Oddly Influenced

Welcome to oddly influenced, a podcast about how people have applied ideas from *outside* software *to* software. Episode 14: David Graeber’s three kinds of economies.

This is the second episode based on David Graeber’s book /Toward an Anthropological Theory of Value/ and chapter 5 of his later book /Debt: The First 5000 Years/.

Here, I’ll talk – to the extent my first cold in three years will allow me – about Graeber’s idea that any society is a mixture of what he calls baseline communism, exchange, and hierarchy. Gifts play a role in each. Each society will have its own variant of each of the three economies, in differing proportions.

My hope is that if you’re aware of how, say, continued exchange works, you might be able to apply the ideas to the mini-society of your own team.

What Graeber calls “baseline communism” has nothing to do with collective ownership of property, so don’t be alarmed. The core idea is Louis Blanc’s slogan, “from each according to their abilities, to each according to their needs”.

That’s a common way families work. My own family works that way. If one of us needs something, someone else will try to provide it.

I think pretty much all cultures consider the family style of baseline communism completely normal, so we think it odd that, for example, in 1881, on his 21st birthday, Ernest Seton’s father gave him a bill for all of the expenses his parents had incurred since his birth, including the doctor’s bill for delivering him. Ernest paid his father back, and then never spoke to him again. I bet very few people fault him for that.

Graeber gives examples of non-family baseline communism.

Suppose a child falls off the subway platform. We wouldn’t find it odd for someone to jump down and save the child, even though:

- it’s a cost for the person doing the jumping
- there’s no reason to think the act is an investment, that the child will someday, somehow “pay you back”.

Instead, the need is huge and the ability (to rescue) is much smaller. So the rescue happens.

Another example would be a smoker with an unlit cigarette approaching a smoker with a lit cigarette and asking for a light. It would be very weird for the lighter-having smoker to refuse. There’s a need. The ability is there. It’s a tiny cost. So almost anyone would help. No smoker would fret that the cost of giving a particular light to a particular person won’t be repaid, because everyone knows that anyone in the community of smokers will give a light when asked.

The key thing is that these relationships are expected to be *eternal*. There’s no need to keep track of who’s done what for whom because the base assumption is that everyone will provide according to their ability to help those who have need.

The second type of economics is exchange. Exchange implies that both parties end up roughly equally well off. So if you’re bargaining to buy a car, you can go back and forth until both of you are satisfied with the deal. You prefer having the car to having the agreed-on price in cash, and the dealer prefers having that money to having the car. So you settle at a point where you’re both satisfied.

I’m going to divide exchange up into three types: commercial exchange, balanced exchange, and continued exchange. These are not hard-and-fast categories – since, after all, almost nothing human comes in hard-and-fast categories. Nor do these categories exactly come from Graeber, so don’t blame him for them.

A key property of commercial exchange is that it’s impersonal. Once it’s over, so could be our relationship. In principle, I have no more relationship to a clerk selling me a Diet Coke than I would have to a vending machine doing the exact same thing. Oh, because we do share a society, the clerk and I will probably exchange pleasantries, but once I walk outside of the store, I could easily never see that person again, and neither one of us would find that odd.

Commercial exchange takes place, conceptually, at a single instant. I say “conceptually” because sometimes one half of the exchange is delayed to a later instant. Back in my consulting days, I’d make a deal with a company for me to spend, say, a week there. After my visit, I’d send an invoice and they would pay me, usually reasonably promptly, sometimes after I nagged them. It was still a commercial exchange because they were obligated to pay the amount we’d agreed on when we signed the statement of work or contract or whatever. (Perhaps surprisingly, in almost 30 years, I never had a company outright fail to pay up.)

A *balanced exchange* differs from a commercial exchange in that one half of the exchange happens an *unknown* time after the first. A typical example would be my inviting you to dinner next week, and – at some comfortable interval after our meal – you would invite me out to dinner. The dinners will seem to each of us to have the roughly the same value, which is what makes it a *balanced* exchange. You *could* fail to ask me to the return dinner, but that very well might put an end to our relationship. Certainly I would no longer do you any favors.

Graeber points out that, quote “As anthropologists have long been in the habit of pointing out, the very existence of such customs – especially the feeling that one really *ought* to return the favor – can’t be explained by standard economic theory, which assumes that any human interaction is ultimately a business deal and that we are all self-interested individuals trying to get the most for ourselves for the least cost or least amount of effort. […] Why, if I took a free-market economic theorist out to an expensive dinner, would that economist feel somewhat diminished – uncomfortably in my debt – until he had been able to return the favor?”

It’s important to note that balanced exchange happens between people who are roughly equal. Quote “If Bill Gates or George Soros took him out to dinner, [our economist] would likely conclude that he had indeed received something for nothing and leave it at that.” I personally think it would be a bit weird for a mere professor to invite Bill Gates to dinner in order to return a favor.

The same can happen in the other direction. Quote: “If some ingratiating junior colleague or eager graduate student [made a dinner invitation], [our economist would] be equally likely to conclude that he was doing the man a favor just by accepting the invitation.” And so he wouldn’t reciprocate, because his mere presence at the first-and-only dinner had immediately repaid any debt.

Competition does come up in balanced exchanges, but in a different way than in commercial exchange. In a commercial exchange, competition shows itself in the attempt to get the best deal for yourself. In balanced exchange, competition tends to show itself in an attempt to be more generous than your counterparty. This can be playful one-upmanship, but it can easily edge into aggression when the more generous party is flaunting his wealth with a gift that can’t be reciprocated. That starts looking like the third economic system, hierarchy, which I’ll cover later.

A balanced exchange sets up a relationship that can be ended after the gifts balance out. I might never invite you to dinner again, and that would be OK. A *continued exchange* is just one in which the cycles of gifting continue on indefinitely. From Graeber’s examples, it seems to me that a continued exchange is easier the less comparable individual gifts are. In the example I used last episode, I gifted my neighbor help fixing a bike and he gifted me mowing the lawn. It’s not worth worrying about whether those two things balance out, so we’re both encouraged to keep giving under the assumption that it will all even out in the end.

Another thing that encourages continuous gifting is increasing the number of people involved. The lawn mowing wasn’t exactly a gift to *me* because Dawn’s the one who mows the lawn. That’s maybe not the best example, since I think most people treat the two of us as a unit. However, I’ve also fixed Phyllis’s bike down the street, and Dawn helped the guy at the other end of the street with his garden, and Steve across the street let Dawn borrow his truck to haul rocks, and… and… and…

I think the trick here is that adding complexity about who owes what to whom keeps everyone gifting over time. We just assume that “what goes around comes around” – it’ll all even out in the end. That means the social relation of “neighbors on the same block” keeps getting recreated over time. There’s still *some* amount of keeping track going on, and someone who’s clearly taking advantage would end up excluded from the relationship, but much less effort is spent worrying about being “cheated”.

It seems to me that continued exchange is moving from commercial exchange toward baseline communism. The difference, Galison says, referring to the continuous gifting of the Tiv women from the last episode, is that, quote “There was certainly a trace of communism here – neighbors on good terms could also be trusted to help each other out in emergencies – but unlike communistic relations, which are assumed to be permanent, this sort of neighborliness had to be constantly created and maintained, because any link can be broken off at any time.”

The last of the three economic styles is hierarchy, where one party is superior to the other in terms of status or power. The relation is often *described* in terms of reciprocity. For example, feudal societies often described themselves as being composed of peasants, who feed everyone; aristocrats, who protect everyone; and priests, who pray for everyone. However, the idea of peasants deciding how much wheat is owed for a certain amount of protection is absurd. Instead, the debt is regulated by custom: how much did their ancestors give? how much was given last year?

And the top of the hierarchy has the most power to determine what the custom is, barring the occasional – and usually unsuccessful – peasant revolt. The medieval historian Marc Bloch describes it like this: quote

“In the ninth century, when one day there was a shortage of wine in the royal cellars at Ver, the monks of Saint-Denis were asked to supply the two hundred hogs-heads required. This contribution was thenceforth claimed from them as of right every year, and it required an imperial charter to abolish it. At Ardres, we are told, there was once a bear, the property of the local lord. The inhabitants, who loved to watch it fight with dogs, undertook to feed it. The beast eventually died, but the lord continued to exact the loaves of bread.”

In a hierarchy, gifting can’t be used to change or establish status because gifts can’t be compared. No amount of wheat delivery will make a peasant a noble. In heroic societies like the Greece of the Iliad and Odyssey, or the Kwakiutl, the extravagant gifts of the “Big Man” are reciprocated by having his followers… follow him. That is, fealty is the return gift. Often, that includes praise that sounds incredibly over-the-top to our ears as a way of reinforcing who’s up and who’s down in the relationship. Here’s Bach’s dedication of the Brandenberg Concertos to the Margrave of Brandenburg-Schwedt:

quote

As I had the good fortune a few years ago to be heard by Your Royal Highness, at Your Highness's commands, and as I noticed then that Your Highness took some pleasure in the little talents which Heaven has given me for Music, and as in taking Leave of Your Royal Highness, Your Highness deigned to honour me with the command to send Your Highness some pieces of my Composition: I have in accordance with Your Highness's most gracious orders taken the liberty of rendering my most humble duty to Your Royal Highness with the present Concertos, which I have adapted to several instruments; begging Your Highness most humbly not to judge their imperfection with the rigour of that discriminating and sensitive taste, which everyone knows Him to have for musical works, but rather to take into benign Consideration the profound respect and the most humble obedience which I thus attempt to show Him.

close quote

Mind you, that’s just the first *sentence* of the dedication. And a margrave isn’t all that big a deal as aristocracy goes: it’s somewhere between a count and a duke. But the dedication shows that some random margrave is far above one of the greatest musical minds in history, and that’s just the way it is.

Now, this may be a stretch, but it seems to me that violations of customs associated with hierarchy are met with more outrage than is the case with the other economic relationships. It’s that special kind of outrage that comes from the violation of the “natural order of things”, “the way things ought to be”, or: the way things have always been.

At least in the USA, we’ve seen that outrage from restaurant owners now that we’ve decided the pandemic is over. They frequently complain that “no one wants to work any more”. To me, it looks like restaurant staff were using increased bargaining power to get higher wages and, especially, better working conditions. Which ought to be, in a nominally free market system – that is, in a commercial exchange system – morally neutral, but the restaurant owners didn’t see it that way. They saw minions getting uppity.

That’s not to say people who live in non-hierarchical systems don’t get upset with rule-breakers. Here’s a quote from Graeber: “One extreme example [of breaking off a relationship] is the Maori story about a notorious glutton who used to irritate fishermen up and down the coast near where he lived by constantly asking for the best portions of their catch. Since to refuse a direct request for food was effectively impossible, they would dutifully turn it over; until one day, people decided enough was enough and killed him.”

Now I’m going to look at how all this applies to a software team. Specifically, I’m going to talk about the teams that Einar W. Høst analogizes to “terrariums”. Quote: “It feels like parts of the agile community still promote a ‘productive terrarium’ model for software development, ‘sheltering the team’ from the environment, trying to control and optimize the conditions inside the terrarium, […] asking if the busy silkworms are feeling safe and thriving.” I’m going to focus on that situation because that’s my personal experience – coaching teams on how to be thriving silkworms, safe behind glass walls – and it was in the context of such teams that I started thinking about gift economies.

I’ve often explained Agile as a deal between the development team and product management: if you (management) tell us what to do but not how to do it or how long it will have to take, then we promise to deliver you working software at frequent intervals *and* not to whine if you change your mind about requirements. (Note: I’m oversimplifying here by lumping together technical management and product management. I don’t think that loses much.)

This is pretty clearly a pure commercial exchange. As such, it’s fragile. One reason is that the deal is typically made before the team knows *how* to reliably deliver working software at frequent intervals. As such, there’s a reasonable chance that product management will be disappointed someday, meaning they’ll likely perceive that they were cheated, that the team didn’t hold up its side of the bargain. For example, it seems fairly common for a new Agile team to overdeliver; that is, to deliver features faster than it builds in the internal quality needed to support later change. That sets early expectations that later iterations will fail to live up to.

Because commercial exchanges don’t build lasting relationships, product management is disinclined to be forgiving of things like that. Unlike balanced or continued exchange, there’s no natural incentive toward generosity. So things fall apart.

Another problem is that commercial exchanges are supposed to be between rough equals, but it’s likely that product management and the team are in a hierarchical relationship. “These things (like making estimates) are our business: butt out” is not typically what those lower in the hierarchy get to tell to those who are higher.

*Except*, there is one person available who probably *is* a rough equal to – and already embedded in – the product management, um, society: the product owner. From now on, I’m going to call the product owner “Nick”, because treating people as job titles is the kind of impersonal relationship we’re trying to get beyond here. And besides, Nick is the name of the product owner in the team that was the clearest example I’ve seen of a commercial exchange turning into a continued exchange.

I want to note that that pleasing change was not the result of any perceptive coaching on my part, nor was it a conscious effort on anyone’s part, as far as I know. To be frank, it happened largely because of Nick’s personality, and I think it was greatly helped by the fact he spent about half his time in the team room, so that the rest of the team could respond to all of him, not just him-as-a-role-that-provides-stories. That is, Nick lived in the team; he wasn’t just a tourist. Tourists get commercial exchange.

All that said, I think it’s more useful to you for me to present a somewhat fictionalized version of the story, one in which the team intentionally builds continued exchange. “Just stumble into it” is not all that helpful as advice goes.

To the rest of the team, Nick’s most obvious role was to describe stories and accept (or, less often, send back for rework) finished work, finished stories. Looks like it’s a commercial exchange: paychecks for code. But let me be a bit fanciful and recast it into a Maori-style balanced exchange. Quote: “In eighteenth century New Zealand, English settlers soon learned that it was not a good idea to admire, say, a particularly beautiful jade pendant worn around the neck of a Maori warrior; the latter would invariably insist on giving it, not take no for an answer, and then, after a discreet interval, return to praise the settler’s coat or gun.”

The equivalent here is that Nick admires the potential in the team to produce a particular wonderful feature and the team hastens to provide it. This actually fits the old XP idea of a story as, not a specification, but an invitation to conversation. Rather than being ordered to produce something, the team is invited into a discussion about what would *really* satisfy Nick. That matters because we’re trying to build a mini-society here. Graeber describes a society as everyone whose opinion *of* you matters *to* you. The conversation shows Nick that his opinion of the rest of the team and their helpfulness matters to that team.

Let me give another example. When a story was finished and all the acceptance tests passed, someone would show off the story to Nick. Nick would sometimes begin to emit “yes, this is what I asked for, but it’s not really what I wanted” vibes. The natural reaction would be to see if you could get Nick to agree to accept the story as is, and schedule changes for some later iteration. That was especially true because this was back in the bad old days of story points and velocity, so rework threatened the velocity and thus the reputation of the team to its management.

A better reaction was more in the direction of insisting, Maori-warrior-like, on taking the story back and reworking it because, after all, that’s what would improve Nick’s opinion of the team. Quite likely, it would then be *Nick* who’d suggest putting off the improvements and seeing if the story would work OK as is. (Note that, by doing so, he’s giving a small gift to the team.)

But what does Nick *mainly* give to the team? I want to say it was not paychecks. Paychecks had nothing to do with the relationship. It was not a motivator. What Nick gave the team was *protection* for the terrarium walls. He represented the team and their work to the product management, and put them in a good light. He turned all sorts of crazy outside demands into a steady stream of well-described, reasonably small stories that *he* had to make sure satisfied the various stakeholders. He didn’t fall into the “feature factory” trap where each iteration was a jumble of stories, some from stakeholder A, some from stakeholder B, and so on. Instead, he created a coherent theme for each iteration, which the programmers liked because it helped them pace and organize their work. He did that even if it meant that he had to mollify stakeholder B who’d be partially *funding* the iteration but not getting anything out of it. Strikingly, one time the team had to request a “refactoring sprint” in which they did nothing but clean up code. Yes, an ideal agile team would alway be able to fold cleanup work into product work and yadda yadda yadda, but they needed this favor. So Nick gave them that gift, even though it’d be him that took the hit of explaining to product management that “nothing would get done” for two weeks.

It seemed to me that the rest of the team thought of Nick as the person who left the terrarium to go into the hard, cruel world to present the team to people skeptical they’d gotten their money’s worth and who would blame Nick if they were dissatisfied. Nick needs our help! was the attitude.

My impression matches Graeber’s description of the Tiv women, so I’ll finish by repeating his quote: “There was certainly a trace of communism here – neighbors on good terms could also be trusted to help each other out in emergencies – but unlike communistic relations, which are assumed to be permanent, this sort of neighborliness had to be constantly created and maintained.”

Even to *me*, what I’ve just said sounds like kind of an over-intellectualized description of what ought to happen naturally. After all, societies do form relationships of continued gift exchange edging toward baseline communism without having first read hundreds of pages of anthropology and pondering how it applies to their circumstances.

But the thing is, when it comes to development teams, what should be natural doesn’t happen often enough. Why not?

I have thoughts about that, but it’s something of a rant, and this episode is already long, so I’ll save them for a bonus episode.

For now, let’s just say that, whatever the reasons, the mini-societies of software teams settle easily into commercial exchange and hierarchy economies. So if you’d prefer the more pleasant sorts of economies, a little-overintellectualized pushing might be in order.

In any case, I hope you’ll find the terminology and ideas of this episode helpful as you think about your own teams. My gift to you.

And thank you for your return gift, which was your attention. Eh, that’s corny.