YES
Pete Lunn
NO
Tim Harford
Dear Tim
12th August 2008
These are exciting times to be an economist. The whirligig of international finance has come crunching to a halt. The British housing market inflated until we could hardly bear to watch, then popped in a destructive, sticky instant. The prices of food and oil are yo-yoing on speculative strings. And the textbooks still tell us that markets populated by rational, selfish, independent agents allocate resources efficiently.
Meanwhile, a revolution is under way in economic thought. Behavioural economics is no bell or whistle on the contraption of traditional economics; it is a big departure which will deliver a revolutionary new way of understanding the world. The founding assumptions of orthodox, neoclassical economics—that people can be thought of as rational, selfish and independent—are collapsing under the weight of empirical refutations.
Here is one example: the "ultimatum game," which typifies the story of behavioural economics with a curious yet simple experiment. As you know, in this two-player game, the "proposer" is given a sum (say £10) on condition that he or she offers a proportion to the "responder." If the responder accepts the offer, each player gets the amounts agreed. If he or she rejects it, both get nothing. Orthodox economics says players are selfish, and so predicts that the proposer will offer just a penny and the responder, preferring a penny to nothing, will accept. But this is not what happens. The most common offer is half the total sum, and offers of less than 30 per cent are almost always rejected. If the proposer's offer is seen as unfair, the responder will decline free money.
What do you make of this violation of orthodox theory? In my experience, most economists argue that the proposer is currying favour, or that the responder will turn down a few quid for the sake of revenge. But the results described above hold when players are anonymous and when the sum involved is up to three months' income. When offered unfair shares, we have a strong economic instinct to say: "you can keep it."
This experiment reveals an instinct that might apply to any transaction where one side proposes the price or wage. So we might expect to see large deviations of prices and wages from their supposed equilibrium between supply and demand. And in fact such deviations do occur. When consumers quote the maximum they will pay for something, their answer changes not only with how much they want it, but also with their perception of what it costs to provide. When bargain-hunting clothes shoppers discover the wages of those who do the stitching, it changes what they are willing to pay. When unions decide wages are too unfair, they are prepared to jeopardise the future of the whole enterprise. When people work in more profitable sectors or companies, they are paid more than the market wage. Our instinct for fair shares flouts the "law" of supply and demand.
Other classic experiments of behavioural economics are challenging the traditional assumptions too. We often overcome selfishness to reach beneficial common solutions, as good managers and motivators understand. As for rationality, behavioural economists have recorded hundreds of instances where our economic instincts run counter to the traditional definition, especially when risk or uncertainty are present. And the assumption that economic agents are independent ignores our strong instincts to seek economic alliances and to be influenced by the opinions of others about what things are worth—those bubbles again.
Orthodox economic models are not wrong as such, but rather sloppy, biased approximations of how our economy works. They present a cartoon characterisation of economic life, greatly exaggerating one side of our nature at the expense of others. Behavioural economics has started to paint a more realistic picture. Its progress is following the pattern of a scientific revolution. At first the findings seem to be anomalies and oddities, but on closer inspection they yield new principles and regularities. Economists trained in the conventional approach were initially resistant—many still are—but over time the more open-minded and, interestingly, younger economists are being enticed towards the new field; testing explanations, deriving implications.
Yours in anticipation
Pete
Dear Pete
13th August 2008
I'm disappointed. I had assumed, given your opening paragraph, that you were going to explain how behavioural economics might have predicted the credit crunch and the commodity boom, or offered a new way to regulate banks. That really would have been revolutionary. Instead, you offer the dear old ultimatum game. I was not surprised to read your lop-sided and unconvincing caricature of orthodox economics ("textbooks" are such a convenient straw man), but I was astonished to see you present a caricature of the subject you claim to be championing.
I am not sure that a "simple experiment" like the ultimatum game does typify behavioural economics. Such experiments have their place, but they also have serious weaknesses. Take this love of fairness, and its application to wages, which you emphasise. No doubt you are familiar with the laboratory work on how workers respond to wage offers. In one celebrated experiment, behavioural economists divided their subjects into "employers" and "workers." They discovered that when the "employers" paid unexpectedly generous "wages," the "workers" reciprocated by working unexpectedly hard.
It's a classic of the field. But the real world remains intractable. The economists John List and Uri Gneezy recently repeated the lab study in the field, advertising real jobs, hiring real workers and paying real hourly rates. They used a controlled trial to see what happened when workers were paid unexpectedly generous rates. And they discovered that the lab results were evanescent: after a couple of hours the gratitude evaporated and the workers slacked off, reverting to the rational self-interested behaviour described in those pesky textbooks. I would not advise personnel departments to rewrite salary scales on the basis of an effect that does not survive past lunch on day one.
Perhaps I am being unfair. Economics is a work in progress, and some of that progress is coming from behavioural research. Yet the idea that the very foundations of economics are being undermined is absurd. One reason for that is that our inner Mr Spock is closer to the surface than we tend to think. In the last few years, economists have found plenty of examples of rational behaviour in the unlikeliest territory. Consider sex, dating and marriage.
For example, American teenagers respond rationally to more restrictive abortion laws by having less, or safer sex. Whenever a particular US state adopts such a law that affects only teenagers, the prevalence of sexually transmitted diseases in the teenage population falls relative to the adult population.
An analysis of speed-dating shows that our tendency to propose dates is responsive to—there's no nice way to put this— "market conditions." Our romantic standards adjust to whether we think we're in a buyers' market or sellers' market.
The introduction of easier divorces in the US in the 1970s altered the bargaining power of women, even though it had little permanent impact on the divorce rate. Again, whenever a US state changed the law, behaviour changed: domestic violence fell by up to a third; spousal murders and suicides also fell.
In short, for every laboratory result showing that we are irrational, economists are discovering rational behaviour from the most unexpected people in the most unexpected situations. Despite all this, I agree that behavioural economics promises to offer more realistic descriptions of the way we act, even if the research is still at an early stage. But that would not demolish the achievements of orthodox economics, because the aim of economics is to understand complex systems, not to predict how a few students behave in a lab.
More realistic models of human behaviour may give us more insight into how economies work, but they also may not. If I was having this debate with Prospect's science columnist Philip Ball, he would be arguing that economists do not have a good understanding of complex systems or the dynamics of networks. He would trumpet the potential of techniques developed by physicists and mathematical sociologists. And these techniques depend on even simpler, less realistic models of human behaviour than economists typically use. It is naive to think that psychological realism is the acid test of economic analysis.
Economists must strike a balance between more sophisticated modelling of individual behaviour and more sophisticated modelling of economic systems. While both psychology and physics have much to offer, in many cases the mainstream compromise continues to work.
I don't deny that behavioural economics is an exciting branch of economics, but it is just one among many.
Rationally yours
Tim
14th August 2008
You were disappointed, I'm astonished. Textbooks are a straw man? To me they are like a vice on the mind of economists, who get tetchy when you present results that reveal the behaviour of real people to be not as those textbooks assume.
When the "dear old ultimatum game"—which most economists I meet have never heard of—is finally given its place in those texts, the revolution I assert is under way will have planted its flag. We will, at last, train economists to understand that when people trade, selfishness is countered by other motives, meaning that prices and wages often fail to match supply and demand—a point which, interestingly, you chose not to dispute.
You raise the issue of drawing conclusions from laboratory experiments. Now, we could trade interpretations of academic papers like List and Gneezy's on workplace motivation until our readers head for other sections. Instead, let's trade evidence on workplace motivation as a whole, from both field and lab.
Here's my take. For decades, orthodox economists have published studies addressing whether workers are primarily motivated by personal gain, and have managed to tie themselves into amusingly contorted knots trying to explain their findings. Most studies show that linking productivity to pay doesn't increase it, that even if it did managers are unwilling to enforce the schemes, and that the managers are probably right because such schemes undermine teamwork.
In passing, these economists have noted that there is one kind of pay scheme that tends to increase individual productivity: profit-sharing. But why would one worker increase their effort when the extra revenue they generate will be split with everyone else in the company?
This is like a public good game that behavioural economists get groups of people to play in the lab. Everyone can choose to contribute some amount to the "pot," which is then doubled and split between all the players. If everyone contributes, they all end up better off. But from a selfish point of view, it is better not to contribute and to free-ride on the generosity of others. What happens? Most people contribute. Events in the lab match events in the workplace, neither of which turns out as textbook economics predicts.
We should of course be careful extrapolating from lab experiments. But where strong and consistent economic instincts are found in the lab, their influence is usually found outside in the real economy too.
Speaking of which, it is curious that some economists appear to have abandoned trying to explain how that economy works and taken up trying to explain anything from teenage sex to cheating schoolteachers. Of course people respond rationally to incentives to some degree in all walks of life. This hardly constitutes evidence that rational choice theory—the idea that people take rational decisions in their own self-interest—is the best way to understand how the economy works.
Lastly, I agree that it would be naive to think that psychological realism is the acid test of economic analysis. Economic models have to simplify human nature in order to shed light on what happens when people come together to do business. The question is: what simplification best describes us? What are our most fundamental economic instincts?
In my view, the answer is unlikely to come from mathematically convenient assumptions cooked up from the comfort of a senior common room armchair. It will come instead from observing and understanding people's real economic behaviour, which is why behavioural economics has begun to revolutionise our subject.
Instinctively yours
Pete
Dear Pete
14th August 2008
I'm sorry to hear that you're surrounded by a tribe of economic neanderthals. Keep up the good work in bringing them enlightenment. But I don't need to judge economic orthodoxy by your descriptions of your acquaintances; I can simply look at what is published in the top mainstream journals. Behavioural economics is popular there, and its main findings widely known. But surely that is not a "revolution." A revolution would sweep away much of the existing consensus; that has not happened and I doubt that it will.
There are three reasons for this. The first is that your question, "What simplification best describes us?" is the wrong one. You have confused economics with psychology. As I wrote in my first letter, one of the most fertile areas in economics is the modelling of complexity. Three examples: the evolution of firms; the development of national economic clusters such as South Korea's memory chip industry; and the spread of social norms like honesty, obesity or smoking. These new complexity models are producing brilliant new results despite riding roughshod over psychological insight.
Second, even when we want to incorporate behavioural insights, it's not always clear how to do so, because they remain disparate. I accept your claim that people have a sense of fairness and reciprocity. But what then? You breezily claim that "most studies" (in the lab? of real companies?) show that workers are motivated not by individual incentives but by a share in a group's profit. Tell me more: which studies? Under which circumstances? You are too vague: arguing with you feels like trying to arm-wrestle a hologram.
I could reel off citations showing the reverse: that individual performance pay schemes do boost performance. I am thinking of Ed Lazear's investigation of a US car repair company, studies of farm workers by Oriana Bandiera and her colleagues, and Robert Drago and Gerald Garvey's analysis of Australian firms. I realise that it is tedious to be so specific, but your hand-waving is getting us nowhere.
The conclusion I draw is that behavioural economists have found evidence—sometimes dubious, sometimes compelling—of a patchwork of exceptions to a general rule. They may one day be able to develop these exceptions into a better theory, one that is simple enough to incorporate into economic dynamics yet realistic enough to beat the rational choice model. But so far, it's case-by-case stuff.
The third reason there has been no behavioural revolution is that the orthodox, rational-choice approach continues to work. Take a step back and look at the big picture. According to the laboratory experiments on public goods you describe, there is no such thing as the free-rider problem. If only that were true. It would mean that there was no climate change problem, because people would voluntarily restrict their carbon emissions to preserve the planet for strangers and the children of strangers. It would mean that fish stocks were healthy because fishing crews realised they were dealing with a common resource. London's congestion charge would be counterproductive, because people do not respond to individual incentives: drivers would have willingly left their cars at home in order to leave the roads congestion-free for others.
No matter how many experiments you allude to, the discomfiting rational self-interested model explains our environmental predicament perfectly. It is also the inspiration for solutions such as a carbon tax or a cap-and-trade scheme. Don't take my word for it; such policies have been recommended by Richard Thaler, the world's most respected behavioural economist.
Instead of focusing on the flimsiest areas of behavioural economics, backed up by airy generalisations, why don't we discuss areas where it has made important contributions, such as understanding how people choose and use financial products like pensions and credit cards? It would be a shame if readers concluded that behavioural economics has nothing to offer.
Your logical friend
Tim
Dear Tim
15th August 2008
Arguing with me is like arm-wrestling a hologram because you are struggling to grasp the unorthodox. For instance, you refer to "what is published in top journals," but those journals are not always the best guide: three of them rejected George Akerlof's seminal 1970 paper "The Market for Lemons," which launched information economics and changed the discipline forever.
You make three substantive points. First, you refer to the success of complex systems models. But this does not undermine the behavioural approach. The two often reinforce each other. (Part of my book explains why your first example, firm evolution, is more behaviourally realistic than standard theory.)
Second, I was avoiding "paper tennis," but you insist. The academic Canice Prendergast reviewed over 100 studies, including those you cite, and found evidence of effective performance pay schemes only in "simple" jobs where performance is easily measured, not in most jobs. Then, expressing puzzlement, he concludes that profit-sharing schemes work. Or consider a recent paper by Sebastian Kube and colleagues, which found that fairness does affect productivity—suggesting that your interpretation of Gneezy and List is askew.
Third, I didn't claim there is no free-rider problem—of course there is. Behavioural economics has revealed our instincts when faced with it, and the importance of communication, group identity, punishment mechanisms and reducing uncertainty. Your description of environmental problems, which wrongly assumes instinctive selfishness, precludes solutions based on this evidence of how to promote instinctive co-operation.
Lastly, this recurring "economics can explain everything" claim is an ideological fiction used to sell books. It also exemplifies "confirmation bias"—looking only for evidence that fits the thesis. Statistically significant changes in behaviour across large samples show that some people respond to changed incentives rationally, to some degree. It does not imply that rational choice theory is the best way to understand human behaviour, any more than statistically significant evidence for wind resistance tells you how vehicles are propelled. The logic is flawed.
Great leaps in scientific understanding are mostly driven by refutation, not confirmation. The irony is that while we economists jump on the Freakonomics bandwagon, trumpeting theories and exaggerating their power, our core theoretical propositions are being refuted by expanding bodies of empirical evidence.
Yours
Pete
Dear Pete
15th August 2008
My frustration was not at your unorthodox conclusions but at your tendency to generalise without visible means of support. Thanks for correcting the omission; I'm looking forward to reading the Prendergast survey. It sounds as though he finds that when pay can be linked to performance, it motivates, but when it can't, it doesn't. Is that really so unorthodox?
As for the journals, I am puzzled. I pointed to them as a sign of how behavioural economics is being embraced; you respond by suggesting that they might be shutting out behavioural economics. Well, they might. But they aren't. You are trying to characterise behavioural economists as intellectual outcasts, unknown by the profession, denied access to the journals, waiting for the revolution. You know that's misleading. Look at the three top figures in behavioural economics. One, Daniel Kahneman, has a Nobel prize. One, Matthew Rabin, has the even rarer John Bates Clark medal. The third, Richard Thaler, has had a regular column in the Journal of Economic Perspectives, one of the most widely read academic journals, since the mid-1980s. This is hardly a conspiracy to suppress behavioural economics.
You began by saying that it is an exciting time to be an economist. I agree—not because economics is about to be replaced by psychology, but because of the big questions we are now struggling to answer. What are the fundamental causes of poverty? Are there solutions? How do ideas, habits and information spread across social networks? What makes innovation happen? Can governments help? Can anything improve the awful record of economic forecasting? How do economic clusters develop? What kills them off? How does competition work in the face of fallible consumers?
Behavioural economics is already making the running on the subject of consumer decision-making, but in other areas it has little to offer. That is no surprise, because there is so much more to economics than a theory of individual decisions.
Economics is stronger thanks to the behavioural economists, that much is clear. Yet the picture you paint—of a great intellectual tide, pent up and ready to wash away most of what economists hold true—is not one I recognise.
Yours
Tim
Click here to discuss this piece at First Drafts, Prospect's blog
