlawyer says:

    @postman
    “There are three kinds of economists. Those that can add, and those that can’t.” – Hamish McRae (Governor of the Bank of England)

    “In a word, to understand the real world, one has to forget microeconomics.” – Emmanuell Benicourt and Bernard Guerrien, “Is Anything Worth Keeping in Microeconomics?” (earlier version of paper in Review of Radical Political Economics, Volume 40, No. 3, Summer 2008, 317-323) available online at: http://bernardguerrien.com/Radicalecoisanything.pdf

    “Make everything as simple as possible, but not simpler.” – Albert Einstein

  14. postman says:

    @ Mark E. Hoffer:

    If each observation’s errors in a regression equation are independent and identically distributed, free of autocorrelation and heteroskedasticity, and no error is correlated with any component of the vector of independent variables, then ordinary least-squares coefficient estimates are unbiased, miniumum-variance, and consistent. (There are numerous econometric methods to apply when one or more assumptions do not hold, but they are more complex to state, let alone the computer programming involved to apply the methods to a database.)

    @ former lawyer

    Thank you for the reference ,which given its radical pedigree, is likely not acceptable to most economists (which does not mean it is wrong). It’s a very long article, and I haven’t yet read it.

    More to the point, the methods I am referring to are empirical. Those criticized in your reference are theoretical. The former, not the latter, are widely-applied. The theory is largely restricted to academe.

  15. rmasand says:

    Ina similar vein, a professor of mine, big on eschewing obfuscation would say – ‘The depth of a book is inversely proportional to its thickness”

  16. formerlawyer says:

    @Robert M

    Its not a bug it is a feature!

    Actually, as I understand it Barry “hi-grades” his and other people’s posts as to what the general readership may find interesting. The links at the top of the page going forward and backward go to other posts that the devoted fan, and I count myself in that category, may find interesting as well.

  17. formerlawyer says:

    @ postman and Mark E. Hofer

    To my recollection – part of the derivative meltdown related to quant modelling – not so much the actual models (although that was also a part – see Black Swan events) but rather the use that was made of them by trading desks.

    Traders didn’t understand the models and the quants could not explain them in plain language – especially the significance of the limits and assumptions embedded in the models. To the traders this was a black box from whence the goodies came from. In fairness though, quants were championed as the “new masters of the universe” and inevitably it went to their heads.

  18. postman,

    w/this: “If each observation’s errors in a regression equation are independent and identically distributed, free of autocorrelation and heteroskedasticity, and no error is correlated with any component of the vector of independent variables, then…”

    I wonder if any such ‘errors’ are recognized, when found, if they do, even, exist..

    but, you’re stripping the exercise from its context makes it that much more difficult to discern, let alone explain..

    not sure what you were trying to exhibit..

  19. postman says:

    @ Marc E Hoffer

    You wanted an important econometric statement that required a math/stat background to understand . QED

  20. Robert M says:

    former lawyer
    The dates and times are in sequence. So where can i find the table of contents?

  21. postman,

    w/this: “…And the theorems and their proper applications can’t be easily explained in plain language to someone without the requisite background…” (from above, @13:13)
    +
    “…but, you’re stripping the exercise from its context makes it that much more difficult to discern, let alone explain…” (@16:50)

    see “…and their proper applications…”

    do you see, now, why I was mentioning: “you’re stripping the exercise from its context” ?

  22. postman,

    also, to be clear, this: “I wonder if any such ‘errors’ are recognized, when found, if they do, even, exist..”, was meant as a Q: , to you..

  23. formerlawyer says:

    @Robert M

    Good question. I don’t know.

  24. constructivemedia says:

    @ Mark E Hoffer & Postman

    Econometrics is a valuable tool for testing intuitions/hypotheses through empirical data analysis. Econometrics, done correctly, requires a significant amount of math & stat. Also, without the requisite background one will not be able to differentiate between junk econometrics and honest hypotheses testing. It is necessary for the econometrician to be able to speak in mathematics and statistics. (I believe this is postman’s point).

    However, upon finishing the process the findings can be expressed in plain English. If they could not be expressed in plain English they would serve very little purpose being that actions happen in the material world and do not take place in the realm of statistics and math, which are both useful tools for recording actions. (I believe this is Hoffer’s point).

    So you are both correct!

    Hoffer, as to the question about the ‘errors’, yes they are. The error term (disturbance) allows for unobserved factors affecting the dependent variable, measurement errors in the dependent and independent variables and is always a parameter of the model.

    Great quotes. Galbraith is always good company. I always enjoyed the jab he would take at tenured professors who would spend a career touting the value of labor insecurity for a healthy economy.