Conception of economics: Comment from Andrew W

In the same post that Dr Doyle commented on, Andrew W linked to an article which he wanted us to discuss here (article is here).

As a critique of applied economics I accept the article (as there is definitely issues associated with the desimenation of economic tools into policy making), as a critique of economic science the article is horrendously off the mark.

One of the main issues I have with the article is the way it view economics vs physics – it is far too simplistic.  For example, the article states:

But statistical regularities should emerge in the behaviour of large populations, just as the law of ideal gases emerges from the chaotic motion of individual molecules

That is just the thing – the behaviour of individuals is not equivalent to chaotic motion (although I am sure many people would disagree 😛 ) because individuals make choices.  This additional element makes the whole study of macroeconomics (which I believe they are attempting the criticise) that much more difficult.

The “axioms” that the article criticises are all assumptions about this choice – factors that economists decided over 100 years ago that they would have to assume because they CANNOT observe choices in a sufficient fashion.  Of course, since then empirical and observational techniques have improved such that “the observation of the process of choice” is becoming avaliable.  As a result, these axioms will be (and in fact are being) challenged and changed.

I think it would have been good for the article to look at work on the methodology of economics before assuming that they could just pull out the critique of 17th century science and apply it directly here.

  • Andrew W

    The article at Nature is now behind a pay-wall, but it’s also here:
    http://xxx.lanl.gov/PS_cache/arxiv/pdf/0810/0810.5306v1.pdf

    I would compare the problems with economic modeling the that of modeling extemely complex physical systems rather than just chucking it all in the too hard basket by saying “individuals make choices”. We will never have perfect economic models mainly for the same reason that we’ll never have perfect atmospheric models, but climate scientists would – and do- make a huge mistake when they make a departure from science and start puting their own subjective spin on their observations.
    Perhaps thats the difference between “science” and “art” sticking to the objective vs allowing yourself to be swayed by the subjective. By surrendering to the temptation of including their own biases, individual economists ensure that they will be more wrong than they would otherwise be.

  • Andrew W

    There is another problem with creating accurate objective economic forecast that I don’t dispute, which is that while I’ll argue that while the actions of masses of people should, in principle, be predictable, what makes them far less predictable is influential individuals whose actions can’t be foreseen.
    I see the main driver of human actions as their instincts, and Human instincts are just animal instincts in a bigger cranium. How a mob of cattle acts is usually fairly predictable, except when you get that one #@^**#! animal that breaks from or turns the mob unexpectedly.

  • JEFFREY M DOYLE

    Economics Needs A Scientific Revolution

    My name is Jeffrey M Doyle. I ATTENDED

    THE UNIVERSITY OF MICHIGAN, ANN

    ARBOR AND MICHIGAN STATE UNIVERSITY,

    RECEIVING MY Ph.D. IN 1977.

    Economics is not physics. The behaviour of

    human beings is not as predictable as

    matter in physics [be it sub-atomic or

    macro.] While I understand the desire of

    many Economists to regard their Discipline

    as a deterministic, mathematically precise

    science, the reality is [with notable

    exceptions conceded] it is not a

    mechanistically elegant discipline like

    physics.

    INTERDISCIPLINARY UTILIZATION AND

    INTEGRATION OF PHYSICS, BIOLOGY AND

    ESPECIALLY THERMODYNAMICS

    INTO ECONOMIC ANALYSES, WILL GREATLY

    ENHANCE OUR ABILITY TO ADDRESS AND

    SOLVE CURRENT SOCIAL PROBLEMS–

    ESPECIALLY IN THE THE AREAS OF OF

    ENERGY AND ECOSYSTEM STABILITY

    I HAVE NOTHING AGAINST MATHEMATICAL BASED MODELING-TO THE CONTRARY I EMBRACE IT WHOLEHEARTEDLY. MY PROBLEM IS THAT UNLIKE [FOR EXAMPLE] METEOROLOGY, ECONOMICS DOES NOT WELCOME THE INTEGRATION OF NEW RELEVANT FACTORS INTO THEIR PRECISE MATHEMATICAL MODELS.

    WHEN METEOROLOGISTS ATTEMPT TO MODEL THE ATMOSPHERE OR PREDICT THE PROJECTED PATH OF A TROPICAL SYSTEM, THEY USE MULTIPLE MODELS, BASED ON DIFFERENT ASSUMPTIONS. IF THERE IS A GOOD AGREEMENT AMONG THE POSSIBLE OUTCOMES OR PROJECTIONS, THEN THEIR IS ALSO A MUCH HIGHER PROBABILITY THAT THE PROJECTED OUTCOME IS ACCURATE AND RELEVANT.

    I think the reputation of economics is strong enough that people think we have things to offer. So, as long as we are realistic and humble, we will have an audience and people will take economics more seriously.

    In the world these days, the complexity is such

    that people know that nobody knows

    everything. That’s not the expectation, the

    expectation is that something useful can be learned from economics.

    Blind perpetuation of the pursuit of internal consistency at the expense of external validity does not make Economics a science-it makes it a religion of collected ICONIC IDEALS-whose relevance cannot be questioned

  • Excellent blog! Interesting article and very informative! I will necessarily subscribe for this blog. http://lowsalego.com/map.html

  • “I would compare the problems with economic modeling the that of modeling extemely complex physical systems rather than just chucking it all in the too hard basket by saying “individuals make choices””

    But economics doesn’t put it in the too hard basket – we start from the premise “individuals make choices” and then try to understand how those choices produce phenomenon. Merely moving off statistical regularities without taking into account how these regularities occurred does not tell us anything.

    Economists cannot observe preferences, however we can create an objective description of how people make choices given their preferences. Once we have this we can apply some subjective definition of what we think these preferences are in order to get a result. The first bit is economic science, the second bit is applied economics/political economy/welfare economics.

    “Perhaps thats the difference between “science” and “art” sticking to the objective vs allowing yourself to be swayed by the subjective”

    Exactly, I think we are actually on the same page here.

    “Blind perpetuation of the pursuit of internal consistency at the expense of external validity does not make Economics a science-it makes it a religion of collected ICONIC IDEALS-whose relevance cannot be questioned”

    Although I think we agree in terms of a lot of points I don’t agree with this. The focus on internal consistency and model building is a separate task to that of policy making. External validity is only ignored in the case where the modeler is not aiming to provide a prescriptive portion to their model but is instead interested in checking the logical consistency of some fundamental premises.

    A model without value judgments will not provide a result. Value judgments are evaluated based on “external validity”. So the actual practice of applied economics will use this. The fact that some economists focus on checking the logical consistency of mathematical models does not make economics the same as “dogma” – these economists are providing a different service to the overall body of knowledge.

  • JEFFREY M DOYLE

    MATT:

    I THINK WE AGREE MORE THAN YOU MIGHT THINK. THE STUMBLING POINT, MAY HINGE ON MY STATEMENT THAT “…Blind perpetuation of the pursuit of internal consistency at the expense of external validity does not make Economics a science-it makes it a religion of collected ICONIC IDEALS-whose relevance cannot be questioned.” NOTICE THE USE OF THE PHRASE “…BLIND PERPETUATION…AT THE EXPENSE OF EXTERNAL VALIDITY…”

    YOU ARE ENTIRELY CORRECT IN YOUR STATEMENT THAT “…A model without value judgments will not provide a result…”. THE PROBLEM IS THAT RESOURCE ECONOMICS, ENVIRONMENTAL ECONOMICS, BEHAVIOURAL ECONOMICS ETC., ARE CONSIDERED A SUBSET OF MAINSTREAM ECONOMICS-IN FACT THEY ARE CONSIDERED BY MOST AS “HETERODOX” ECONOMICS. IF YOU CONSULT YOUR FRIENDLY DICTIONARY YOU WILL NOTE THAT SYNONYMS FOR HETERODOX INCLUDE: UNORTHODOX AND HERETICAL.

    MAINSTREAM ECONOMICS-WHICH IS THE PRIMARY PRODUCT OF EDUCATION-AT LEAST IN THE GOOD OLD USA-IS THE MOST INFLUENTIAL IN SHAPING THINKING. YES, THEY ACKNOWLEDGE THE TERMINOLOGY, BUT HAVE BEEN VERY SLOW TO INTEGRATE THAT THINKING INTO PREDICTIVE MODELS.

    MY ANALOGY TO METEOROLOGY IS WORTH REPEATING. JUST YESTERDAY I READ THAT NEW SUPERCOMPUTERS OPERATE AT OVER ONE TERABYTE PER SECOND. CERTAINLY RUNNING VARIOUS MODELS, WITH VARYING ASSUMPTIONS IS FEASIBLE, IF PREDICTIVE ACCURACY IS OUR GOAL. THIS IN NO WAY COMPROMISES THE NEED FOR INTERNAL CONSISTENCY WITHIN EACH MODEL.

    LET ME KNOW WHAT YOU THINK.

    JEFFREY

  • “I THINK WE AGREE MORE THAN YOU MIGHT THINK”

    I agree.

    I guess in New Zealand we don’t have the feeling that behavioural, environmental, and resource economics are peripheral. As a result I do view the core of economics as being somewhat larger.

    My primary distinction is between “economic science” and “economic policy”. Economic science is the pursuit of internally consistent models – piles of them – where as many of the relevant parameters are made obvious as is technically possible. Economic science provides us with a framework for analysing issues – but no direction on how to.

    Economic policy involves the use of value judgments and certain empirical “facts”. When these are combined to generalised models we get a policy conclusion – but not before.

    Ultimatlely, if I had to make a hard core of economics it might not include behavioural economics – but then it would not include macro-economics or applied microeconomics either. It would just be a mathematical description of choice and trade-offs.

    However, behavioural, environmental, and macro-economics would all enter at the same time – all apply different value judgments to the overall core.

  • all these capital letters are making my eyes hurt!