John Burnheim on theory and practice in understanding the world

In an exchange, John Burnheim sent me an email which seemed to me to be the effective condensation of a lot of good thinking. It certainly chimed with my own thoughts. So I suggested he clean it up and I’d reproduce it here, which I reproduce below. Because it is the conclusion of three emails, the first from him, followed by my response, I reproduce those two previous emails immediately below his final email – which is immediately below:

Don’t despair. Just face the problem that the more we know the more difficult it is to handle our problems. Give up looking for a kit of tools each of which has a specific use and collectively can do anything you need to do. Even in physics the variety of models that turn out to be relevant to some apparently simple problems tend to keep on multiplying. complications.

For much of my life I have been a devotee of sailing as “the poetry of motion” and a fascinating set of problems of design involving a host of diverse considerations, ranging from structural and accommodation problems, the capabilities of various materials, through hydrodynamics of flow, the interactions of hull with waves, to the aerodynamics of sails, each of which involves a different sort of theorisation, but also the impossibility of getting exact measurements such as might enable the relevance of variations of design to performance to be calculated from the theory, which itself is often an approximation. What theory does is to enable us to design ways of testing a particular way of dealing with a problem that identify its weaknesses.

Nobody who knows anything about such design problems imagines that progress will take us towards a single model that incorporates and integrates all those factors, or even that some of the apparently well establishes theories of such things as laminar flow are as simple as we have supposed. An old friend of mine, Tom Fink, was at one time prof of aeronautics engineering at Sydney University. He was very critical of the reliance of his Cambridge teachers on empirical generalisation in aerodynamics and set out to get the theory right. He once announced to me with pride that he had solved the problem of supersonic flow in TWO dimensions. I said “Surely it’s at least a three-dimensional problem. He replied that even though that was true, the development was a big step forward, and not without its practical use in circumstances where you could treat the third dimension as given, assign values to it and see what happens.

The trouble with economics is that the matters it deals with are much more complicated than the problem of sailing, but people want it to at least supply them with the sort of second-order considerations that they hope will enable them to make good decisions about particular matters even where there are no clear answers available to many of the questions that arise in the particular case. Philosophers are even worse addicted to simplification. Religions worse still. Emotionally for each personally, there is a strong hankering for clear direction, while collectively we seem to need agreed principles if there is ever to be unforced collective action on any other basis than arbitrary power. I think the agreement we need does not need consensus but compromise, based on awareness of the limitations of our knowledge and willingness to experiment honestly and prudently.

What is particularly dangerous is that people constantly fall for oversimple “truths” to rule out certain things as impossible. My favourite example of such an assumption is that heavier gases fall, displacing lighter. So clouds consisting of colder and heavier vapour should never be able to form, but the molecules of H2O should immediately fall to the ground. Why they do not do so is a complicated story involving quite different forces that do not evade the law of gravity but succeed in countering that force in certain circumstances. Economists habitually assume that certain sorts of motivation cannot be countered by other motivations in certain circumstances. The paradox is that many people embrace such simplifications precisely because they recognise that human motivations are so various and complex that no realistic analysis of them is possible. In fact, there is a strong case for saying that motivations are irrelevant to economics, which should concern itself with the consequences of acing in a certain way in certain hypothetical circumstances, including the actions of others.

One important advantage of such an approach to economic questions is that it facilitates constructive compromise. In assuring certain consistency and impersonality it should arrive at a representation of various outcomes that can, prescinding from other factors, be rated clearly in relation to the real problem. People may alter their prior subjective preferences in negotiation with each other in view of the possibility of achieving agreement to at least make a start at tackling the problem seriously. Moreover, enabling better negotiation between people who need to reach a compromise can, by exploring how much of the likes and dislikes of each party, accommodated them in relation to that particular proposal and its context. But that assumes that the models particular proposal is comprehensive and precise enough to minimise the danger of omitting seriously important factors. In practical situations what is needed very often is frank discussion of the likelihood of deceptive omissions and of insufficient weight to be attached to risks. Very often the risks involved in doing nothing turn out to be much less acceptable than those of some positive action. This is normally the case in business and in administrative decisions in most kinds of institutions that are forced to adapt to changing circumstances. Indeed, this sort of decision seems to me to be the central paradigm of practical rationality for people with diverse desires but a loosely defined important common interest in a very complex world acting on sound but limited knowledge to experiment with due meet the threats and opportunities the changing situation offers. Seriously considered action at least has some chance of improving a deteriorating situation.

Its obvious downside is that it is not emotionally satisfying, offering no prospect of a decisive triumph of some collective identity, no simple second-order rule or objective, and no assurance of salvation from fatal dangers. All of these satisfactions have their place in other contexts. Not many people find analysing and negotiating solutions to our problems a congenial occupation. What I hope for is that we will evolve and reward institutions that undertake those tasks, that the recommendations ar which they arrive will have a strong influence on public opinion, and that legislators and administrators will be constrained to follow those opinions.


On 7 May 2018, at 12:53 am, Nicholas <> wrote:

I doubt you’re out of touch, but economics is. It doesn’t really take this kind of thing very seriously. It didn’t in the Cambridge Capital Controversies where the Cambridge (England) guys won the intellectual battles, but it had no impact on the neoclassical model building – because (I think) model building and pressing on with the discipline is an academic imperative. Even if you know it’s wrong.

No good in physics or hard sciences of course because you’ll be shown up. But in economics refutation is the least of your worries – no-one’s ‘theories’ are very securely supported by the evidence over others. I expect there’s some good thinking on the subject – or good-ish – but I don’t know of it as I doubt it’s much use to my own endeavours and I’m very confident that it will end up in some academic cul-de-sac – professionally speaking.

That’s just not the way modern social sciences – or at least economics – develops any more, which is why I’m not that interested in a lot of contemporary academic economics. I don’t find it very insightful about anything other than the technicalities of its sub-field.

Cheers, NG

On 30/4/18, 6:15 pm, “John Burnheim” <> wrote:

Hi Nicholas,

A brief word about something that has troubled me for years. The classic Smith-Ricardo theory of markets seems to me to have been concerned almost exclusively with commodities, things that can be produced for consumption, and the role of markets in establishing the choices open to those involved in production and exchange of commodities in that sense.

The value of assets like land, natural resources and works of art obeys a different social dynamic from commodities because they cannot be produced. And so does money in so far as it canvey produced freely as credit.

Although each of the three categories depend on relative scarcity and relative desirability to motivate exchanges, it seems obvious to me that it is a mistake to treat money as the metric of scarcity and desirability. The crucial reason as I see it is that many needs for commodities are very inflexible on both the production and the supply sides, while both assets and money are instantly transferable.

This has numerous effects, most of which are occluded by such measures of economic well-being as GDP, basically because they do not take account of the diverse dynamics of the markets for different forms of exchange. The power of physics is that it has continually found more precise understanding of the dynamics of the very different forces, having long since abandoned the attempt to reduce those forces to a single vague category, even though the result of force is always measured ultimately in the same terms. I think that for economics to be successful, the way to go is not the present path of more sophisticated psychology, but more analysis of the dynamics fo markets.

Or am I a bit of touch?



This entry was posted in Economics and public policy, Science. Bookmark the permalink.
Notify of

Newest Most Voted
Inline Feedbacks
View all comments
5 years ago

Hi, I love economic models, but I always recall Paul Krugman’s warning not to take models too seriously. But to regard models as just a tool.

There is always a mental model somewhere lurking in our subconscious mind. Explicit modelling forces us to re-examine our assumptions, and take a more ibjexrice perspective of the world.

I also often recall Amartya Sen’s warning not to judge how the world is doing based on a single metric. Again, we do need a set of metrics to assess how we are going but we ought not let a single set of metric determine how we see the world.

I often wish that we had alternative ways of measuring progress The media seems to focus on GDP growth, inflation and unemployment. But what about inclusiveness?

Know Teeth
Know Teeth
5 years ago

Thanks Nicholas Gruen & John Burnheim.

John wrote “Even in physics the variety of models that turn out to be relevant to some apparently simple problems tend to keep on multiplying. complications.”… and Nicholas, maybe this is a footpath out of the cul-de-sac…
“”The availability of large data sets, combined with advances in the fields of statistics, machine learning, and econometrics, have generated interest in forecasting models that include many possible predictive variables. Are economic data sufficiently informative to warrant selecting a handful of the most useful predictors from this larger pool of variables? This post documents that they usually are not, based on applications in macroeconomics, microeconomics, and finance.””

And… the all that’s new is old again… we have not yet made copyright drag obvious as Common Knowledge. [ I’ve only recently come to understand Common Knowledge. It is important.] I hope someone asap replicates this study…

“Copyrights grant publishers exclusive rights to content for almost a century. In science, this can involve substantial social costs by limiting who can access existing research. This column uses a unique WWII-era programme in the US, which allowed US publishers to reprint exact copies of German-owned science books, to explore how copyrights affect follow-on science. This artificial removal of copyright barriers led to a 25% decline in prices, and a 67% increase in citations. These results suggest that restrictive copyright policies slow down the progress of science considerably.””

Effects of copyrights on science.
Barbara Biasi, Petra Moser 26 May 2018

And a bonus for John;

Know Teeth
Know Teeth
5 years ago

Phone fingers now crossed! I tried to link properly… my apologies.
Here is the link to …””The availability of large data sets, combined with advances in the fields of statistics, machine learning, and econometrics, “”…

Know Teeth
Know Teeth
5 years ago

Third time… pressed link and appeared in comment window but not published? Replace spaces. It is worth the effort. Toes crossed also.
libertystreeteconomics newyorkfed org /2018/05/economic-predictions-with-big-data-the-illusion-of-sparsity html

paul frijters
paul frijters
5 years ago

the problem with market interactions has been recognised by many for a long time. But doctrine is a very powerful force.
Where there is truly a streightforward prediction race, both mainstream economics and econometrics have lost out to more pragmatic and less doctrinaire approaches. Financial markets are the prime example of where machine-learning algorithms and pragmatic statistics now drive trading heuristics, despite the fact that it lacks religious appeal (ie a ‘foundational’ story).
The crown-jewels of economics are that scarcity forces hard choices, that trade in produced final goods is nearly always a good thing for the exporting country, and that market concentration is a bad thing with known counter-moves. What is fascinating is that knowledge of these crown-jewels precede economics by at least a century, that mainstream economic stories about trade are essentially wrong as to why it is so useful, and that the key stories told about market concentration are not consistent with the key doctrines of mainstream economics. Furthermore, the most important elements in running a country are in fact inconsistent with standard economic thinking (ie raising taxes needs tax compliance and market imperfections; and the basis of the capitalist system is not individual responsibility, but its very opposite, namely limited liability).

The disconnect between mainstream doctrinaire storytelling and the actual basis of our economy and the most useful activities of economists is fascinating. The mental contortions are amazing. We really are a very odd animal.