The El Farol problem goes digital

I saw this tweet and it immediately reminded me of Brian Arthur’s El Farol equilibrium story.

Please develop an app that simulates Waze recommendations to assess which routes will open up based on everybody else follow Waze.
13/05/2015 07:12

For those who don’t know it, El Farol is a bar in Santa Fe, where Arthur is based. It has great Irish music but the problem is that it’s no fun if the bar gets too crowded. What is the outcome in this situation where people are trying to choose based on beliefs about what others will choose? There is no ‘deductively rational’ solution to this choice problem. In his paper, Arthur writes: “If all believe few will go, all will go. But this would invalidate that belief. Similarly, if all believe most will go, nobody will go, invalidating that belief. Expectations will be forced to differ.” Simulating behaviour of 100 agents repeatedly shows that mean attendance at El Farol quickly converges to 60, but with large swings from period to period, and changing identity of those attending – it isn’t always the same 40 with a varying number of extras. “while the population of active predictors splits into this 60/40 average ratio, it keeps changing in membership forever. This is something like a forest whose contours do not change, but whose individual trees do. These results appear throughout the experiments, robust to changes in types of predictors created and in numbers assigned.”

The Waze problem looks similar. If you see congestion on your planned route, you’ll switch to an alternative – perhaps – for you also have to predict how many other people will switch too, and whether the initial congestion will stay or vanish if there are enough other users of similar apps.

Later work on El Farol found one game theoretic solution: individual agents adopt a mixed strategy, whereby each has a fixed probablility of choosing either El Farol or an alternative bar. Another is that after a period of learning, agents sort themselves into groups, those who always go and those who always stay home. But I don’t think these work for Waze-style congestion problems which are not repeated.

Indeed, real-time apps are surely creating more of these kinds of co-ordinated decision problems.

Brian Arthur’s book [amazon_link id=”B00SLUR9HI” target=”_blank” ]Complexity and the Economy [/amazon_link]is a nice introduction to his work.

[amazon_image id=”B00SLUR9HI” link=”true” target=”_blank” size=”medium” ]Complexity and the Economy: Written by W. Brian Arthur, 2014 Edition, Publisher: Oxford University Press, USA [Hardcover][/amazon_image]

Markets in stories

Yesterday I attended a very interesting conference, on Non-Equilibrium Social Science, which stands for making economics more realistic and interesting – looking at the economy in terms of non-linear dynamic systems. As ever at a good conference there were some great opportunities for conversation – and book recommendations. I came away with Marion Fourcade’s [amazon_link id=”0691148031″ target=”_blank” ]Economists and Societies[/amazon_link], and W.E.G.Salter’s [amazon_link id=”0521095689″ target=”_blank” ]Productivity and Technical Change[/amazon_link], and William Hazlitt’s 1805 [amazon_link id=”1151650102″ target=”_blank” ]Essay on the Principles of Human Action[/amazon_link].

[amazon_image id=”0691148031″ link=”true” target=”_blank” size=”medium” ]Economists and Societies: Discipline and Profession in the United States, Britain, and France, 1890s to 1990s (Princeton Studies in Cultural Sociology)[/amazon_image]   [amazon_image id=”1144113512″ link=”true” target=”_blank” size=”medium” ]An Essay On the Principles of Human Action: Being an Argument in Favour of the Natural Disinterestedness of the Human Mind..[/amazon_image]

Also, after hearing him speak, David Tuckett’s [amazon_link id=”0230299857″ target=”_blank” ]Minding the Market[/amazon_link]. His argument was, in a nutshell: “Financial markets are markets in stories.” He classed as standard models both the rational choice and ‘behavioural’ approaches to decision making, because both assume there is an objective reality about which an optimum can in principle be known or calculated. Instead, Prof Tuckett argued, the decision problem is one of making any sense at all of how to act now on the basis of information now about an ontologically uncertain future. His answer is that we act on the basis of ‘conviction narratives’ which are collective interpretations of what today’s ‘facts’ (all based on our sense perceptions) mean for tomorrow. He described some empirical work looking at the stories that run in financial markets, as extracted from unstructured texts such as newswire reports, brokers’ notes and Bank of England reports.

[amazon_image id=”0230299857″ link=”true” target=”_blank” size=”medium” ]Minding the Markets: An Emotional Finance View of Financial Instability[/amazon_image]

This is intriguing. It certainly chimes with my unease (as in my Tanner and Pro Bono lectures) about the way economists talk – especially in the context of policy – as if they are omniscient outsiders, not part of what they are analysing. My question – which Prof Tuckett agreed is still-unexplored territory – is how reality interacts with our narratives, sometimes changing them. I’d like to have followed up by asking if the idea is a [amazon_link id=”0226458121″ target=”_blank” ]Kuhnian[/amazon_link] paradigm shift, but in the domain of financial markets, or economic life in general, rather than scientific exploration.

Another conference highlight was Bridget Rosewell on the inadequacy of our standard cost-benefit approach for deciding whether or not to go ahead with transport infrastructure projects, inadequate because they use a comparative static, equilibrium framework for something that is bound to change the dynamics of the economy. She gave many examples of projects that would never have passed a modern cost-benefit assessment, from Bazalgette’s London sewers to the Jubilee Line Extension and – as she is scrupulous – an example of one that wasn’t making it in terms of her tale of self-fulfilling visions that deliver the benefits they subscribe. (This is the fantasy airport on Boris Island, which she supports.) Bridget touches on this infrastructure question in [amazon_link id=”1907994149″ target=”_blank” ]Reinventing London[/amazon_link].

[amazon_image id=”1907994149″ link=”true” target=”_blank” size=”medium” ]Reinventing London (Perspectives)[/amazon_image]

Economists and humanity

Peter Smith sent me his new book T[amazon_link id=”0957069707″ target=”_blank” ]he Reform of Economics: How the complex systems approach is building a realistic and humane alternative to laissez-faire[/amazon_link]. In a letter accompanying it, he said he has two motivations. One is to get economics out of the trap of over-simplifying so that models can use linear algebra and thus be made ‘tractable’. This is one of the things that makes complexity economics and agent-based modelling appealing; virtual economies run on a computer do not need to be solved algebraically.

[amazon_image id=”0957069707″ link=”true” target=”_blank” size=”medium” ]The Reform of Economics[/amazon_image]

The other aim is to make economic methodology something more like normal scientific methodology. Economic method consists of choosing some basic postulates and making deductions from them. The deductions can then be tested against data. Normal science involves both induction and deduction. Careful empirical observation will shape theory.

The book dates the choice of the purely deductive path to Lionel Robbins and his 1935 essay [amazon_link id=”B001037AGS” target=”_blank” ]The Nature and Significance of Economic Science[/amazon_link]. He defined economics as the science of constrained choice, which, “Not only excludes uncertainty, but it also excludes from the scope of economics both institutions and the medium-term evolution of economic systems.” This isolates economics from the institutional framework of the economy, and hence from what determines the availability of resources over time – it makes economics an inherently static subject.

Natural scientists do regard economics as bizarrely non-empirical – I’ve been in multi-disciplinary conferences about both macroeconomics and behavioural choice at which biologists exclaim about how rarely economists discuss data, for all that they might go away and test hypotheses. One of the joys of being on the Competition Commission for eight years was how profoundly evidence-based the process is, and hence a real insight for an economist used to generalising about how companies behave. There aren’t many business people who think about marginal cost curves and production functions.

[amazon_link id=”0957069707″ target=”_blank” ]The Reform of Economics[/amazon_link] is a game of two parts (not halves). It is mostly a critique of economic methodology but also has a useful introduction to agent based modelling. It ends on an upbeat note I very much like:

“Economics is becoming a much more interesting area in which to work and learn; and we have every hope that a more realistic and effective reformed science of economics will also be a more humane one. For, ultimately, economics is about the well-being of humanity.”

Resources on complexity economics

Following my posts last week on complexity and economics, Professor Leigh Tesfatsion of Iowa State University sent me this very useful website with links to loads of resources on the subject – including an introductory self-study course.

Prof Tesfatsion wrote to me that the complexity approach has real momentum but added: “In macroeconomics, however, bitter resistance has been encountered, particularly from those who have devoted themselves to mastering DSGE modeling.” However, there is also some work on agent-based macroeconomics.

One general book I spotted in this list, one I’ve not read, is Mark Buchanan’s [amazon_link id=”1408827379″ target=”_blank” ]Forecast: What Physics, Meteorology, and the Natural Sciences can Teach Us About Economics[/amazon_link]. Nate Silver’s [amazon_link id=”0141975652″ target=”_blank” ]The Signal and the Noise[/amazon_link] doesn’t feature agent-based modelling but does talk about the macroeconomy as a complex (non-linear multivariate dynamic) system.

[amazon_image id=”1408827379″ link=”true” target=”_blank” size=”medium” ]Forecast: What Physics, Meteorology, and the Natural Sciences Can Teach Us About Economics[/amazon_image]

Serendipity, complexity, and loneliness

No sooner (literally)  had I written about the complexity economics of the new book by David Colander and Roland Kupers, [amazon_link id=”0691152098″ target=”_blank” ]Complexity and the Art of Public Policy[/amazon_link], than (in one of the many instances of serendipity in life) another book  on complexity turned up in the post, courtesy of its author, Peter Smith. The book is [amazon_link id=”0957069707″ target=”_blank” ]The Reform of Economics: How the complex systems approach is building a realistic and humane alternative to laissez-faire[/amazon_link].

[amazon_image id=”0957069707″ link=”true” target=”_blank” size=”medium” ]The Reform of Economics[/amazon_image]

The book looks like it argues for a more realistic alternative to mainstream economics by actually developing it, using agent-based modelling. In a covering letter, Dr Smith says the intelligent agents: “learn by experience how to respond to market conditions. … They can engage in price exploration, and learn to manage their inventory, plant renewal and cashflow (or go bust), all starting from far-from-equilibrium states.” He also describes his research, and his search for a post-crisis renewal of economics as “a mite lonely.” I think it might be less lonely than he fears.