Oh, the other thing I had was the thought that the people who are really successful in our current economy don't let that whole "I don't have any money in my account" thing stop them, as my hypothetical buyer does. It takes money to make money and they'll borrow aggressively, putting themselves in millions of dollars in debt, in order to create whatever venture it is they want to make.


Edited at 2009-07-09 06:25 am (UTC)

(Reply) (Thread)

I have heard of a situation which kind of went to the opposite extreme. It was a babysitting time-bank, where hours were the medium of exchange (possibly with a fudge factor for lots of kids? I don't recall). It got into a state where everyone was hoarding their hours because they were afraid of not having hours when they needed them, so nobody was willing to pay hour-points for babysitting, so nobody felt like they could easily earn hour-points, so everybody hoarded the hour-points they had, so.... The administrators of the system ended up solving it by adding hour-points evenly to everyone's account, so that everyone felt more flush. Even though they were still all in the same relative positions, in a closed system, this worked, because emotions were part of the endless spiral of doom.

If I could remember more about it, I'd point you at the actual story and not just my memory of it.

Anyway, the way they solved the problem you referenced was to not have negatives allowed. Everyone started with (number pulled out of air) 20 hour-points and when the crunch happened, got increased to a point as if they'd started at (pulled out of air) 30 hour-points.

(Reply) (Thread)

On a larger scale, conspicuous consumption is a standard, but not entirely necessary way of playing the social standing game, which seems to be mandatory in human societies. It seems like if you were in a society that was sufficiently thought-divorced from American-standard, you might be able to shift the ways in which social standing is expressed to ones that worked better with your proposed economic system. They'd still be nasty ways of putting other people down, alas, but they could be different nasty ways.

(Reply) (Thread)

oh, boojum, you misunderstand. my king-sized bed of caviar (that must be replaced every day) isn't about being conspicuous or proving that I'm better than you, it's just the sleeping surface I happen to find gives me the best dreams at night...

(Reply) (Parent) (Thread)

heart.

(Reply) (Parent) (Thread)

Hee!

I would love to read an (accessible to the interested layperson) analysis of people pretending to not be engaged in social standing activities as a way of upping their standing. The closest I've seen is _Orbital Resonance_, which is fiction about post-apocalyptic twelve-year-olds in space -- fun, but not heavy on evidence or analysis.

(Reply) (Parent) (Thread)

Robin Hanson writes (at overcomingbias.com) about signaling activities all the time. His main thesis is this:

Food isn’t about Nutrition
Clothes aren’t about Comfort
Bedrooms aren’t about Sleep
Marriage isn’t about Romance
Talk isn’t about Info
Laughter isn’t about Jokes
Charity isn’t about Helping
Church isn’t about God
Art isn’t about Insight
Medicine isn’t about Health
Consulting isn’t about Advice
School isn’t about Learning
Research isn’t about Progress
Politics isn’t about Policy

(Reply) (Parent) (Thread)

is "yes, this." done yet? Cause, well, this.

(Reply) (Parent) (Thread)

The timebank (why is it called that?) sounds a little like how little Maya and Lindsey handled the notional money in our My Little Ponies' society. We didn't want to bother with currency, so we just had a notebook at the bank with everyone's balance written in it, and from time to time we'd shift money from one account to another, keeping the total balance constant.

But when I run with those rules in an economy where the buyer doesn't stop when he hits $0, I get a situation where the buyer never stops buying and the seller never stops selling Any exchange-for-currency is available at any time, which leads me to wonder why anyone's bothering to do any accounting at all.

Bounds on term sizes. In relational programming, where, for instance, you want to find all integer values of X, Y, and Z that satisfy the relation X + Y = Z, you have to somehow put a cap on the growth of one or more of those terms if you don't want an infinite loop. In the case of the timebank, could you just cut a person off from spending if they drop below a certain threshold?

(Infinite loops, it turns out, are all over the place in relational programming, so much so that at least one whole chapter of my friend Will's dissertation is devoted to Stuff You Can Try To Do To Make It More Likely That Your Program Will Stop Sometime.)

(Reply) (Thread)

It's called a "timebank" because the unit of exchange is hours. If you copy-edit Will's dissertation for an hour, you get one time dollar. You exchange it with me for one hour of finding and tuning up a bike for you to ride during your next visit. I exchange it with the acupuncturist down the street and she puts some metal needles in me for an hour, etc.

As for the threshold, if it's static, it seems like you're still using a fixed, centrally-issued money supply, just one where the chairman decides to issue $N to everyone once in their life. If it's dynamic, well, that makes things much more interesting. Also, complicated.

(Reply) (Parent) (Thread)

So, are all kinds of work worth one time-dollar per hour? I could see that system being abused. Say Paul and I are both willing to mow your lawn. Mowing your lawn takes about two hours, regardless of who does it, so it should cost you two time-dollars. But because I want your business, I pull you aside and whisper that I'm only going to charge you one time-dollar for it. I start doing this for everyone, and soon I have all the lawn-mowing business. I make half of what it should be worth according to the system, but I don't care, because everyone goes to me.

I like the idea of the bank, but I'm just not sure how much time should have to do with it. If I get better and better at making hats to the point where it takes me only an hour to make a hat that used to take five hours to make, should I only get one time-dollar in exchange for the hat? And what's to keep me from telling everyone that it did take five hours? Or ten hours, for that matter?

Maybe timebanking works best with exchange of services, like babysitting, not of goods, like hats. But even babysitting seems prone to the kind of abuse I mention in the lawn-mowing example. How are they dealing with these issues?

(Reply) (Parent) (Thread)

Huh! That's essentially the opposite of the concern I hear most frequently, which is "I want to charge more for my brain-surgery skill", not "what if scabs undercut my market."

I don't know. I'm not even sure if it's a problem yet. Although it is, now that I think about it, a slightly different expression of a relative-efficiencies problem I was thinking of earlier:

Say you and I both mow lawns. But you're lots better at mowing lawns than I am -- you can mow three lawns in the time it takes me to do one. So, you get all the lawn mowing business. Until peak mowing season hits and you run out of capacity for more jobs, at which point I pick up your overflow business.

So you're mowing lawns, and I'm mowing lawns, we're both doing good honest work and getting paid for it. Yeah, I'm getting paid more per-lawn than you, but there's still a limit to how many hours are in a day, so we're making about the same wages per day and together we're meeting the world's lawn-mowing needs.

Except the people who are paying me are spending more for the same service; consequently, it's like their money is worth less, just because they happened to not be lucky enough to get one of your lawn-mowing slots. What the hell? That sucks.

We could balance this out with a lawn-mowing co-op, where you and I get paid out of the co-op's joint account and the co-op charges some sort of average rate... that's about as far along as I've thought on that line.

(Reply) (Parent) (Thread)

Comparative Advantage says you and Lindsey can make the world a better place by doing what you're each better at. So instead of mowing lawns, you sit babies -- you can sit 3 babies in the same hour, while Lindsey can only really keep track of one baby at a time. And then, where before only one baby and one lawn could be satmowed, 3 can be.

This is one of the ways that price is supposed to be a signal in the marketplace. Low price says that you can do it efficiently, and that means more of it can be done for everyone.

(Reply) (Parent) (Thread)

Well, sure, until we get more lawns than Lindsey can mow by herself...

(Reply) (Parent) (Thread)

Wwell, your thought-problem (that people will spend infinitely) is probably stopped by a social understanding of the likelihood of cheaters. If I see someone with a large negative balance, I'm much more unlikely to offer them a service. Someone with a high positive balance, much more likely.

But I think the only helpful thing I have is that some folks in academia seem to call this community economics.

(Reply) (Thread)

Oh, so you won't take my money because I'm a cheater? My money isn't as good as anyone else's?

So here my account balance isn't so much of a amount-of-tradeable-assets, it's more of a deadbeat-meter. It seems like we're almost talking about a reputation system here. And you were the one who was talking to me about the "my personal worth as a human being is determined by my bank account" problem, which is certainly the direction I start to go with this...

although, maybe "I am my bank account" is only a problem because it encourages hoarding, and if you change the metric to be something like Transactions / (k + ABS(Balance)), (where k is some constant to prevent dividing by zero for balanced accounts), you exchange the hoarding characteristic for an entirely new set of behavior.

(Reply) (Parent) (Thread)

If I give you hours, and you're a cheater, I'm enabling a cheater. I'm pretty hardwired not to do that. Of course, your large negative balance isn't a guarantee that you're a cheater, but it's the best hint the numbers give.

This is functionally the same, not of taking your money, but of loaning you money, if I think you're already in debt. Or of taking an IOU from you if you're already in debt. Both of those most people would avoid doing.

Okay, so your metric is basically this: high numbers means that you've had lots of transactions, and/or your give/receive is relatively close to balance. low numbers means you've either had few transactions, or you almost entirely give or receive services.

So you're punishing the altruistic.

I'm at least far more interested by this metric than the old one.

(Reply) (Parent) (Thread)

it's more of a deadbeat-meter. It seems like we're almost talking about a reputation system here.

We are. Or rather, it seems like we have to be. This is actually a common phenomenon in economics, where it often makes more sense to talk about the exchange of debits than it does to talk about the exchange of credits. The way the world market works is one example of this. Ironically, if everyone paid off their debts, then the system would have no momentum and no country would have reason to lend another money. (There's a connection to programming in here actually, http://www.springerlink.com/content/m2105282ru426654/ )

The tricky thing about debt is that it's not only deadbeats who own debt. Someone who has a (relatively small amount of) debt, is someone who has a need. Since they need something you as a lender can provide, you have a reason to trust that they'll hold up their end of whatever bargain. Someone with a non-negative balance, on the other hand, isn't someone you can trust because you hold no power over them, they could break contract at any point.

In order to get a real appreciation for the deadbeat vs the altruist, you need to track more than one number. You need to track both credit and debt, which is how we do things in the current economy too. If someone has a lot of debt but also a lot of credit, then you know they're invested in the system (they're a good consumer, they just bought a house or something). If someone has a lot of debt and no credit, they're a deadbeat. If they have little of each, then they're someone who doesn't use the system (which can be interpreted as good or bad, depending on your own role in the system). If they have a lot of credit and little debt, then they're an altruist (who will be taken advantage of by deadbeats, and can't be trusted by lenders).

(Reply) (Parent) (Thread)

A mutual credit system is not without limits. Like some people mentioned. Right now you don't need money to get started trading with US$. You get a credit card. Nothing stops people from shopping till they drop, except maxing out their credit cards and the credit card companies making a decision that the person no longer has the ability to pay off their obligations.

In a mutual credit system, replace the credit card company with your neighbors and trading partners. Either the currency system imposes fixed commitment limits, or (as I hope is possible) the trading partners themselves can be encouraged to make good decisions. They can decide when to not accept money from someone with a balance so committed (negative), their ability to fulfill the commitment in reasonable amount of time is questionable.

Another option for limiting the amount of credit a person has, is for there to be a common rule, but instead of a flat limit (say $-300) it's based on a number of things, possibly including the cost of living. The community might say, we are willing to give a person the essential goods for 2 months, and then they have to start contributing back before they receive more.

A business man's goal is not to make piles of money. Some have that goal and nothing more, and they deserve to be awakened. The goal of business is the same as a job, to trade what you can do or make for what you can't do or make. We just use money as the token to keep track of what people have generated vs consumed. Using these tokens creates a sense of fairness and encourages people to give up what they have (item A) for something they want more (money to buy item B).

National currencies are currently so prevalent and controlled that the money it is equivalent as having the goods themselves. If you change to a currency that is not govt controlled, every time you go to trade, you have to ask yourself, by receiving this money, am I going to be able to eventually trade it for what I really want? If we can adjust the answer to that question based on the past trading behavior of the buyer, then the seller becomes the tool to encourage buyers to limit their consumption relative to their productive output. If we can automate the answer to that question, we can overcome a lot of friction in alternative currency transactions.

(Reply) (Thread)

Some have that goal and nothing more, and they deserve to be awakened.

Oh, so, my economy will only work with awakened people? Because one unenlightened broker can resell goods to an army of deadbeats, and easily keep _his_ balance in the black to, at first glance, seem like a productive member of society. So I either need a gatekeeper to only let in enlightened people, or my do-business-with metric needs to look a lot deeper in the graph than just one step.

If we can automate the answer to that question, we can overcome a lot of friction in alternative currency transactions.

Yes. Because "am I going to be able to spend this money" is really not a question I want to have to ask every time I sell a six-pack and a bag of chips at my convenience store. So I think it's got to be very low-friction to work in any wide-scale sort of way.

(Reply) (Parent) (Thread)

heart.

(Reply) (Parent) (Thread)

This is exactly like how real money works. Your confusion in one case directly applies to the other. The main difference is that we have decided that credit card companies and banks should be the arbiter of who will be able to make good on their claims, and complementary currencies seem to leave it up to the individual.

(Reply) (Thread)

Okay. That's a pretty valuable service. I think I'm quite happy to delegate that (the "who will make good on their claims") to specialists. And I'm not really sure I trust laypeople to do it in general.

Which suggests to me that I am either:

1. having my "specialists" be an open-source algorithm baked in to the accounting machinery, or
2. probably not interested in a complementary currency with an unrestricted membership

Edited at 2009-07-09 06:00 pm (UTC)

(Reply) (Parent) (Thread)

If this is just an alternate currency with a group of appointed experts to handle the economy, why not just use US$?

(Reply) (Parent) (Thread)

we have to get them from somewhere. people know their worth relative to the outside world, so they will demand their worth, and that's a lot of startup capital to make the group be able to function. this is a way that people with free time can create currency.

(Reply) (Parent) (Thread)

Your confusion in large part also mirrors a confusion that actual economists had until they started actually measuring these complementary currencies. You give all of these examples of how one might collude or cheat or otherwise-weird-up the system, but all of these things also all have social pressure not to do them. Previously, it was thought that money was a neutral carrier of value, and, while there were exchange rates between currencies, the upshot was that there was no difference between buying something in dollars versus euros versus local-exchange-time-hours. It turns out that this is false. For reasons we don't understand relating to human psychology, we treat different currencies differently. The best example of this is the studies regarding the Japanese time-banking system for elder care where people could pay in time-hours or in yen, and weird things happened.

This also means that different currencies have different in-practice problems, despite the fact that any reasonable mathematical model would treat all of the currencies as the same, and therefore the only problems "should" only be related to things like the size of the monetary base of each currency. The models that we intuitively make seem to be in-practice incorrect. http://www.predictablyirrational.com/?p=317&date=1

(Reply) (Parent) (Thread)

This largely mirrors what our local PDX Timebank ringleader says when she sits down with a few of us software engineers. We say "but there's this attack vector here and what about this and that and the other thing," and she says "yeah, but... I've never seen that actually happen."

So I guess in this discussion I'm trying to find out two things:

1. How do I explain the system so that geeks are willing to trust its integrity.

2. What elements are important to encode in the software? Do we need strict control of account provisioning, do we need a hard limit on negative balances, do we need people's balances or their transaction histories to be viewable to the rest of the community, etc.

   Because a "anyone can do any transaction at any time" platform is, if nothing else, much quicker to write. ;)

(Reply) (Parent) (Thread)

I would be very nervous about instituting this in software. We relate to physical things much better than numbers on a screen. People aren't willing to steal currency, but usually are quite wiling to be dishonest about numbers or to take small items from work. Use physical tokens is my advice.

This smells of attempting to use technology to solve a social problem, which is generally not the right thing unless you know both teach and people at a deep intuitive level. If I had to implement it, I would go for maximum transparency and utility and allow any transaction at any time, but also allow rollback by either party at any time. Make it a wiki of fake money, and give it the same robustness that a wiki has - it can be attacked, but the damage can be easily undone.

No geek will trust it until they see it working for a few years. Even then, they will delight in pointing out how it is unreliable and can be attacked. Hardcore security nerds still say that *Wikipedia* should be 100% untrusted, despite the fact that it contains more material at a measured accuracy only slightly behind the Encyclopedia Britannica. No geek will intuitively trust this system unless they are also already either a wiki-type or a hippie or both. And even then they will probably want more technological controls than is appropriate or good.

(Reply) (Parent) (Thread)

...know both teach and people...

know both tech and people

(Reply) (Parent) (Thread)

the name WikiWikiMoney has been claimed by the Twollars people.

(Reply) (Parent) (Thread)

For an answer to 1, I would say you should point to Bruce Schneier's comments on credit card security and fraud. transaction authorization and compliance. And build your software from the point of view of changesets.

(Reply) (Parent) (Thread)

also, how do social pressures change with questions of scale? Do the same systems that work with a local elder care community work if we scale it up to the thirty thousand people in the Woodlawn neighborhood? Does it work with the half a million people in the city of Portland? Does it work with 85 million ebay users?

(Reply) (Parent) (Thread)

Nobody knows. Anyone who claims to know is either a researcher in the field or a liar or both. At least, nobody knows the answer to YOUR question. People and groups thereof tend to generalize very poorly.

What is known is that big is different in kind from small. Solutions for large are usually incorrect for small and vice-versa. My mom knows all of the baggage checkers at her local airport (which operates only 3 days a week) and yet they still check all her luggage and make her take of her shoes and everything else. What is appropriate at a large airport is ludicrous at a small one. The most important thing to understand is that there are no known "laws of people" like their are laws of physics or principles of program design. People are hideously complicated bizarre nonlinear things.

http://alistair.cockburn.us/Characterizing+people+as+non-linear,+first-order+components+in+software+development

(Reply) (Parent) (Thread)

Speaking of the monkeysphere, your answer in general to trying to extrapolate any facet of human behavior will be "no, that doesn't apply here" regardless of what that is, where here is, or what direction you're generalizing.

This is famously one of the problems with socialism of the Marxist or hippy-commune variety. When groups are small enough that everyone knows everyone else, our tribal biology kicks in and we have a hard time taking advantage of Sally Sue down the street who has three kids to feed and is working her way through college. But when the community scales to a size where, with decent probability, you don't know the victim of your attack, all those instincts go out the window and it's all just numbers on a screen. Interestingly, there isn't generally a hierarchy here: once the system is big enough, you don't care if it's Sally Sue you take advantage of or some random person on the other side of the city (modulo any other communities you and Sally may mutually belong to).

The social and psychological aspects are going to be far and away the dominant issue in any sort of economy (whether financial, timeshare, reputation, or whatever). At the small scale where most participants know each other, having something full-transparency like a wiki will probably work (that is, will work under the assumption that the participants are all in it for the same goal). At a large scale, the equivalent of spammers, scammers, hucksters, and charlatans will script their attacks on any sort of technological backing the system has. Proving semantic security of the system against manipulation requires a deep understanding of the human psyche as well as the chops at proving security. Technical feats aside, all the hard-core security folks I know are cynics down to the core, which cuts them off from a wide range of the human psyche.

All of this together means that you'll need to have some sort of cellular structure, like terrorism made so popular. This way the cells are small enough to ensure tribal bonding, thus limiting the range of attack vectors. The downside here is that each cell will need its own currency and the mechanism for exchanging with someone outside your cell will be an involved one-time deal between cells, since it requires negotiating the terms of exchange and analyzing the trustworthiness of the other cell. To reduce the friction of inter-cell exchange, people will tend to set up alternative economies with different cellular boundaries so that they can temporarily change allegiance and perform an intra-cell exchange in this alternative economy. Because people are participating in multiple concurrent economies, the power of any one economy is diminished and excessive control of it will lead to people converting to use a different economy instead. Which is human history in a nutshell :)

(Reply) (Parent) (Thread)

I like the idea of an asymptotic limit. Say once you get more than a certain amount in the hole -- 10 hours of favors owed -- your value decreases 10%. So if you ask someone for another favor, she helps you for an hour, but she only gets 0.9 credits. Ramp that up and you'll hit a point where you can't get any farther in debt. But in practice, you won't get that far, because there will be a rising disincentive for other people to put you off until you rebalance yourself.

(Reply) (Thread)

Er, incentive. Or disincentive. You know which I mean.

(Reply) (Parent) (Thread)

So, it's not that services become more expensive when you're in debt, but people providing services to debtors receive less for their time. Interesting! That does certainly provide the seller with a disincentive.

I can think of ways that might go weird -- would sellers take disproportional affront to being paid at 90% and not do business with debtors at all? Would debtors deliberately surf that line to get cheap labor? That might be harder to exploit for service exchange (as is the primary focus of a timebank), but if you were exchanging goods, it could get exploited pretty quickly.

(Reply) (Parent) (Thread)

This is exactly what credit cards and loans do/are. Better credit risks get better rates, worse ones get worse rates.

(Reply) (Parent) (Thread)

Err, except the way I read it, I think what Andrew is suggesting is flipped 'round from that. The worst debtors get the best rates -- at the extreme end of the spectrum, they get services for epsilon-per-hour. The trick is that sellers will be decreasingly willing to provide services for the crazy rates.

(Reply) (Parent) (Thread)