5 suggested common themes for an Economics that takes its subject matter seriously
from Bruce Edmonds
This is another contribution under the theme of Geoff Davies’ essay “The Nature of the Beast”, which considers the question of how do we build a narrative linking the various heterodoxies. 5 possible uniting themes are suggested.
1. Economic phenomena are social phenomena
Economics involve all sorts of intelligent, social and adaptive behaviour including: social norms, fashions, identity, context-dependency, trust, friendship etc. Economic exchange can involve all these things, is embedded in our social and cultural life and can often only be fully understood in its social context. The deliberate exclusion of “non economically rational” elements of sociology and psychology is simply not supported by evidence. Economic phenomena is just that part of social phenomena which involves the exchange or transfer of items of value. Social behaviour undoubtably came before economic behaviour in the development of humankind, it is the more fundamental category.
2. Evidence should be paramount
Although all science involves theorising and confronting theory with evidence, evidence should be the ultimate winner of any dissagreement between them. The balance in the physical sciences is more towards the evidential end of the representional spectrum, with many researchers making careers in collecting evidence and in inventing new ways of collecting evidence. A shift back towards evidence and away from theory needs to occur in Economics. It should be a fundamental principle that evidence is never ignored without a very good reason for doing so. This means various kinds of evidence should be taken seriously, including the narrative evidence collected in the “softer” of the social sciences (e.g. ethnography).
3. Accepting the complexity of Economic phenomena
Social phenomena are very complex. Whilst there are sometimes identifiable broad trends, this should not be mistaken for understanding. Many complex systems exhibit broad trends in overall behaviour but can still surprise us when the circumstances change – we should expect economic systems (e.g. real markets) to be the same. A change of expectations to the phenomena we seek to understand is in order – away from looking for a “clever” simple model/mechanism that will explain a wide class of phenomena towards dealing with the contingent, context-dependent, complex, various phenomena we observed. Economics might turn out to be more like zoology than theoretical physics. This is indeed a dissapointment, but simply ignoring the difficulty of our subject matter will not facilitate real progress. A mature science seeks to limit claims beyond its capacity, even in policy advice and grant proposals.
4. Rejecting analytic formalism where this distorts good analysis
(i.e. using appropriate tools)
Following from the above, there is no reason (apart from blind optimism) to hope that formal models that are adequate to our modelling goals and the phenomena we face will be at all simple, and certainly not analytically tractable. Previously there was no choice if one wished for a formal model, but we now have the option of simulation models – there is now simply no need to distort our phenomena to make it fit the tool of analytic mathematics. Analytic models that require assumptions that are implausibly strong or (worse) there is evidence against should simply be binned. (This does not mean there is not role for analytic maths, for example it can be used to check and analyse the properties of complex simulation models). Precise models that can be replicated and asnalysed are important to science, but there is simply no longer any need to use the wrong tool.
5. Recognising the need for clusters of related models of many kinds and levels
Following on from the last point, we are faced with a dilemma – complex models that relate more directly to what is observed but are hard to understand and analyse (i.e. relevance); or simpler models that dont relate to observations (at best to our ideas about what we observe) but that can Social phenomena are not only complex but that can be thoroughly understood (i.e. rigour). The truth is we need both rigour and relevance, which means we will not achieve this using one model or one technique. Rather we will have to make do with “clusters” of related models capturing the phenomena – different aspects, at different granularities, and at different levels of abstration. So, for example, we might acquire a series of representations of evidence from many different sources (ethnographic, statistical, social network, interviews, observations, lab experiments etc.), theser might be related to complex “data integration” simulation models that are consistent with as many of these as possible. Then this complex simulation might be a safe target for simplification and abstraction in other, simulation and alaytic models, since these can be adequately tested for relevance against the simulation model they are about.
Complex, relatively rich, but precise representations (records of evidence and data-integration simulation models, etc.) might be the closest thing to a common language for relating the multifarious stands of a wider economics that takes its subject matter seriously.
Leave a comment
—– look inside —– $5.94 / $20.00
—– look inside —– $4.90 / $8.00
—– look inside —– $15.99
—– look inside —– $5.99 / 12.99
—– look inside —– $5.93 / $12.99
—– look inside —– $4.97 / $9.90
—— Ugarteche, Puyana and Madi ——
Gerson Lima / Maria Alejandra Madi
Edward Fullbrook and Jamie Morgan
————— Michael Hudson ————–
Maria Alejandra Madi / Jack Reardon
————- Edward Fullbrook ————-
—————— Steve Keen —————–
————— Richard Smith —————
————– Gustavo Marques————
– Victor Beker and Beniamino Moro –
————– Lars Pålsson Syll ————-
—————– Stuart Birks —————-
Edward Fullbrook and Jamie Morgan
———— Armando Ochangco ———-
Shimshon Bichler / Jonathan Nitzan
————— Mauro Gallegati ————–
————— Herman Daly —————-
————— Asad Zaman —————
—————– C. T. Kurien —————
————— Robert Locke —————-
Real-World Economics Review Blog 
You miss the train absolutely. How many angels etc. Who cares?
Economics should help us to make decisions. Not in narrow sense of decision science. If so, it should pose the hard problems — how to achieve long-range sustainable survival of mankind in acceptable condition. In first approximation, how to avoid catastrophes looming in this “century of black swans” — natural, geopolitical, political, and economic. Formulate and solve the corresponding “optimization under radical uncertainty” problems. Derive dual prices, use them as corrections to market prices.
The problem is that such approximations do not work. They certainly don’t tell us important things such as when the market will crash. Understanding the hard problems will only come from some hard research – simply “wishing/pretending” that plainly inadequate techniques tell us anything real about these systems does not help. A wrong prediction expressed with some confidence is worse than no answer at all.
“…we will have to make do with “clusters” of related models capturing the phenomena – different aspects, at different granularities, and at different levels of abstration.”
This is the testimony of Robert Solow in a Congressional Hearing entitled Building a Science of Economics for the Real World. Sidney G. Winter, Scott E. Page and David C. Colander also gave compelling testimony. One criticism was that NSF funding for economic research was, in practice, confined to work based on the Dynamic, Stochastic General Equilibrium Model, which was brutally critiqued, and one significant recommendation was to bring reviewers from other economic approaches and other disciplines into the grant review process.
See: http://science.house.gov/publications/hearings_markups_details.aspx?NewsID=2876
What are the goals of humans since the enclosure of the commons? Mainly, to amass wealth, to acquire “things”. There is no commons anymore. There is only greed, corporatist goals accompanying a loss of personal meaning and effectiveness, and a loss of expertise (except perhaps that of stock market traders).
Unless the goals of individual people change to goals which include all people, what is the good of “Economics” theory?
I think these are excellent points. Perhaps I can elaborate on some of them a little from my perspective as a scientist studying a rather messy object – the Earth. Earth is not as complicated as people and social phenomena, but it’s much messier than laboratory physics, so it has given me a broader appreciation of how to proceed scientifically.
Point 1. No argument, well put.
Point 2. Evidence should be paramount. Yes, no question, if the aim is to understand the world we observe.
However in practice it’s not necessarily simple. I have personal experience of people claiming an observation invalidated a hypothesis. However some more detailed modelling revealed that the hypothesis accounted for the observation after all, and the hypothesis was supported. Given models that are simplified to be tractable (even with powerful computers), it is often not clear whether an observation is inconsistent with the hypothesis being modelled (imperfectly), or if the model is too simple. There is often room for judgement as to which observations are crucial, and which misfits fatal. Climate modelling provides many examples.
A separate point, evidence doesn’t have to require complicated econometrics. You look for a key observation that provides a nugget of insight. One of my favourites is the 1987 stock market crash. Market valuations changed by 30% or more in a day, though there was no significant change in the real world – 30% of the world’s factories had not been bombed overnight. This is a simple and powerful bit of evidence against the quaint “efficient markets hypothesis”.
Point 3. “A mature science seeks to limit claims beyond its capacity, even in policy advice and grant proposals.” Especially then! Any science worth the name must define the limits of its applicability. If the answer is “I don’t know” or “It could be anywhere between this and that”, then that is the answer and policy makers need to know that. Everyone muddles through life coping with uncertainty. The role of science is to reduce the uncertainty, not to pretend it can eliminate it.
A technical point. “Complexity” has a technical meaning in systems science. Complicated is a useful term that avoids confusion with the technical meaning.
Point 4. The role of “analytic formalism”. Scientists (mostly? often?) use models to gain insight, not to simulate. I once did a back-of-the-envelope, order-of-magnitude estimate and learned that a magma ocean would cool in a few thousand years, not a few million years, an important insight about the early Earth. If I could do all my science with such estimates, or with a few pages of calculus, I would do it, but I can’t, so I use numerical codes to simulate mantle convection. However I use them in the same way – to observe their behaviours and see if they resemble what we know about the Earth. Sometimes I can calculate an “observable” from a model and do a direct comparison, sometimes I can only observe general behaviour. Certainly economists could profitably wean themselves from elaborate analytical solutions of (still) highly idealised equations and learn to use numerical methods, but you still have to understand the limitations of your models and the limitations of your observations, and qualify your conclusions accordingly.
Sometimes people construct quite elaborate models, such as ocean general circulation models or climate models. One then ought to work hard to establish which model outputs are robust under input uncertainties. Some are more conscientious in this than others, but I think generally scientists are far more aware of the need than economists with their elaborate general equilibrium models.
Point 5. “Clusters of related models”. Yes, you can’t hope, at the beginning, to make a model that includes everything that might be relevant. Even if you did, you wouldn’t understand its behaviour any better than you understand the real world. You have to start with simplified models that are not only tractable (with or without computers) but whose behaviour you can understand. This is true even though we know we’re dealing with a system that has emergent properties, and so you can’t use a simple reductionist approach. An example of a good approach I think is Steve Keen’s macro modelling of Minsky’s financial instability hypothesis. He has been progressively adding factors, and looking at the resulting (nonlinear, sometimes counter-intuitive) behaviour to see if it captures anything of the qualities of real world behaviour. (See, for example, http://www.debtdeflation.com/blogs/2010/07/07/naked-capitalism-and-my-scary-minsky-model/ . That’s all right Steve, you can scratch my back sometime.)
As you accumulate simplified models of various phenomena, you have to worry if they’re compatible. In my field, geophysicists and geochemists came to quite different pictures of Earth’s mantle, one layered, the other not. Such incompatibilities tell you there’s something important you (collectively) don’t understand. It has taken about three decades to begin to bring the two disciplines into compatibility, and the arguing is far from over.
Through it all, you have to try to maintain some humility. Not a quality traditionally associated with scientists, and certainly not with neoclassical economists. However the good scientists know the limitations of what they know. Otherwise they’ll fail to progress.
This leaves the question of policy advice. The simple answer is that economists have been far to arrogant in offering advice. In many respects we would be better off if local shopkeepers offered economic advice. At least they know about balance sheets. Politicians go by the seat of their pants all the time. So do most of us in daily life. So being honest about what you don’t know will do the world a far greater service than pretending you’ve got it all figured out. A century or so of economic dysfunction testifies to the truth of this.
<– "sharon"
I think this is a little pessimistic. Political parties, environmental campaigns, baby-sitting circles, charities, hobbies, amateur sports leagues, religious events, families, wikipedia, academic research etc. are all not about acquiring "things". In all of these items and services of value are transferred from person to person (whether or not this is essential to the core activity) but not in a classical market/contractual/economic manner – i.e. there is a broader class of social activity which involves transfer of value but with a goal of cooperation and sharing. These activities are largely ignored by economists. This is the point of the SCIVE workshop (http://cfpm.org/scive) – to study some of these and bring them back to general attention.
What I find so striking about this great discussion is the lack of surprise.
That economic systems are embedded within social systems; or that their study should be evidence based; or that they are complex; or that our choice of analytical tools should be within the context of the problem at hand are surely not controversial.
Except in economics.
Perhaps we can summarize: the world moved on, including the all of science. But economic theorizing remained stuck with a worldview, toolkit, and problem set developed decades ago. All that goes on now is ever finer parsing of what is a well worn set of ideas. Neoclassical economics has entered an inert steady state from which even the failure of some of its core concepts cannot shake it. The torrent of criticism is meaningless: the subject is so disengaged from reality that examples of its failure drawn from the real world have no relevance. They can all be dismissed as examples of distortions in what, otherwise, would be an ideal. And it is the ideal that is taught and studied.
Hence our frustration and the lack of surprise I noted above.
Peter, my take is that neoclassicism was cloned off the kinetic theory of gases in the late nineteenth century, and it has been isolated ever since. Not only did it miss the later thermodynamics that included the second law (no perpetual motion), but it didn’t even get the essence of the science it was copying, namely that you have to compare your theorising with observations. Of course it has also missed the twentieth century, including anthropology, sociology, ecology, cognitive psychology, computers, far-from equilibrium systems, etc etc.
Neoclassical economists are appallingly ignorant people.
The essence of science has been observation of measurable data and prediction of cause and effect to the extent it is currently possible.
In economics the data includes logistical objective data on units, weights, distances, etc., and price non-data which begins with the fact that it is subjective opinion not a factual measure.
Added to this unscientific original data, is the gross error of money systems without purpose– in the real world of necessary consumption of the products of work to stay alive.
It is true that with these foundational flaws, nations manage to produce industrial miracles. So something right must be occurring as well as all the things wrong. Separating the two is a purpose of economics that current dominant systems ignore.
When Functional Finance insists purpose must guide money’s power to summon very hard work, it is not obeyed. Rather, money power’s to support good and bad results is accepted as creating purpose– one that remains obscure and even non-existent.
The system as a whole can best be described as “the rule of law”. And legal systems are notorious for the absence of purpose and glorification of mindless markets to distill mob rule.
In my experience, it is only in the game of war that science and technology trump law as the dominant discipline in operations.
This suggests reform of the “game of peace and preparation for the next war” to introduce purpose in as many formal equations as possible.
If money is managed on purpose and never accepted AS purpose, we may find our way out of the mess we are making of life.
<– Peter Radford
I have little hope that mainstream economists will take any notice of their critics – not in fundamental ways (I predict they will resurrect their old and mistaken assumptions in a new languages – neuro-economics, agent-based economics or whatever). There is nothing to stop others doing meaningful research on social phenomena of exchange, the only real bone of contention is the ownership of brand "Economics". It does still seem to be the case that policy makers go to "economists" for advice, despite their track record. What we probably need to concentrate on is those outside academia who think that mainstream economics is worth consulting and educate them.
Peter –
Max Planck (I think):
“A new scientific truth does not triumph by convincing its opponents and
making them see the light, but rather because its opponents eventually
die and a new generation grows up that is familiar with it.” – M. Planck
And Thomas Kuhn, http://en.wikipedia.org/wiki/The_Structure_of_Scientific_Revolutions :
According to Kuhn, the scientific paradigms preceding and succeeding a paradigm shift are so different that their theories are incommensurable — the new paradigm cannot be proven or disproven by the rules of the old paradigm, and vice versa. The paradigm shift does not merely involve the revision or transformation of an individual theory, it changes the way terminology is defined, how the scientists in that field view their subject, and, perhaps most significantly, what questions are regarded as valid, and what rules are used to determine the truth of a particular theory.
So yes, we should address those who are open to a new way of seeing things, and endeavour to displace the neoclassicists from positions of power (political and academic).
good points> I am really still on the fence about Kuhn. I read him just after college ’60s and didn’t really understand completely. I guess I should read him again, before i go out on a limb, but I think he got it wrong in a sort of over correction overshoot sense.
economics is reasonable as long as it is analyzed in ecological terms. Birds build nests collect food have babies and generally do not over populate. Of course we have screwed with that so much that we endanger extraordinary percent of all other creatures on earth.
But classical economics is the basis of the endangerment. Everything that doesn’t agree with the assumptions of clecon is ignored, externalized.
well there you have my comment.
charlie thomas forest biometrician now rancher and land manager, hoping to make a difference to our coevolutionary species.
On ARGeezer and the congressional hearing testimonies on DSGE models: highly interesting reading, but still quite general (except for the plea of Chari for more moeny for macro economic (read: DSGE) research, which implies that Colander is completely right). I’ve tried to deconstruct one DSGE article, to add some more specific criticism to the discussion, see below. The most original critique is (11), the most practical is (4).
The article: Mourougane, A. and L. Vogel (2008), “Short-Term Distributional Effects of Structural Reforms: Selected Simulations in a DGSE Framework”, OECD Economics Department Working Papers, No. 648, OECD publishing.
In the article, the next statements can be found (my numbering), close reading results in the following criticism:
“(1) DGE models are explicitly derived from the optimisation of agent behaviour under constraints.
The Arrow paradox states that community indifference curves are a logical impossibility. Even when one accepts the idea of individual rationality and the concept of utility, it is not possible to aggregate the individual preferences into a community indifference curve.
(2) Their use presents a number of advantages. First, they allow a wide range of structural reforms to be examined and possible spillovers between the variables to be taken into account.
So do, according to my knowledge of these models, traditional econometric models of the economy. Though far from perfect, these do in fact a better job, as they are based upon real data (see 8) and real reactions. These measured reactions are indeed based upon history and not on arbitrary neo classical assumptions – but that’s what exactly what makes them superior to the DSGE models. Simple variables like ‘consumer confidence’ and ‘producer confidence’ perform, as far as I know, even better when short run changes in consumption or production are predicted. It seems quite a waste of time to develop all these DSGE models, which predict less well than existing models and explain even less (see 1, 3). Why shy away from the real world!
(3) Second, these models are less subject to the Lucas critique as they are based on structural equations with sound microeconomic foundations.
A. The Lucas critique has to be restated. Lucas states that macro economic models should be based on ‘sound’ or ‘non trivial’ micro economic foundations. He should have stated that, according to him, they should be stated on neo classical micro economic foundations – a whole array of other approaches is available!
B. Neo classical micro economic foundations are far from sound. Consumer behaviour has, during the past decades, developed into a real science. In fact, rational choice theory is debunked by marketers. They use anything, from mathematics to business economics and from history to theology – except the idea of rational choice, as it does not fit the facts. See (Dutch editions): Kotler, P. m.m.v. G. Armstrong, J. Saunders, V. Wong en F. Broere (2006), Principes van marketing (second edition) (Pearson, Amsterdam, 2007) p. 227. Verhage, B. (2004) Grondslagen van de marketing (sixth edition), Wolters-Noordhoff, Groningen, p. 41.
C. According to me, micro economic foundations of macro models imply that all individual people and businesses are taken into account – which happens when the National Accounts are constructed: millions of individual data are aggregated into a meaningful system. DSGE models shy away from such kind of work – the so called micro economic foundations are not micro at all!
D. According to the National Accounts – and DSGE models state that they do use concepts obtained from the National Accounts – the sector ‘households’ includes hospitals, student dormitories, jails and other institutional households. It is at first sight not entirely clear to me that all these institutional households strive at maximizing the utility of their inhabitants. As with all neo classical models, DSGE definitions are fuzzy and sloppy to the extreme!
(4) Third, it is possible to assess policies through their effects on consumer welfare.
In the Netherlands, the Centraal Planbureau (an economic think tank) uses a model based upon the representative consumer with a Cobb-Douglas utility curve to model the housing market (why a Cobb Douglas function? Not because it fits the facts but because it is easy to manipulate). When interest on mortgages rises, this representative consumer starts to rent a little more, when rents increase, he starts to own a little bit more. However: especially since the mortgage and banking crisis, not everybody can get a mortgage anymore. You’ve survived cancer? Good for you, but you won’t get a mortgage. You’ve got your own business for less than three years? No mortgage! And so on. This is not incorporated by the model of the representative consumer. Of course, this is a practical example of the statement of some of the critics of the DGSE models that they do not incorporate these kinds of differences.
(5) Finally, DGE models encompass dynamic effects and are well suited to examine the adjustment to changes in economic structure and policy.
See 6.
(6) However, the lag structure reflects the optimisation-based micro-foundations and is generally limited and similar across regions or countries.
Right. According to the model, we’ll always end up in a Pareto optimal GE situation. The omnipotent, everlasting and all seeing representative consumer will take care that we’ll be in heaven.
(7) Consequently, DGE results may tend to overemphasise similarities and to attribute differences to shocks rather than to economic structure.
Everything threatening the neo classical world of make believe is simply assumed away.
(8) The empirical validation of DGE models is an important concern, but also an active field of economic research.
We do have such a lot of macro economic data. The USA National Accounts for instance contain data which show that Asian households, on average, earn about 125 to 130% of average USA household income (very slightly rising since 2000), whites earn about 105% (more or less stable since 1967), hispanics earn about 75% (more or less stable since 1982) while afro Americans earn about 68% (stable between 1967 and 1994, very slight rise afterwards). These differences are not incorporated by the DSGE representative consumer (think of the Arrow paradox!), just like differences between Gross Domestic Product and Net Household Income (about 40%). Why don’t we develop models directly based upon the National Accounts, which can use these data? We do have the data, we do have models and systems that incorporate these data – and still DSGE models are being developed… it’s like physicists who deny the existence of the law of gravity.
(9) In order to focus on the distributional effects of reforms, a heterogeneous household sector with two groups of consumers is considered.
See 10.
(10) The first group maximises intertemporal utility over an infinite planning horizon in the presence of habit persistence.
Not just unlikely, but plainly wrong. According to DSGE models, nothing ever really changes. The rise of literacy, the internet, birth control, better health – it just does not make any difference in the way people try to shape their lives. Also, according to the neurologist Victor Lamme, people make choices first, then, after the unconscious part of their brain has made these choices, become aware of their own choices and consequently make up a story about the rationality and coherence of these choices – even if these stories are admittedly and measurably false (he calls our consciousness ‘the chatterbox’, Lamme, V. (2010), ‘De vrije wil bestaat niet. Over wie er echt de baas is in het brein’, Amsterdam.). In this perspective, DSGE models are just the most extreme variant of this rationalization process.
(11) The second group is liquidity-constrained … has no access to financial markets for intertemporal income transfers and consequently spends its disposable period income entirely on current consumption.”
Now, this is going to be a really interesting comment. There indeed are people who do not have access to the banking system. Collins e.a. however show that even the poorest of the poor are occupied with a costly and time consuming process of continuous saving and lending outside this system. Life events (marriages, religious duties and festivals, funerals were dozens or even hundreds of guests have to be fed and entertained as well as small scale investments) force them to be occupied constantly with a process to match their liquidity with their financial obligations. As money seems to lead to such a system, the assumption above of the liquidity constrained second group denies the very essence of money; money might very well lead to the problem of the ‘mutual coincidence of wants instead of solving this problem. Disposable period income is never entirely and precisely spent in the same period. It is easy to dream up such a situation – but it is entirely wrong and (though consistent with the neo classical idea on money) it denies the very essence of a monetary economy. A more evidence and less model based approach should have shown this to Mourougane and Vogel (Collins e.a. (2009), ‘Portfolios of the poor. How the World’s poor live on $ 2 a day’, Princeton).