From: Jason W Solinsky <solman@MIT.EDU>
Date: Tue, 30 Aug 94 08:39:44 PDT
To: Hal <hfinney@shell.portal.com>
Subject: Re: Cyberspatial governments?
In-Reply-To: <199408300105.SAA11868@jobe.shell.portal.com>
Message-ID: <9408301539.AA19053@ua.MIT.EDU>
MIME-Version: 1.0
Content-Type: text/plain


Hal speaks thusly:

> As I understand Jason's proposal, his government does not rely on force,
> but rather it acquires authority by people voluntarily putting themselves
> at the mercy of the government to a certain extent.  The principal
> mechanism I have seen suggested is for people to put some money into
> escrow or a bond which they will surrender (according to agreed-upon
> rules) if they break the laws of the government.

This is my simplest suggestion. I use it most frequently in situations
requiring a level of trust which does not exist. But it is difficult to
imagine the existance of a _powerful_ cyberspatial government that does
not base its power on the non-linearity of the value of information (i.e.
the fact that the act of communication tends to leave its participants
with a more valuable set of resources afterwards than they had before
the communication.)

In my most recent use of the idea, I once again suggested it as a
substitute for trust when cryptography will not suffice. You were
writting about the problems of anonymous entities and suggested that
you would have difficulty dealing with such entities because there is
no way for you to know when a company you are dealing with undergoes a
substantial change. But look at the physical realm. What is it that makes
companies disclose changes in their upper management? Why do they announce
major deals publicly? Why do they discuss strategy in their quaterly
filings? They might well be motivated to disclose positive things without
SEC regulations, but negative events show up because a government is forcing
them to make those disclosures.

I forget which financial magazine I read it in, but I recently saw an
article discussing precisely the same problem with reguard to mutual
funds. It seems that the federal requirements on disclosures of major
changes in mutual fund management are sufficiently infrequent that on
a number of occasions they have not been announced for months. Of course,
the article recommended more regulation :-(.

But the question is: How does an entity (any entity, not just cyberspatial
entities and not just anonymous entities) convince the people it deals
with that they will be notified immediatelly if any significant changes
occur? Its an issue of trust. I don't see how it is possible to guarantee
such a trust unless the entity put itself in hock to the extent that the
people it deals with would be hurt if it broke that trust.

> Now the simple objection I offer is that most people don't have enough
> cash lying around to effectively obligate themselves.  Most people,
> unfortunately, spend their money rather than saving it.  Even people who
> do have large sums of cash are, for that very reason, able to tolerate
> larger losses, so they will apparently have to put up very large bonds,
> which would have to be a strain on their liquid capital as well.

In just about 100% of these situations, I would expect an insurance company
to be involved. So while the cybergovernments and escrow agents or whatever
we call them will set a flat price, the insurance agent then has the ability
to enter into a more personal relationship with the entity being insured
(note there is absolutely no reason why the insurance agent needs to be a
third party [in fact there are many good reasons why this might not be ideal]
but the abstraction is a useful one.)

These agents could lower their deposits in exchange for controls over
parts of the decision making process. More importantly they could insure
one entity for multiple potential violations (thousands in fact) and thus
lower the deposit that way. Finally one would expect the least valueable
certifications offered by cybergovernments [my use of the word implies the
existence of some cybercitizenry to which access is made substantially
cheaper by possesing the cybergovernment's certification] to be subsidized
by the citizens of those governments. This last point should allow any
honest business to achieve higher levels of certification by establishing
a reputation. Will there be some entities which find themselves in a
situation in which it makes economic sense to break the rules? Sure, we
have plenty of them today. The cybercitizens and insured entities pick up
the cost. And each time such a loss occurs, information becomes available
that allows us to better match the constraints placed on receiving
certification and the variety of certifications to the economic value
derived.

I would also expect proliferation af a trend we are seeing in the physical
realm insurance business today... particularly in the employer
indemnification business. The insurance companies combine their insurance
with consulting on how to avoid the risks being insured for in the first
place. This has the advantage of making the consulting part of the business
receive compensation matched to the value it derives. It also allows
companies to displace the uncertainty traditionally associated with
government decisions. Normally it takes a significant amount of time before
businesses can be sure what the government means by a new law. The close
relationship between insurance company/consultant and customer allows the
customer to price the governmental risk of all possible decisions, and
immediately make the best decision and enter into a contract with the
insurance company that hedges all risk. (The insurance company presumably
has a sufficiently large protfolio to allow it to absorb the risk).

> And, for people who do have the money, how can they tolerate tying up a
> large sum of cash for such a long period of time?  Does the government
> offer interest?  How are the funds invested - safe or risky?  Low
> returns or high?  People want to diversify their investments, and I
> don't think they are going to be willing to put all their cash into
> this one lump sum bond.

What I have suggested previously, and what I still think is the best idea,
is to structure the deposit such that it can be spent by agreement of both
parties (if the government imposes a fine and the fined entity agrees or
if the entity decides to give up its certification and the government
says it is payed up) or (in the case of a dispute) by the decision of an
arbitrator (selected by a method determined at the time of the deposit).
Under this method the deposit is in cash and it doesn't go anywhere. BUT,
the type of cash can be any that both parties agree to. Since most
cyberspatial currencies will be invested in some way, the deposit will
increase in value [on average].

> When people do save money, it is often with the intention of spending
> it later.  They save money to put their kids through college, or for
> retirement.  Sooner or later their comes a time when they have to start
> consuming the nest egg.  Will this entail withdrawal from the benefits
> of the cyberspace government?

If you have been following the rules for a long time, the insurance
company will allow you to decrease your deposit. Most car insurance
companies do the same thing. It is possible the the government will
interact with the insurance company in ways that allow the insurance
company to also lower its (larger) deposit.

> To sum up, I don't think most people's lives are structured in such a way
> that they can credibly obligate and commit themselves to a potentially
> risky contract.  With physical governments people might say "as long as I
> live on this island I agree that the government can shoot me if I kill
> someone," and I will be inclined to believe that they will not try to
> commit murder.  But that promise is much less credible if all they will
> do is forfeit a $2,000 bond, if that's all the money they've managed to
> save.

All that is important is that the value that is to be lost match the value
that is to be gained by commiting the crime. It is my belief that most
people will be able to offer non-monetary assurances to insurance companies
that allow them to dramatically reduce their deposit.

Cheers,

Jason W. Solinsky