Is economics nothing but a library of models?

28 Nov, 2022 at 22:32 | Posted in Economics | 8 Comments

Chameleons arise and are often nurtured by the following dynamic. First a bookshelf model is constructed that involves terms and elements that seem to have some relation to the real world and assumptions that are not so unrealistic that they would be dismissed out of hand. monocle_chameleon_2The intention of the author, let’s call him or her “Q,” in developing the model may be to say something about the real world or the goal may simply be to explore the implications of making a certain set of assumptions … If someone skeptical about X challenges the assumptions made by Q, some will say that a model shouldn’t be judged by the realism of its assumptions, since all models have assumptions that are unrealistic …

Chameleons are models that are offered up as saying something significant about the real world even though they do not pass through the filter. When the assumptions of a chameleon are challenged, various defenses are made (e.g., one shouldn’t judge a model by its assumptions, any model has equal standing with all other models until the proper empirical tests have been run, etc.). In many cases the chameleon will change colors as necessary, taking on the colors of a bookshelf model when challenged, but reverting back to the colors of a model that claims to apply the real world when not challenged.

Paul Pfleiderer

As we all know, economics has become a model-based science. And in many of the methodology and philosophy of economics books published during the last two decades, this is seen as something positive.

In Dani Rodrik’s Economics Rules (OUP 2015) — just to take one illustrative example — economics is looked upon as nothing but a smorgasbord of ‘thought experimental’ models. For every purpose you may have, there is always an appropriate model to pick. The proliferation of economic models is unproblematically presented as a sign of great diversity and abundance of new ideas:

Rather than a single, specific model, economics encompasses a collection of models … Economics is in fact, a collection of diverse models …The possibilities of social life are too diverse to be squeezed into unique frameworks. But each economic model is like a partial map that illuminates a fragment of the terrain …

Different contexts … require different models … The correct answer to almost any question in economics is: It depends. Different models, each equally respectable, provide different answers.

But, really, there have​ to be some limits to the flexibility of a theory!

If you freely can substitute any part of the core and auxiliary sets of assumptions and still consider that you deal with the same theory, well, then it’s not a theory, but a chameleon picked from your model library.

The big problem with the mainstream cherry-picking view of models is of course that the theories and models presented get totally immunized against all critique.  A sure way to get rid of all kinds of ‘anomalies,’ yes, but at a far too high price. So people do not behave optimizing? No problem, we have models that assume satisficing! So people do not maximize expected utility? No problem, we have models that assume … etc., etc …

Clearly, it is possible to interpret the ‘presuppositions’ of a theoretical system … not as hypotheses, but simply as limitations to the area of application of the system in question. Since a relationship to reality is usually ensured by the language used in economic statements, in this case the impression is generated that a content-laden statement about reality is being made, although the system is fully immunized and thus without content. In my view that is often a source of self-deception in pure economic thought …

200px-Hans_Albert_2005-2A further possibility for immunizing theories consists in simply leaving open the area of application of the constructed model so that it is impossible to refute it with counter examples. This of course is usually done without a complete knowledge of the fatal consequences of such methodological strategies for the usefulness of the theoretical conception in question, but with the view that this is a characteristic of especially highly developed economic procedures: the thinking in models, which, however, among those theoreticians who cultivate neoclassical thought, in essence amounts to a new form of Platonism.

Hans Albert

A theory that accommodates any observed phenomena whatsoever by creating a new special model for the occasion, and a fortiori having no chance of being tested severely and found wanting, is of little or no real value at all.


  1. Why can’t you generalize the current consensus gravity model to galaxies, without adding an aether-like dark matter “epicycle”; and if physics has gravity wrong again, why shouldn’t I conduct my life under my own, much better, model, which allows nature to be as inconsistent as it wants?
    In other words, if bridges (most of the time) stay up because safety factors double the model-predicted tolerances, am I able to construct, for my own use, a more complete physics model that includes natural, and logical, inconsistency?
    In quantum computation, does linear algebra allow you to include contradictions inside matrices, thus hiding these natural inconsistencies from the higher-level math abstractions; and thus facilely implying that non-contradiction holds everywhere because it’s still there at the high abstract level? Yet, does the trivialist logic of simultaneous A and not-A yet persist, hidden away within vectors?

  2. Kuhn and Lakatos have a slightly different account of good and bad models. Kuhn’s Copernican Revolution is case study of the shift from Ptolemaic to Copernican astronomy. Copernicus’ first iteration of his model assumed circular orbits, and at the beginning was a worse fit with the data than the Ptolemaic model. But the Ptolemaic model achieved its fit with the data by arbitrarily adding epicycles until it did, not driven by the logic of the model.
    It took about one hundred years for Copernicus model to outdo the Ptolemaic model, but won converts long before that, because it improved the fit of model to data by making the model more realistic in its assumptions, not by adding arbitrary bells and whistles.
    Lakatos calls research which improves fit between data and model by making more realistic assumptions a Progressive Problemshift. Whereas adding arbitrary bells and whistles, which appears to be what you are talking about with your chameleons, is a regressive problemshift.
    Both Kuhn and Lakatos talk about new insights in progressive research programs being driven by the logic of the model. Whereas adding bells and whistles gradually reduces the empirical consequences of the model.

    • What about Aristarchus and his 3rd century BC heliocentric model, predating Copernicus by some 1800 years? And, wasn’t the Ptolemaic model driven by its logic that circles were the naturally perfect path for orbits?
      Could Dark Energy be a “bell and whistle” added to physics to reconcile observations of galactic orbits with the current consensus theory of gravity?
      In short is physics too just a library of inconsistent models, one for quantum scales, one for our scale, another for galactic scales, etc.? (Why does the vacuum energy predicted by quantum electrodynamics theory differ from the observed value by 10^120? Why does F = ma seem to hold here, but not for stars orbiting galaxies?)
      Thus, why even envy physics?

      • Physics is an example of a successful science that has improved our capacity to act in the world. Arguably, biological and ecological sciences are a better analogue for economics. The history of science is one of successful and unsuccessful analogy. Testing the analogy and choosing a better one is the trick. (See Mary Hesse Models and Analogies in Science)

    • Some people get carried away with the logic of logic and forget the logic of the world. The base presumption of Enlightenment science is that the world (reality) is a logical place, constrained and ordered by a functional mechanics, whether we understand it or not. What exists is thus implicitly a test of the logical necessity of what is proposed in analytic theory as possible. Building an operational model “mapping” the observable becomes a test of the logic of the world as opposed to the logic of the analysis, but the discovered logic of the world may have implications for the analysis, for example by exposing additional necessary and sufficient elements required for sound analysis. Elliptical orbits may not have seemed as ideal as perfectly circular ones, but inertia and a force of gravity could make an ellipse the outcome of a systematic relation between masses — a tremendous insight into the “logic” governing the world.
      It is worth noting the critical importance of the telescope and systematic measurement in astronomical observation, as well as the progress of an earth-bound physics seeking after abstract principles in changing modes of thinking that went way beyond the plausibility in outline of the Copernican model.

      • ‘The logic of the model” I am referring to is not analytic logic, but analogue logic. Lovelock’s Gaia model was first modelled with black and white daisies, which generated data approximating the pattern of empirical data. Then colored daisies were added to the model, then rabbits (plant eaters) and foxes (predators). This made the model more complicated but more realistic, and results of model runs both became more stable and more like historical data.
        In Copernicus’ case, improvement of the telescope led to revision of some of the astronomical data, and improved fit between Copernicus’s model and data.

        • An heuristic then to explore the case rather than a realized theory of the case

          • Not sure what your point is. Theories are always works in progress. Analogue models are the engine.
            See Nancy Cartwright Why the Laws of Physics Lie. Causes are true or not, theories aren’t.

Sorry, the comment form is closed at this time.

Blog at
Entries and Comments feeds.