In the rough-and-tumble world of science, disputes are usually settled in time, as a convergence of evidence accumulates in favor of one hypothesis over another. Until now.

On April 10 economist John R. Lott, Jr., formerly of the American Enterprise Institute, filed a defamation lawsuit against economist Steven D. Levitt of the University of Chicago and HarperCollins, the publisher of Levitt’s 2005 book, Freakonomics. At issue is what Levitt meant when he wrote that scholars could not “replicate” Lott’s results …

That’s from Michael Shermer’s Skeptic column in Scientific American. Shermer is “executive director of the Skeptics Society, bold debunkers of all things supernatural”, according to Salon which has a long interview with him [free, if you are willing to watch an ad]. I don’t know if artificial intelligence would be considered ’supernatural’, but here is an interesting article in the Skeptic magazine (flagship of the Skeptics Society) debunking the lofty claims made by AI enthusiasts.

… to post excerpts. Here are the links anyway:

• FT’s review of Yochai Benkler’s Wealth of Networks: How Social Production Transforms Markets and Freedom”
• Tyler Cowen’s latest NYTimes column on gastronomical economics.
• Daniel Gross on why businesspeople love to quote Chinese proverbs.
• Tim Harford on why some people cheat, and others don’t.
• Stuart Jeffries on why happiness is overrated.

After writing this post about economics, physics and econophysics, I was poking around the web, looking for Philip Ball’s articles. Ball is the author of the piece that I linked to in my post, and has written quite enthusiastically about “sociophysics” which seems, to me, to be mostly simulations in which independent entities (particles, people, institutions) act and react according to specific rules. From statistical physics simulations of interacting particles, we know that complex behaviour could emerge even with simple interactions among the particles, and I guess the hope in sociophysics is to show a similar correspondence between simple interactions among entities (‘agents’ seems to be the preferred term in sociophysics) and (emergence of) complex behaviour in the aggregate.

Philip Ball has a huge footprint on the web, a testimony to his prolific output, not only as a regular columnist for the Nature group of publications, but also as an author of quite a few books. Check out his website. One of his recent books, Critical Mass: How One Thing Leads to Another has specifically been about sociophysics. Some of the ideas appeared earlier in the form of a short article with a catchy title Physics of Institutions (pdf); see also this rather nice popular science piece titled Utopia Theory in PhysicsWeb.

Here are some of the reviews of this book: Bruce Edmonds, James Buchan for the Guardian, Steven Strogatz for Nature, and Tamás Vicsek for PhysicsWeb. The ‘Reviews’ section of Ball’s website has links to more of them.

Let me quote from Bruce Edmonds’ review:

… It is, in its way, the first “popular science” book covering a substantial section of social simulation, and talks about many of the main figures up to about 1990 (it does cover later work but not so comprehensively, which is understandable). Thus the work of Thomas Schelling, Ilya Prigogine, Brian Arthur, Alan Kirman, Robert Axtell, Joshua Epstein, Robert Axelrod, Paul Omerod, Martin Nowak, Per Bak, Duncan Watts, are all discussed.

In all of this the book is quite careful as to matters of fact – in detail all its statements are cautiously worded and filled with subtle caveats. However its broad message is very different, implying that abstract physics-style models have been successful at identifying some general laws and tendencies in social phenomena. It does this in two ways: firstly, by slipping between statements about the behaviour of the models and statements about the target social phenomena, so that it is able to make definite pronouncements and establish the success and relevance of its approach; and secondly, by implying that it is as well-validated as any established physics model but, in fact, only establishing that the models can be used as sophisticated analogies – ways of thinking about social phenomena. The book particularly makes play of analogies with the phase transitions observed in fluids since this was the author’s area of expertise.

This book is by no means unique in making these kinds of conflation – they are rife within the world of social simulation. The culture of physics is a complex of different attitudes, norms, procedures, tools, bodies of knowledge and social structures that are extremely effective at producing useful knowledge in some domains – it is not for nothing that physists have gained status within our society. However when this culture is transported into new domains, such as that of modelling social phenomena, the culture does not travel uniformly. Thus we have seen (and Critical Mass documents) an influx of simple, physics-style simulation models into sociology but they have arrived without the usual physists’ insistence that models predict unseen data. It is part of the culture of physics to aspire to the simplest possible model of phenomena but a model which only acted as a sort of vague analogy with respect to its phenomena would get short shrift in traditional physics domains. Yet frequently one reads social simulation work which takes the form of physics-style models and yet uses only vague, hand-waving justifications to justify its relevance (and, at best, a rough fitting of known, aggregate data). Models need to be constrained by the subject matter they are supposed to be about – there are two main ways of doing this: by ensuring the model is designed to behave as we know it should do (typically the parts of the model); and by checking the resulting behaviour against corresponding observed behaviour (often in aggregate). Sociophysics models tend to avoid either: they impose over-simple behaviour onto the design and don’t validate strongly against unseen data. Thus whilst such models may have interesting behaviour there is little reason to suppose that they do in fact represent observed social behaviour.

A point Edmonds makes is this:

[C]omplex behaviour can result from the interaction of lots of simple parts. This is now well established, but the implied corollary that the complexity we observe is a result of lots of simple interactions (or that it is useful to model this in this way) does not, of course, follow. Grounds for hope does not make it a reality.

This seems to be an intensely difficult ‘inverse’ problem, no? A related problem, which seems to be common to many ‘emergence‘ phenomena is the following: suppose you rig up a model with a certain set of rules (for interactions among the agents). And suppose that this model exhibits some complex behaviour. You are certainly within your rights to feel satisfied. However, how can we be sure that this is the only set of interaction rules that will lead to this ‘complex’ behaviour? If there are two (or more) sets of rules that give rise to (broadly) the same complex behaviour in the aggregate, which one should we choose? Even then, how can we be sure that that is the one that governs the real interactions among the agents?

An interesting paper in Science (a publicly available summary is here; link via Brain Ethics) shows that human beings’ propensity to punish (unfair acts by others) is correlated with their altruism. This finding is based on a pretty large scale study involving populations in no less than 15 different societies or tribes. The implication is that these two cultural traits co-evolved. Here’s a key quote:

A hallmark of humanity is that people help other people–not just relatives and friends but even complete strangers. Such altruism, which goes beyond the mere exchange of favors and forms the scaffolding of large-scale cooperation in human societies, has long been an evolutionary mystery. On page 1767, anthropologist Joseph Henrich of Emory University in Atlanta, Georgia, and his colleagues take a crack at solving the puzzle, concluding that such helpful behavior may have arisen as a result of punishment.

Reporting on experiments they conducted in 15 different societies on five continents, the researchers argue that altruism evolved hand in hand with a willingness to punish selfish behavior. Their results lend support to models of gene-culture coevolution that propose that cultural norms such as the punishment of unfair actions drive the selection of genes favoring altruism.

What was interesting (to me, at least) was the use of three fairly simple prototype games that allowed the researchers to assess the participants’ inclination towards “costly punishment” and “altruism”. For assessing punishment, they used the Ultimatum Game and Third Party Punishment Game; for assessing altruism, they used the Dictator Game. In case you are not able to access the paper from the Science website, Brain Ethics has a description of the three games (with links).

Over at Reason Online, Ronald Bailey has an article explaining this ‘punishment-and-altruism’ research, and this is his concluding paragraph:

The results are intriguing. It turns out that the societies in which the player ones in the dictator game were willing to give more to the player twos are also the societies in which people were more willing to punish less generous players in the other two games. In other words, societies that punished strongly were also the most likely to have strong altruistic impulses. The moral of the story is that if you want to live in a world of caring generous cooperative people, make sure that you thoroughly thrash all the greedy, chiseling scoundrels you come across. It may cost you, but the world will be a better place. [bold emphasis added]

An academic paper by Marianne Bertrand, Simeon Djankov, Rema Hanna, Sendhil Mullainathan examines this issue with drivers in New Delhi. Here’s the abstract:

We follow 822 applicants through the process of obtaining a driver’s license in New Delhi, India. To understand how the bureaucracy responds to individual and social needs, participants were randomly assigned to one of three groups: bonus, lesson, and comparison groups. Participants in the bonus group were offered a financial reward if they could obtain their license fast; participants in the lesson group were offered free driving lessons. To gauge driving skills, we performed a surprise driving test after participants had obtained their licenses. Several interesting facts regarding corruption emerge. First, the bureaucracy responds to individual needs. Those who want their license faster (e.g. the bonus group), get it 40% faster and at a 20% higher rate. Second, the bureaucracy is insensitive to social needs. The bonus group does not learn to drive safely in order to obtain their license: in fact, 69% of them were rated as “failures” on the independent driving test. Those in the lesson group, despite superior driving skills, are only slightly more likely to obtain a license than the comparison group and far less likely (by 29 percentage points) than the bonus group. Detailed surveys allow us to document the mechanisms of corruption. We find that bureaucrats arbitrarily fail drivers at a high rate during the driving exam, irrespective of their ability to drive. To overcome this, individuals pay informal “agents” to bribe the bureaucrat and avoid taking the exam altogether. An audit study of agents further highlights the insensitivity of agents’ pricing to driving skills. Together, these results suggest that bureaucrats raise red tape to extract bribes and that this corruption undermines the very purpose of regulation.

There is some commentary by Brad DeLong and Alex Tabarrok. In particular, Mark Thoma has quite a few links in his discussion of the question: “Does corruption improve economic efficiency?”.

Here’s a primer:

Let’s start with how economists reach our conclusions. Non-economists might be forgiven for presuming that we construct our arguments in the same way that they do: apply their preferred mixture of values and interests in order to decide whether or not they like a policy, and then assemble arguments to support that position. This might explain why many would view economists as opponents: by the simple act of disagreeing, aren’t economists making it clear that we don’t share the same values?

Except that’s not how we work. Our starting point is always a model: a stylized representation of how the economy works. Once we’re satisfied that we have a model that incorporates the main features of interest — this step necessarily involves a certain amount of subjective judgment — we compare what the model would predict if the policy were in place with what would happen without it. The difference between the two predictions is the effect of the policy.

After describing the problems that arise from this fundamental misunderstanding, the author (Stephen Gordon of the University of Laval at Quebec City) goes on to offer some suggestions about what can be done:

Clearly, economists can make a more concerted effort to explain to non-specialists what it is they are saying, and why. This isn’t a simple task — economics is a difficult and technical subject — and it’s made more complicated by the fact that there are any number of commentators who have built their careers on misunderstanding and misrepresenting what economists have to say.

But it would be easier if progressives made an effort to set aside their distrust of economists and actually listen to what we are trying to say. Yes, you may be forced to re-examine some long-held opinions, but is that really a bad thing? And you may be pleasantly surprised to learn that we too are preoccupied with finding solutions to the problems of poverty and inequality.

Ttwo Princeton professors, economist Alan B. Krueger and psychologist and Nobel laureate Daniel Kahneman, in collaboration with three others from other universities (psychologists David Schkade of the University of California-San Diego, Norbert Schwarz of the University of Michigan and Arthur Stone of the State University of New York-Stony Brook) are reporting something quite interesting:

While most people believe that having more income would make them happier, Princeton University researchers have found that the link is greatly exaggerated and mostly an illusion.

People surveyed about their own happiness and that of others with varying incomes tended to overstate the impact of income on well-being, according to a new study. Although income is widely assumed to be a good measure of well-being, the researchers found that its role is less significant than predicted and that people with higher incomes do not necessarily spend more time in more enjoyable ways.

… The new findings build on their efforts to develop alternative methods of gauging the well-being of individuals and of society. The new measures are based on people’s ratings of their actual experiences, instead of a judgment of their lives as a whole.

The study is being published in Science, in the issue dated 30 June 2006. Here’s the abstract:

The belief that high income is associated with good mood is widespread but mostly illusory. People with above-average income are relatively satisfied with their lives but are barely happier than others in moment-to-moment experience, tend to be more tense, and do not spend more time in particularly enjoyable activities. Moreover, the effect of income on life satisfaction seems to be transient. We argue that people exaggerate the contribution of income to happiness because they focus, in part, on conventional achievements when evaluating their life or the lives of others.

Jon Wiener, professor of history at the University of California at Irvine, recounts the story of the investigation by a faculty committee into allegations of academic misconduct by Ward Churchill, “Native American activist and professor of ethnic studies at [the University of Colorado at Boulder]“. It appears to be a balanced story, and there are lots of details there that I was not aware of (I wasn’t following this case closely enough, I suppose).

Janet Stemwedel, who’s “working the ethics beat at ScienceBlogs“, offers a crib-sheet on the important ethical questions in this case. She concludes thus:

It would have been nice if the University of Colorado had investigated the earlier allegations of academic misconduct. It would have been swell if the careful investigation that [actually] happened hadn’t been precipitated by a politically motivated firestorm. But given the facts in evidence, he has to go.

For a long time, physicists have had a reputation for boldly venturing into other disciplines. Indeed, in a recent Physics Today article recounting the history of physics since 1931, Spencer Weart specifically mentions the rise of ‘hyphenated physics’ (bio-physics, geo-physics, etc) during this period as a key development.

The natives of the other disciplines, of course, would grumble because they felt that many of these wandering physicists were promiscuous (with no long term commitment to their field) and, more importantly, arrogant. I remember a wanderer saying several years ago, “You know, these metallurgists know a lot of stuff about X. I don’t know how they know so much, but they just do!” Among the natives, the joke is that these promiscuous physicists were just looking for interesting problems, because there weren’t any in physics. I suppose all this is a part of a healthy disdain for other disciplines that scientists imbibe and develop.

I am reminded of all this by this paragraph, quoted in Peter Klein’s post (which was triggered by an earlier post):

Economists are extending the range of their studies to include all of the social sciences. . . . What is the reason why this is happening? One completely satisfying explanation . . . would be that economists have by now solved all of the major problems posed by the economic system, and, therefore, rather than become unemployed or be forced to deal with the trivial problems which remain to be solved, have decided to employ their obviously considerable talents in achieving a similar success in the other social sciences. However, it is not possible to examine any area of economics with which I have familiarity without finding major puzzles for which we have no agreed solutions, or, indeed, questions to which we have no answers at all. The reason for this movement of economists into neighbouring fields is certainly not that we have solved the problems of the economic system; it would perhaps be more plausible to argue that economists are looking for fields in which they can have some success. [from Ronald Coase's 1978 paper titled "Economics and Contiguous Disciplines".

Just replace 'economics' and 'social sciences' with 'physics' and 'natural sciences', respectively, and you have a perfect analogy!

[Peeter Klein's posts also discuss and critique the 'freakonomics' kind of incursions into other fields; do read them.]

* * *

Last year, the New York Times proclaimed ‘econophysics’ as one of the most noteworthy ideas of the year. Given the reputation of physics and economics in their respective domains (natural and social sciences), econophysics sounds like a marriage between two domineering individuals. Has it been a marriage filled with joy and peace? Hardly!

In a recent article in Nature (subscription required), Philip Ball (author of this survey article on interating agent models in sociology) describes the scene rather well. Here’s how the article opens:

For the past two decades, some physicists have been trying to apply their ideas and tools to an area that seems a long way from traditional physics. They are exploring the notion that there might be a kind of physics of the economy — an ‘econophysics’, as it has been dubbed1. Last year, some of these econophysicists even went as far as to suggest that economics might be “the next physical science”.

But now this unlikely marriage is showing signs of turning sour. Even those economists who at first welcomed econophysics are starting to wonder whether it is ever going to deliver on its initial promise. Early successes in modelling financial markets have not led to insights elsewhere, some complain. Matters came to a head at the Econophysics Colloquium, held at the Australian National University in Canberra last November. A group of economists attending the meeting were so dismayed with what they saw many physicists doing that they penned a forthcoming paper entitled ‘Worrying trends in econophysics’.

To me, this paragraph is telling:

So why have some of these physics-friendly economists become fed up? Although Ormerod and colleagues are highly critical of mainstream economic theory, they point out that “economics is not at all an empty box.” The Canberra critique accuses econophysicists of ignoring the existing literature — a charge also levelled at physicists when they began to dabble seriously in biology.

* * *

Over a year ago, I covered a different kind of interdisciplinary war: the one between sociologists and physicists about the theory of social networks.

Roland G. Fryer and Steven D. Levitt (2006): Testing for racial differences in the mental ability of young children. Here is the abstract:

On tests of intelligence, Blacks systematically score worse than Whites, whereas Asians frequently outperform Whites. Some have argued that genetic differences across races account for the gap. Using a newly available nationally representative data set that includes a test of mental function for children aged eight to twelve months, we find only minor racial differences in test outcomes (0.06 standard deviation units in the raw data) between Blacks and Whites that disappear with the inclusion of a limited set of controls. The only statistically significant racial difference is that Asian children score slightly worse than those of other races. To the extent that there are any genetically-driven racial differences in intelligence, these gaps must either emerge after the age of one, or operate along dimensions not captured by this early test of mental cognition.

And, here are two interesting paragraphs from the discussion section:

The debate over racial differences in intelligence is among the most divisive in the social sciences. Utilizing a newly available, nationally representative data set with measures of mental function among children before their first birthday, we find little evidence of systematic racial differences. Some substantively small, but statistically significant differences are present in the raw data. Including controls for age, socio-economic status, home environment and prenatal environment largely erase these small differences. A simple calibration exercise suggests that many of the basic facts in the data can be generated from a model in which there are small mean differences in intelligence across races, but large environmental differences across races that become increasingly important as children age. [...]

Although damaging to the hypothesis that genetic differences are at the root of racial gaps in intelligence, the results of our analysis do not preclude a possible role for a genetic contribution to racial differences in intelligence for a number of reasons. First, one could reasonably argue that the control variables we include in the regression analysis are themselves partly genetically determined. By controlling for factors such as socio-economic status and birth weight (which systematically differ across races), we may indirectly be parsing out important channels through which genetics are operating. The fact that the raw differences in test performance across races are so small, however, makes this argument largely moot.

* * *

Hat tip: Brad DeLong.