Suppose I borrow € 1 mil and buy a machine to make widgits. I expect to be able to make € 150 thousand every year, which should be enough to pay back the machine over ten years.
Now, if I only make € 75k every year, the bank is going to take the machine (which is collateral for the loan) and scrap it. But the scrap value of the machine is only half a mil. If I had been able to keep the machine, and the bank had written down my loan to half a mil, the bank would still have gotten the half-mil, and I would have gotten a profit too (if 150k can service one mil of debt, 75k can service half a mil). So clearly this is a net loss for society as a whole. There actually was a free lunch if the bank would just have written my loan down to half a mil, instead of liquidating and getting the same half mil.
Moderate inflation forces continuous writedowns of loans, which means less liquidations chasing sunk costs.
- Jake If you only spend 20 minutes of the rest of your life on economics, go spend them here.
The point of the loan is feudal tribute to the bank. Those who can't pay tribute must be punished.
The fact that this destroys productive economic activity and puts people out of work isn't relevant.
The net loss of prestige to the bank when debtors default is considered more socially damaging than the net financial loss.
q.v. Greece, etc.
But even if banks were run by honest, upright people who had the best interest of society at heart...
... there would still be genuine uncertainty about how much my machine could make, and how much it would sell for when my firm is liquidated. The bank does not know that it will only get half a mil until it has sold the machine. Nor does it necessarily know that I make only 75k a year. Maybe I'm making 100k a year and taking a suitcase full of 100-euro notes to Switzerland every time I go there to ski in the alps.
Even if banks were not run by gangsters and neo-feudalists, writing off sunk costs in a controlled and predictable manner would, I claim, still be a good idea. Because otherwise, reasonable people will end up doing unreasonable things in an attempt to recover this sunk cost.
If you want a system which aims at maximising collective benefit and the Greater Good, you'll need a new accounting system.
Currently the concept of 'risk' is a thin cover for 'loss of aristocratic prestige.'
I'd suggest that if a machine has been bought and paid for, there is no longer any risk to anyone who matters. So the concept of sunk costs doesn't apply.
Investment cash isn't a thing or a finite conserved substance.
In practice it's a nod of aristocratic assent for a project or service. And return on investment is really a social reward that increases prestige and power for those who 'invest' in projects that are deemed 'successful.'
It's possible to imagine an economy in which goods and services are produced according to overall social value, rather than according to the need to maintain an existing aristocracy or create a new one.
On a local scale, this is what local currency trading schemes do.
What's needed is an international equivalent. But that would mean many changes - not least of which would be the removal of the current layers of aristocracy, and likely the end of nation states too. (Nation states being one of the least efficient ways to organise trade for the planetary good.)
I'd suggest that if a machine has been bought and paid for, there is no longer any risk to anyone who matters.
That's true, but if the investor knows ahead of time that he can hold the bank hostage to the difference between purchase price and scrap value, then he only requires his business plan to provide enough income to cover the scrap value of the machine.
Which, going into the business venture, is bad.
In the example above, if I knew that my machine would only make 75k a year, I would not borrow 1 mil to buy it. Except if I could reliably get my loan written down to the scrap value of the machine, in which case I would (well, I wouldn't, but somebody would). And that would cost society maybe a quarter mil in lost energy, labour and other non-recoverable costs. That is a bad deal.
This is the central problem of credit: How to align both the ex ante incentives of investors and the ex post incentives of creditors with the interests of society. This is a decidedly non-trivial problem. My gut feeling is that you will need both automatic sunk cost writeoffs (like moderate inflation) and discretionary sunk cost writeoffs (like bankruptcies). Both methods have costs, and balancing those costs against their benefits is, again, a non-trivial exercise.
If economics were easy, neoclassical economists could do it.
But the physical capital you can buy with it is. And in a capitalist economic system, physical capital is rationed by price.
And while rationing capital by price is not optimal for all areas of society, it certainly is optimal for some areas of society. Which means you need to figure out how to make that form of rationing play ball with your financial system.
But the physical capital you can buy with it is.
There's no such thing as physical capital. There are only physical resources, human resources, and creative intelligence.
Oh - and politics.
The issue for me isn't whether costs are sunk or not, but whether any given accounting system increases or decreases collective intelligence. Individual opportunity falls out naturally from collective intelligence - which is the reverse of the usual capitalist belief, because the opportunity is of a type that doesn't lend itself easily to capitalist competition.
It's really, really hard to persuade people to start thinking outside the usual frames. It's like trying to persuade medieval deists that god doesn't exist, and they should stop framing beliefs, philosophies and actions in religious terms.
The issue is that aristocratic accounting destroys collective intelligence and opportunity. Not only does it waste (almost) everyone's time and effort, it also limits what a culture is capable of. It's inherently regressive and is a huge brake on the benefit multiplication that becomes possible with real innovation.
Forcing machinery and people to stand idle and destroying valuable resources without long-term planned benefit - all for the sake of accounting fictions - is really the height of atavistic stupidity.
But the concept of sunk costs isn't relevant because it accepts too much of the standard framing.
The real issue is the extent to which physical resources, human resources and creative intelligence are being used in an intelligent and progressive way.
Word tens of millions of people stand to see their lives ruined because the bureaucrats at the ECB don't understand introductory economics -- Dean Baker
Indeed.
But you still need a way to take people's toys away if they go into a project that they know or should know will never earn back the full ex ante cost they throw at it.
The game tree looks like this:
So society needs to pick some sort of mixed strategy - or take a third option - which penalises you for pushing ideas you know (or should know) are bad, but does not penalise you for pushing ideas that you wrongly believe are good.
And no, saying "rational social decisionmaking would not have this problem" is not a valid answer, any more than "assume a can opener" is a valid answer to the question "how do I open this can?"
The point being that if you're thinking in terms of social benefits rather than profits,you've already improved your collective prospects.
But I still maintain that no one should be punished for bad ideas. It's more useful to write off mistakes to trial and error and to work at building useful predictive ability into collective intelligence.
Punishment - or at least restraint - should be reserved for individuals who reliably cause harm, and who seem to be motivated by harm.
And that would include excessive self-interest.
A can opener also has a proven record of fitness for purpose, the problem is whether you can assume you have one at hand.
Since rational collective decision making may never have occurred yet in human history (do you have an example with its proven record of accuracy?) even if you can define it, you still have to assume you have it at hand.
In addition, your definition should hopefully take account of Arrow's theorem. tens of millions of people stand to see their lives ruined because the bureaucrats at the ECB don't understand introductory economics -- Dean Baker
The point is to make executive decision making public and explicit rather than private and implicit.
It's essentially public regulation of politics - which our current pretend-democracy pretends to do, but doesn't.
Okay, do it! What, indeed, is "rational collective decision-making"? Is it like the God of medieval theology? All-powerful, all-knowing and all-merciful and just? Does it lead to the same kinds of paradoxes as how can God be all-merciful and just at the same time, or whether God can create an object so heavy he himself couldn't move? tens of millions of people stand to see their lives ruined because the bureaucrats at the ECB don't understand introductory economics -- Dean Baker
It's not as if this stuff is rocket science.
In fact you pretty much can't do rocket science without it. Do you think NASA got to the moon by waving dead chickens over satellite parts, or getting rhetorically over-excited?
Or did NASA actually manage to do that rational planning thing, at least for a while?
I'm suggesting it would be interesting to use similar techniques in what we call politics.
Or actually to replace politics with something that has more chance of being useful to more people.
May I point out that engineering decision-making is not democratic?
Maybe you advocate enlightened despotism or philosopher-kings? Assuming a rational absolute ruler solves the problem of rational decision making. tens of millions of people stand to see their lives ruined because the bureaucrats at the ECB don't understand introductory economics -- Dean Baker
Do you think NASA got to the moon by waving dead chickens over satellite parts, or getting rhetorically over-excited? Or did NASA actually manage to do that rational planning thing, at least for a while?
"Engineering-style decisionmaking works well for NASA, so it must work for society as a whole."
"Walrasian double-auctions work at the fishmonger's, so they must work for society as a whole."
Compare and contrast.
Migeru:
I think we should realise that the banking system is central to planning function of society seen as a whole. Whether banking is public or private, it has the power to allocate credit and thus influence decisively who can or cannot carry out projects for which they don't have the resources at hand already.
And this:
ThatBritGuy:
The point being that if you're thinking in terms of social benefits rather than profits,you've already improved your collective prospects. But I still maintain that no one should be punished for bad ideas. It's more useful to write off mistakes to trial and error and to work at building useful predictive ability into collective intelligence.
I think we could device a better system then todays. I'll have a go, and then you can find the flaws, m'kay?
So we need to allocate credit. We need to be able to handle small (small company investment/mortgages/car loans) and large projects (factories, shipyards), we want to allocate credit for projects that benefit society and we want to do it in a manner that does not centralise power.
I would propose scrapping the banks lending facility, and instead introduce a program where the government deciding the amount of public financed investment and the citizens deciding where it goes through each citizen getting to allocate their share of the public investment (practicly this could be handled through systems similar to Kickstarter or Flattr). The point is that you do not get cash, but a share of investment to allocate.
The investment then becomes partially owned by the state, perhaps in a contractual relationship where the person or organisation managing the investment can pay off it over a period of time through a share of income flow or such. So a self-serving conman would be stuck with the govenrment partially owning his con-investment and probably have a hard time convincing people to invest in his next project while the mistaken entreprenuer have to take a longer time to pay of his machine.
But a not profit-turning investment should also be allowed if enough people thinks it is worthwhile. The object of investment then stays public property. Like if an artist convinces people to fund his big-ass statue, where the big ass then daily can inspire passersby.
So how would this way to allocate investment be worse then todays? And how could it be even better? A vote for PES is a vote for EPP! A vote for EPP is a vote for PES! Support the coalition, vote EPP-PES in 2009!
Which means that
we want to allocate credit for projects that benefit society and we want to do it in a manner that does not centralise power.
In the system I crudely draw up, if ABB wants to tap the public investment system for their new advanced factory thingy, they would probably run the numbers and see how many people they need.
Next step would be to market the investment in turn to employees (it is necessary for your jobs), owners (it is necessary for your profits), customers (it is necessary to produce the great products you want) and the general public (the world will be better with the products produced in this factory thingy).
So I don't see problems for technically sophisticated, capital-intensive production. I do see problems in marketing leveraged buy-outs to the general public, but then those are not a benefit to society. A vote for PES is a vote for EPP! A vote for EPP is a vote for PES! Support the coalition, vote EPP-PES in 2009!
But I still maintain that no one should be punished for bad ideas.
People should be punished for mendacity. And there are situations where you cannot reliably distinguish mendacity from bad ideas without the assistance of a telepath.
If you never punish bad ideas, the majority of these edge cases will be mendacity, because the proportion of bad luck is constant while the amount of mendacity increases in proportion to the probability that the response will be permissive. Conversely, if you always punish mendacity, then the majority of your edge cases will be bad luck, because you can voluntarily refrain from mendacity, but you can't voluntarily refrain from bad luck.
This means that there is a point somewhere on the scale where the loss to unpunished mendacity begins to exceed the gain from not punishing good-faith mistakes.
Yes, if you assume that you can always perfectly discern whether people are acting in good faith or not, then you can construct a system that never has to accept punishing bad luck as necessary collateral damage in rooting out mendacity. And if pigs could fly, reinforced umbrellas would be a growth industry.
If you never punish bad ideas, the majority of these edge cases will be mendacity, because the proportion of bad luck is constant while the amount of mendacity increases in proportion to the probability that the response will be permissive.
Not necessarily. Mendacity for the sake of it is pointless - or at least evidence of a mental disorder, or an interest in an unusual form of performance art.
The real issue is self-serving mendacity. The self-serving part not being limited to mendacity, but other actions too.
I'd suggest anti-social self-interest is rather easier to discover - e.g. totally open transaction records would be a practical requirement. (For whatever form of records were being kept.)
The issue then becomes symmetrical oversight with feedback loops that encourage open participation and prevent authoritarian and/or narcissistic damage to the common good.
And punishment doesn't apply - revenge is pointless, unless you want to be a biblical patriarch with a bad temper.
Restraint would be enough.
Totally open transaction records does not help you determine whether an investment that generates a positive operating surplus but negative profit was made because the investor expected it to be a profitable investment or because he expected to be able to hold society hostage to the sunk cost.
You want to propose one of those?
And punishment doesn't apply - revenge is pointless,
On the contrary, in an iterated prisoner's dilemma, for example, revenge is absolutely required to create a credible disincentive to self-serving mendacity.
There are insane results in game theory, but that's not one of them.
The real issue is self-serving mendacity.
(simple) mythomania vs agenda-based mythomania... ~"When an inner situation is not made conscious, it appears outside as fate." Karl Jung~
