Home > The Economics Profession > Discussion of the week: Is economics a science?

Discussion of the week: Is economics a science?

from Fred Foldvary, Leon, Bruce Edmonds, Antonio Garrido, Michael Meeropol, Dave Raithel, Ken Zimmerman, David Ruccio, Peter Radford, Jon Cloke, Vladimer A. Masch, Merijn Knibbe and Alice

Comments on Is economics a science? 

  1. Fred Foldvary
    November 22, 2010 at 4:20 pm | #1
    Reply | Quote | Edit

    Is the proposition, that the future is uncertain, itself uncertain? If it is certain that the future is uncertain, then that certainty becomes an axiom, among others, for economics to indeed be a science.

  2. Leon
    November 22, 2010 at 4:53 pm | #2
    Reply | Quote | Edit

    But economics can be a science, even if the current emphasis on mathematical modelling gets in the way. This is the argument made many times over by Tony Lawson. For an accessible overview see:

  3. Leon
    November 22, 2010 at 4:58 pm | #3
    Reply | Quote | Edit

    Sorry weong link. For the Lawson link on Ecomomics and Science see:

  4. Bruce Edmonds
    November 22, 2010 at 5:02 pm | #4
    Reply | Quote | Edit

    It depends on what one is claiming for the “science”. Zoology is a science, but concerns itself with categorizing lots of kinds of species, and then observing the characteristics of each (how they breed, what they feed upon etc.). It could simply be that economics is more like zoology than physics. That is there has been a mistake as to the *kind* of science it is, and what *can* be predicted with reliability using that science. If one can predict that, for example — that the prices in some markets will not be predictable, for well-understood reasons (the self-adjustment of markets to any temporarily successful price prediction scheme)b and many markets observed to see if this statement is backed up with the evidence — then *this* is science. What is wrong is that in current economics, having a precise model, is confused with the ability to predict (a confusion due to Friedman).

  5. Antonio Garrido
    November 22, 2010 at 7:07 pm | #5
    Reply | Quote | Edit

    Tha question is ill conceived.
    The importan thing is not if economy is an science or not (By the way What is SCIENCE?).
    The important is if it is USEFULL or not. Most usefull things, from accounting to a hammer are not science but technique.
    In Cambridge UK they have this idea quite clear: see the box of tools of Joan Robinson and what Keynes said about the economist as a dentist.
    A little modesty wont be so bad…

  6. Michael Meeropol
    November 22, 2010 at 7:35 pm | #6
    Reply | Quote | Edit

    But ultimately the question of whether economics is a SCIENCE boils down to whether we can LEARN ANYTHING from evidence. I believe we can learn some things by appeals to evidence. We can learn that increased inequality, suppression of the incomes of a large percentage of the population, defeat of labor unions, conquering inflation will not produce an investment boom absent a bubble. We have 30 years of evidence to demonstrate that since the imposition of Reagan-Volcker policies in 1979-81.

    The lesson however should not be written in cement because a changed structure (Mussolini style fascism or worse) could permit increased inequality, suppression of the incomes of over 80% of the population, defeat of organized labor and very low inflation to become consistent with an investment boom and more rapid economic growth — even without another bubble.

    Thus, there is “art” to the practice of our science. But I believe that unless we want to argue that it is impossible to learn anything about how economies like ours function from looking at evidence, economics is a science.

    It’s not an EXACT science because we cannot do controlled experiments. But astro-physics cannot do controlled experiments either. (and “medical science” is often quite inexact as well!)

  7. Dave Raithel
    November 22, 2010 at 9:47 pm | #7
    Reply | Quote | Edit

    Here’s another way round of thinking about the question: Are economists SCIENTISTS?


    “……In a new paper, Gerald Epstein, chairman of the UMass Amherst economics department, and Jessica Carrick-Hagenbarth, a graduate student, examined a group of influential economists, scoured publicly available resumes, biographies, articles, and interviews, and found that the majority had made money from financial institutions — but very few had disclosed these connections when writing, speaking, or giving interviews on public policy……”

    There’s some etc. before the paragraph, and a lot more etc. after…

  8. Ken Zimmerman
    November 23, 2010 at 2:00 am | #8
    Reply | Quote | Edit

    These are all good comments. They bring out the basic questions. There is not Science, with a capital “S” but many sciences, small “s.” Some are more experiential oriented, some more laboratory oriented, others more mathematically oriented, and still others more technique oriented. All are partial, incomplete, and none can depict the world. But all can play parts in building worlds and helping the implications of such construction play out. All want to show or explain or predict, or all of the above how and why things work. In this sense they are more or less useful. But they are always actively involved in world construction. None is or can be just a presentation of what is, since they all play active roles in making what is what is. In this sense neo-liberal economics is no different that other so called sciences such as Biology, Sociology, or Astrophysics. But neo-liberal economists are not just scientists, or in some cases may hold no wish to be a scientist at all. They are also professors, deans, private economic consultants, bankers, or hedge fund managers. How this meshing comes out is impossible to say in general and difficult to figure out in specific cases without a great deal of observation and description.

  9. David Ruccio
    November 23, 2010 at 3:15 am | #9
    Reply | Quote | Edit

    Friends, these are all good comments. For my part, I’m fine with Ken Zimmerman’s formulation about many sciences—although I tend to make the distinction based not on methodologies but on theories (e.g., neoclassical, Keynesian, Marxian, etc.).

    But, in all honesty, what I found interesting about Pollock’s essay was not the question—is economics a science?—but, rather the focus on uncertainty. In my view, the presence of uncertainty (radical, unknowable uncertainty, not domesticated uncertainty as in probabilistic formulations) both undermines the certitude with which mainstream economists make their pronouncements (about the causes of and solutions to the crisis) and gets at one of the key issues in the financial shenanigans that led to the crisis in the first place (in the sense of the shifting of risk caused by uncertainty from the big banks onto others). Uncertainty, in this sense, unsettles both mainstream economic theory and the presumed stability of the financial structure of capitalism.

    • Peter Radford
      November 23, 2010 at 8:31 pm | #10
      Reply | Quote | Edit


      I think you are being unusually gentle when you say that uncertainty “unsettles” mainstream theory. In my view it undermines it completely.

      Also, I think the “financial structure of capitalism” as it existed in 2007/2008 was an artifact of neoclassical design: most of the risk modeling, underwriting, and other aspects of finance were based on mainstream economic theory. All those processes changed radically over the last three decades after the hegemony of neoclassical theory emerged. Prior to that finance was far more “subjective” and less dependent on assumptions of rationality. This is why I feel economics needs to fix itself – it produced technologies for finance that failed to work. In that vein I have a different interpretation of what happened to the banks: they were mostly gullible, incompetent and lazy. They gave up “old fashioned” banking, for “modern” risk assessment based on idea they didn’t understand or think through. All those MIT professors and PhD’s flooding Wall Street were hard to argue with!

  10.   Jon Cloke
    November 23, 2010 at 12:02 pm | #11
    Reply | Quote | Edit

    This is surely where chaos theory comes in; the future is neither necessarily certain nor uncertain in a complex dynamic system, depending on the time-scale and how that time-scale is relevant to the lifespan of the system under consideration. Within chaotic systems (which is what the socio-economy is) there can be periods of apparent calm and stability which tend to overlie changes in hidden fundamentals. Thus the alleged NICE era (which in reality was no such thing) in which apparent stability of visible measurables concealed the substantial, fundamental changes being brought about by the massive liberation and revolution of financial services mechanisms.

  11.   Dave Raithel
    November 23, 2010 at 2:38 pm | #12
    Reply | Quote | Edit

    Mr. Ruccio might find the following of interest:

    Working Paper 2009-020A

    A Simple Model of Trading and Pricing Risky Assets Under Ambiguity: Any Lessons for Policy-Makers? by Massimo Guidolin and Francesca Rinaldi

    It pretty much beat me up, but I think they argue that even uncertainty can be priced; which of course means (since we now have a number) the mathematical project can proceed. How they get from there to the claim that one cannot inflate out of a financial crisis, I don’t know. But I get that Mr. Ruccio and others find “uncertainty” a uniquely different kind of problem (it’s not a phenomenon, right, so at most you can only measure people’s response to “it”) whereas some people do not ….since you can measure people’s response to it …? But hey, I’m mostly a spectator here, pointing to things I see in the spectacle going on around me…

  12.   Vladimer A. Masch
    November 23, 2010 at 6:16 pm | #13
    Reply | Quote | Edit

    Evolution recognizes only one type of success: long-term sustainable survival, avoidance of catastrophes. This century will be awash in threats to the humankind. Can economics (and adjacent disciplines) help in solving the arising problems? To be or not to be — that is the question.

  13.   Merijn Knibbe
    November 23, 2010 at 7:50 pm | #14
    Reply | Quote | Edit

    Dear Dave,

    I started glossing over your article – but decided not to read it as I encountered the next paragraph (see below). They are still modelling risk as a known variable, with a knowable distribution, a knowable variance and a knowable mean. The whole point is that in the case of uncertainty, the distribution, the variance and the mean are NOT KNOWABLE. Insurance companies do calculate the risk that people aged between 50 and 60 will get colon cancer next year. They might make a pretty good prediction. But they can’t calculate the chance that something like a new HIV-virus will pop up (by the way: why do the always use a normal distribution?).

    ” We model two kinds of risk: risk that affects the entire market (systematic risk ) and risk that just reflects circumstances peculiar to the specific
    firm represented by a stock (idiosyncratic risk). The systematic component of the stock payoff is normally distributed, with mean μS and variance σ2 S . Also the idiosyncratic component is normally distributed, with parameters μI and σ2I. The policy-maker/central planner performs two functions. First, using un-modelled
    economic policy tools, it sets the inflation rate at some level i. Second, the policy maker may ex-ante change features of the environment to favor “better” outcomes. Call d the total payoff on the stock. Because systematic and idiosyncratic risks are independent, we have d ∼ N (μI + μS, σ2I+ σ2S).

    • Peter Radford
      November 23, 2010 at 8:18 pm | #15
      Reply | Quote | Edit


      That paragraph seems to describe what is called the Capital Asset Pricing Model, which is [was?] a mainstay of finance in the aftermath of the triumph of neoclassical theory. It asserts that the price of a firm’s stock price depends on the two forms of risk mentioned. CAPM is just one of the technologies derived from neoclassical theory and in use throughout the financial world.

      I think there are other interesting points raised above.

      First: my view is that the entire premiss for a business firm is that it seeks to translate [I call it mediation] uncertainty into risk. It does this by making “plans” which are simply educated guesses that it tries to impose on the unfolding reality it confronts. If the revealed reality conforms closely enough with the plan, the firm makes a profit. If not it fails. This does not get rid of uncertainty – nothing can, it simply makes it tractable.

      Second: any theory that ignores uncertainty is detached from reality and has no relevant content with respect to reality.

      Third: I think economics qualifies as a science in the sense Ken points to. He seems to go further than I would though. I believe there is a set of principles that are common to all sciences, with a small case “s” that combine to form a notion of Science with an upper case “S”.

      Fourth: there is an abundance of scientific thinking that relates to uncertainty outside of economics, much of which needs a good look at by economists as they seek to rebuild from the beating the discipline took during the crisis.

      And, yes: we need to get rid of those Gaussian distributions!

  14.   Dave Raithel
    November 24, 2010 at 2:50 pm | #16
    Reply | Quote | Edit

    I appreciate the replies. I’ve probably got more questions than capacity for asking them. Here’s what I personally find to be the biggest problem you and yours have in making headway on your project here in America: We’re easily intimidated by mathematical bullies (just listen to any “business news” network.) I managed to slog through two semesters of calc and one of discreet math and some stats back in the ’80s. Never-the-less, since I do none of that for a living, reading arguments employing that language just makes my eyes bleed, so to speak. That’s one of the reasons I appreciate Steve Keen so much. The first time I met the Prisoner’s Dilemma, I thought and intuited and believed that any theory of rationality which implies rational people can make themselves worse off than they would be if irrational HAS to be wrong, and just stupid. The most gifted among you are going to have to find ways to reassure the rest of us – who are literate and analytical but simply poor mathematicians – that our intuitions are more correct than not; else most of us (well, not me, I just stay angry and suspicious) succumb to the intimidation. I can see that’s what the Kick-it-Over campaign is about – but it’s about 30 years too late for me…

    Again, thanks…

    • Alice
      November 25, 2010 at 9:00 am | #17
      Reply | Quote | Edit

      I will be so presumptious to say that the most mathematically gifted amongst us have to stop talking to themselves and start coming up with something that works in the freal world. Despite the explosion of maths based papers in the past three decades…who amongst economists really talks anything meaningful and who amongst policy makers has had any decent policies..

      for if it were so why are we here in the mess?

      I know, I know…the good economists were silenced by the career mathematicians (it happened – I live it on campus)…leading me to a deep suspician of anyone who speaks in symbols now. They are invariably a bully. The age of the computer model has a huge amount to answer for and how many really know or understand the maths they do or do they just replicate other studies, mumble to themselves and button push. Not a normative noble subjective thought anywhere. No deep seated civic feelings allowed. No duty of care to society permitted. Nothing but a pseudo intellectual arid mathematically myopic void which excludes all but a few rusted on peers.

  15.   Merijn Knibbe
    November 25, 2010 at 12:23 pm | #18
    Reply | Quote | Edit

    Contrary to a widely held point of view, much mathematical economics is not precize bur rather fuzzy – as the variables which are used are not precise but fuzzy. Again and again, it is, implicitely, assumed that a variable is indeed integrable or differentiable. Just ask a mathematical economists for the exact definition of his variables (households, utility, interest, capital, whatever) – and again and again he or she will answer you that, well, that that’s not something we have to think about. Ask again, and you often will find out that they have no clue how the variables relate to the real world. I’m not against mathematics. I’m against fuzzy variables.

    • Alice
      November 26, 2010 at 9:35 am | #19
      Reply | Quote | Edit

      I agree Merijn. The assumptions behind the variables are fuzzy, vague and belong to some other pre computer time…so what good is it when the maths is precise but the underlying assumptions are so disconnected as to be useless (for has not the world changed but the asumptions remained locked in a past era?). It makes no difference at all to the final output, despite the precision of the maths, it remains unworkable and meaningless…and act of artistic sophistry at best.

  1. November 28, 2010 at 1:02 pm

    Economics is an art, but Neo-Classical economics is science fiction!

  2. November 28, 2010 at 2:51 pm

    “Economics is an art. Neo-classical economics is hazardous to your health and science fiction to boot.” I’ll buy that. The missing ingredient for usefulness, (there is not need to make predictions of future events — there are so many of them,) is PURPOSE. Functional Finance demands purpose. And the economics that has let us down needs to be focused on and replaced by Functional Finance. Mathew Forstater and UMKC faculty and friends can take us to Abba Lerner http://positivemoney.us/ff.htm

  3. November 28, 2010 at 4:04 pm

    It is a science if it contributes to the improvement of human life. It stops being a science if it is exhausted in figures and in invisible hands.

  4. November 28, 2010 at 5:57 pm

    I offer classification of disciplines into constructive and destructive sciences. Stop talking about neoclassical, math economics, fuzzy variables etc. The real question is: is the lowest turtle constructive or destructive? Which is: maximization of something, assumption that something is “the best” because it extremizes the value of some variable, neglecting a zillion of externalities.
    Turns out, it depends on the state of society. Is turtle an acceptable approximation?
    Stage 1: 19th century till WW1 – production capacity chases demand. Turtle is an OK approximation — uncertainty mild, mood optimistic. Stage 2: till 1980s: demand chases production capacity. Uncertainty substantial, pessimism. Age of Keynes; turtle OK rarely. Stage 3: now — save the mankind. Absolutely destructive.
    Easy proof: When turtle OK, we can use six criteria of Decision Analysis (five “single” ones plus Hurwicz), math programming and other models etc., as tools of selecting the best, of decision-making. When turtle not OK, uncertainty great and externalities dangerous, confining ourselves to just one tool and (in Hurwicz — just one value of alpha, one point of horizon) is highly counterproductive. New paradigm: catastrophe avoidance. All tools no more than means of analysis, not decision-making. Necessary to analyze (by many tools available, not a single one) all horizon, not a single point of it. Analysis — to ferret out potential dangers under different scenarios and to construct contingency plans to deal with them.
    To achieve the full goal of planning and decision-making, which is both attaining intended consequences and avoiding unintended consequences, by constructing flexible and robust survivors. Good protective equipment decreases need of knowledge about the future. Amen.

  5. Peter T
    November 29, 2010 at 7:50 am

    The implicit definition of a “science” seems very restrictive. Is ecology or evolutionary biology a science? Neither predicts, but both explain. Do “sciences” generalise? Well, most do, but at different levels. There is often an area where things are known to just happen. History as a discipline tries hard to explain what really happened – it tries to mirror reality. But it’s not usually thought of as a “science”.

    Better question is – what should economics try to do?

  6. omahkohkiaayo
    November 29, 2010 at 5:46 pm

    I like the aphorism from Heilbroner that “mathematics has added to Economics rigor, and alas, also, mortis.” Science requires quantification so that variables, functions, parameters, scopes and depths of modeling, once defined, cannot be changed to suit a particular argument. But one can still wind up with elegantly quantified crap. Further, if prediction is the only “test” of the soundness of the logic and assumptions of deduction, there are the usual problems of confirmation bias, cherry picking data, sources, time frames, and methods that will “confirm” the prediction and thus the “logic” and adequacy of assumptions embodied and used in the logic. Where the philosophers are the “scientific salesmen” of the theologicians, the neoclassical economists are the pimps and whores (with all due respect to whores) of the philosophers of philosophical positivism in all of its interests, versions and manifestations.

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: