A description of trade among Australian hunter-gatherers from David Graeber:
In the 1940s, an anthropologist, Ronald Berndt, described one dzamalag ritual, where one group in possession of imported cloth swapped their wares with another, noted for the manufacture of serrated spears. Here too it begins as strangers, after initial negotiations, are invited to the hosts’ camp, and the men begin singing and dancing, in this case accompanied by a didjeridu. Women from the hosts’ side then come, pick out one of the men, give him a piece of cloth, and then start punching him and pulling off his clothes, finally dragging him off to the surrounding bush to have sex, while he feigns reluctance, whereon the man gives her a small gift of beads or tobacco. Gradually, all the women select partners, their husbands urging them on, whereupon the women from the other side start the process in reverse, re-obtaining many of the beads and tobacco obtained by their own husbands. The entire ceremony culminates as the visitors’ men-folk perform a coordinated dance, pretending to threaten their hosts with the spears, but finally, instead, handing the spears over to the hosts’ womenfolk, declaring: “We do not need to spear you, since we already have!”
This prompted a comment from Daniel Davies that it seems a lot of work to organize an orgy every time I want to buy a blanket.
Some people quite enjoy haggling, some like me do not, but it’s always tiring and time-consuming. A Moroccan craftsman must spend an hour a day haggling with customers over his wares, and with his suppliers over their inputs.
Economists have long been fascinated by the process of market negotiation. They even see it as a paradigm for human interaction. Adam Smith defended self-interested arm’s length trade as a more solid economic base for society than unreliable benevolence. Léon Walras upped the stakes with his powerful idea of a general equilibrium: but to construct it, he needed to spin a fairy tale of an autarkic village market, to which peasants bring their produce and discuss prices until everybody is in subjective equilibrium, when the bell strikes and all the contracts are consummated. It’s too bad that no such market has ever been seen. And that counts as realistic by the standards of later general equilibrium models, with perfect information and zero transactions costs across whole industrial economies.
Actual industrial economies have in fact largely abandoned these face-to-face, open-outcry, haggling markets. This was an inevitable result of indutrialisation. According to Wikipedia (sourced to Emma Griffin):
A worker spinning cotton at a hand-powered spinning wheel in the eighteenth century would take more than 50,000 hours to spin 100 lb of cotton; by the 1790s, the same quantity could be spun in 300 hours by mule, and with a self-acting mule it could be spun by one worker in just 135 hours.
(The fine fibres of cotton must be very difficult to spin by hand; handspun wool output must have been much higher). A week’s output – say 60 hours – of one independent worker in 1750 would then have been two ounces of cotton thread: a small bag or spool he or she could haggle over with a weaver or seamstress. In 1800, the owner of a small state-of-the-art spinning mill with 100 workers would be producing two tons of cotton thread a week. He could not possibly sell this in the same individualised way. The practicable pricing systems became: long-term supply contracts; a fixed sales tariff, updated from time to time; an organised central exchange aggregating many buyers and sellers and trading standard spot and futures contracts; and a centralised auction, as with Dutch flowers.
The centralised systems have large setup costs and are only feasible for highly standardised products. You could not for instance sensibly have one for the huge variety of printed textiles. By the time products reach consumers, fixed-offer pricing is the only game in town, with a very few exceptions: real estate, second-hand cars, luxury hotel rooms (but not restaurants). Even most financial markets work on fixed offers, adjusted rapidly. The competitive markets that keep policymakers and buyers awake at night are not those for tires and bolts and shares and breakfast cereals, but precisely the negotiated ones for real estate, second-hand cars, and complex financial products.
We are so used to fixed-offer pricing as the norm that we forget it is quite recent in human history, and don’t notice that it takes us ever further away from Walras’ village general equilibrium and the real-life haggling-based Moroccan souk. When the fixed offers are set wrong, stocks run out or stay on the shelves unsold. Tut tut. Paretian inefficiency!
But so what? Time is money, as they keep reminding us. Negotiating costs are real and large. As we all get richer, it becomes less worthwhile to spend time searching out the lowest price. You and I will shop around for a €300 smartphone, but not for a €3 T-shirt as long as it fits. The Moroccan souk merchant is well aware of this propensity and uses it to extract as much of your consumer surplus as he can. The souk does not yield Walras’ single market-clearing prices, rather a cloud of individual transactions between the buyers’ and the sellers’ reserve prices, scattered round the notional equilibrium by the haggling contests.
Real economies operate of necessity in disequilibrium, as I’ve proposed here before. Actually feasible price-setting systems cannot achieve general equilibrium, and will vary from it in different ways. So even if you take Paretian efficiency as your sole criterion, it’s not self-evident that modern bureaucratic, fixed-offer capitalism is further away from it than the pre-industrial haggling souk. Supermarkets leave around 10% for fruit and vegetables unsold (here, page 10). The profitable French tiremaker Michelin typically ends up with a more surprising 20% of dead stock (here, page 2). I suppose this includes advances to retailers for the replacement market, where they make all their money. Does anybody have comparable numbers for pre-industrial markets? I’d be very surprised if they were lower.
I’m well aware, thank you, that Soviet central planning was much, much more wasteful. But we should be sceptical of the the (Hayek?) story that this had anything to do with a failure to solve the coordination problem and compute a feasible general equilibrium. On paper the material balances arrived at by Gosplan, through trial and error more than computation, were more or less OK. What doomed the centrally planned economy was the failure to create (outside the defence sector) incentives for quality and innovation like product liability and sale-or-return, the absence of a hard budget constraint on firms enforcing Say’s Law and efficiency, and systematic excess demand, the whole ensuring long queues for dud televisions. GE and GM are large command economies too, and they are not run by computed optimisation either. What keeps them working is that the relations between their component factories fully satisfy my conditions. The car assembly plant will return dud gearboxes to the gearbox factory, which will deduct the returns from its sales, and its imputed profits and the manager’s bonus go down. When GM’s overall outgoings exceeded its sales to customers, it was forced into bankruptcy.
In any case, productivity trumps efficiency any day. I’m not defending waste, but prizing market-clearing over plenty is perverse. The real problem for the tourist in the souk is the strain of dealing face-to-face with people for whom 50 dirhams is a lot of money, since their productivity is very low, and closing the deal with you is important. You can’t miss the anxiety and tension. Worrying about prices is a psychic cost.
Marx’s theory of alienation certainly captures important defects of the condition of the worker in industrial capitalism who completely loses control over his or her working day. It is also based on an absurd idealisation of the pre-industrial condition. What was the difference in autonomy between the Lancashire mill worker of 1840 and the household servant in Blenheim Palace in 1740? In the ancient world, domestics were slaves – that’s the etymology of “servant”. Pre-industrial peasants could in theory set their own work agenda, but this freedom was tightly constrained by custom, the environment, and fixed technology. The pre-industrial artisan had more options. But can we really believe Marx’s dithyramb to an independent producer’s relations to a customer:
I would have the direct enjoyment both of being conscious of having satisfied a human need by my work, that is, of having objectified man’s essential nature, and of having thus created an object corresponding to the need of another man’s essential nature. . . . Our products would be so many mirrors in which we saw reflected our essential nature.
That’s not how I saw Lu’s transactions in the souk. On the contrary, each participant was forced into a sports-like contest of will and wit, without the winner’s reward in a real sporting competition of third-party esteem. In haggling, the human relationship is narrowed into a focus on money as alienating as wage slavery.
When I deal with the lady wage slaves at the Spanish supermarket ham or fish counter, neither they not I can affect the prices set by the store management. Our conversation is as free and easy as we want it to be. They are under a little management pressure to be civil, but there’s no supervisor in sight and little monetary incentive to do more than the minimum. I find this a far more human and less alienated relationship than haggling.
I wonder if this helps explain the prejudice of Ancient Greek élites against trade, channelled here by Aristotle in the Politics, I. X – my italics:
There are two sorts of wealth-getting, as I have said; one is a part of household management, the other is retail trade: the former necessary and honourable, while that which consists in exchange is justly censured; for it is unnatural, and a mode by which men gain from one another.
“Gaining from one another“ is, you would have thought, a feature not a bug of market exchange as of sex. But exploitation is intrinsic to haggling. We are better off with its demise, and let Walras rest in peace.
* * * * * *
The other psychic benefit of industrial service is the lessening of rivalry between workers. The independent cotton-spinners are in anxious competition; the mill-workers are all in the same boat, and friendship and solidarity will help them stand up to the boss. Industrial employers still try hard to break down this solidarity and re-atomise the labour contract with individual negotiations and incentives, but their success in this is necessarily limited by their scale of operations.