Joseph Tainter and the cost of complexity

What Tainter did with Gibbon in 1988

The Collapse of Complex Societies came out in 1988 from Cambridge University Press, in the New Studies in Archaeology series, and it opens with an act of filing. Joseph A. Tainter sorts every explanation anyone has offered for why states fall over into eleven themes: resource depletion, catastrophe, intruders, class conflict, social dysfunction, a chance concatenation of events, and the rest. Gibbon, Spengler and Toynbee land together in theme nine, mystical factors. They get a few pages and a shrug. Decadence is not a variable.

That box has real occupants. Gibbon, closing his general observations on the Western empire in 1781, called the decline of Rome "the natural and inevitable effect of immoderate greatness". There's a mechanism buried in that sentence, and Gibbon half-meant it as one. What the tradition after him kept was the organic metaphor: cultures as bodies that ripen and rot, with the historian arriving as coroner. Tainter wanted numbered propositions and a price.

"The decline of Rome was the natural and inevitable effect of immoderate greatness." Edward Gibbon, 1781

Collapse defined so that it stops being a tragedy

Tainter's definition is narrow and deliberately dull. Collapse is a rapid and substantial loss of an established level of sociopolitical complexity: the number of distinct social roles, the layers of hierarchy above the household, the volume of information the system moves, the specialised parts it has to keep fed. Rapid means decades rather than centuries. Substantial means it shows up in the dirt: fewer administrative tiers, less monumental building, shorter trade routes, an end to coinage. Bryan Ward-Perkins, in The Fall of Rome and the End of Civilization (2005), describes a Britain that loses coined money, wheel-thrown pottery and roof tiles within a generation of 400. Nothing in the definition mentions virtue.

Under that sits a claim about what a society does all day. It's a problem-solving organisation, and its standard answer to a stress is another increment of complexity: a new office, a bigger army, a longer supply line, more record-keeping, more people whose job is coordinating other people. Complexity works. It's also a bill, paid in energy, food, fuel, labour and the taxes that convert them, and Tainter's name for the flow that pays it is the energy subsidy. He compresses the whole argument into four propositions. Societies are problem-solving organisations. Sociopolitical systems require energy to maintain. Rising complexity carries a rising cost per head. Investment in complexity as a response to problems reaches a point of declining marginal return.

Membership has a benefit-cost ratio: what the polity gives you against what it takes. Turn that ratio negative for enough people and walking away is arithmetic rather than pathology. Tainter's word for the result is economising. Collapse becomes a drop to a cheaper level of complexity that the land can actually support. Salvian of Marseille, writing around 440, complained that Romans were running to the Goths and the Bagaudae to escape Roman tax collectors. He filed it as a moral scandal. Read it as a price signal.

How to read the curve off a real state

Watch the mechanism run. Stress arrives. The society answers with complexity, and the first increments are cheap and pay well: an irrigation ditch, a clerk who knows where the grain is. Cheap solutions get taken first, so every later increment costs more per unit of problem solved, which is what Tainter means by declining marginal returns on complexity. Past the inflexion point the state pays more to stand still. The reserve that used to soak up shocks is gone. Then something ordinary happens, a bad harvest or a disputed succession, and there is nothing left to absorb it.

Rome is the worked case. Tainter's empire financed itself out of other people's accumulated treasure, ran out of profitable conquests after Dacia in 106, and then had to fund its administration from current production. Diocletian, from 284, answered with more complexity. The army roughly doubled, the provinces were cut from around fifty into nearly a hundred smaller units, the court swelled, and the assessment fell harder on a narrower base. The coinage tells the story from underneath. A denarius close to pure silver under Augustus had become bronze with a wash of silver by the reign of Gallienus, 253 to 268.

For declining returns on administration Tainter borrows a joke that turned out to be data. C. Northcote Parkinson, in The Economist of 19 November 1955, set the Royal Navy of 1914 against the Navy of 1928: capital ships down from 62 to 20, Admiralty officials up from 2,000 to 3,569. The office grew as the fleet shrank. A sharper claim follows from the same model. A state ringed by rivals can't economise by collapsing, because a neighbour absorbs it first. Tainter's name for that condition is peer polity competition, and it means collapse in his strict sense needs a power vacuum. The Western empire had one. A world of competing states doesn't, which is why he thinks any modern collapse would have to be global or nothing.

The objection from Questioning Collapse, and the one that hurts more

The loudest attack arrived in 2010: Questioning Collapse: Human Resilience, Ecological Vulnerability, and the Aftermath of Empire, edited by Patricia A. McAnany and Norman Yoffee for Cambridge University Press. Its main quarry was Jared Diamond's Collapse of 2005, though the complaint about the category catches Tainter too. The Classic Maya did not vanish. The last Long Count date in the Maya area, carved on Monument 101 at Toniná, falls in 909. What stops there is dynastic monument-carving in the southern lowlands, while several million Maya speakers live in Guatemala, Belize and Chiapas today. McAnany's charge is that archaeology reads the elite and monumental record best, and so keeps mistaking the end of an elite for the end of a society.

Tainter has a reply. His definition covers sociopolitical complexity and says nothing about population or descendants, so a collapse that leaves farmers and villages in place is exactly what the theory expects. The Collapse of Complex Societies gives a passage to the benefits of collapse: for a cultivator, losing the tribute schedule and the war captain can raise the standard of living. Continuity of population describes his mechanism running, and doesn't touch the mechanism itself.

The objection I find hardest to answer is internal, and it's about measurement. Complexity and return are the two moving parts of the argument, and neither gets a unit anywhere in the book. The proxies are chosen after the conclusion: coin debasement, which a dozen fiscal stories explain equally well, and Parkinson's clerks, which began as a magazine joke. No curve here is drawn from data. Any outcome at all can be redescribed afterwards as the moment marginal return went negative. A theory that forbids almost nothing does less work than it appears to. This one is very hard to catch being wrong. The single part of Tainter that forbids anything, peer polity competition, is also the smallest part of him.

What the curve is still good for

Used as a diagnostic, the curve earns its keep. Three questions, put to any institution: what problem was this complexity bought to solve, is it still solving it, and who pays the marginal cost. Try them on a compliance department, a standing army, a ninth layer of review, an accreditation body that accredits accreditors. The answers are usually embarrassing. Tainter aimed the same question at industrial society in Drilling Down (2011), written with the petroleum engineer Tadeusz W. Patzek, where the energy subsidy is fossil carbon and the return on each research dollar keeps sliding.

The reason to reach for Tainter in an argument about why civilisations fall is that he explains where others deplore. His account runs on a cost curve and a population doing sums. Lost virtue never enters it. Diocletian's Edict on Maximum Prices, published in 301, fixed ceilings for over a thousand goods and services and made overcharging a capital offence. Lactantius says blood was shed over cheap and trifling things, and the goods vanished from the market. The prices went up anyway.