From: Daniel Fabulich (daniel.fabulich@yale.edu)
Date: Thu May 28 1998 - 01:45:04 MDT
Soory it took so long to reply to this one, but I was making up HTML
documents with images to explain the graphs better. They're referenced in
this e-mail.
On Sun, 24 May 1998, Michael Lorrey wrote:
> If I can demonstrate in court that the information is X percent congruent with the information I was
> licensing (as the information content in and of itself is its own fingerprint), then I can still sue you as
> an accessory to breach of contract, even if I cannot determine the original breaching agent. Keep in mind
> that under a libertarian system, there is no right to refuse self incrimination. Since I would be backed by
> my PPL which has a standard practice of using veradication technology on individuals testifying in court, I
> doubt that you could escape prosecution.
And if I bought the info from someone who had bought it from someone who
had licensed it from you? Then I'm not even an accessory; I'm one step
removed from any crime that took place. I'm also not bound by any
contract to do anything for you; are you going to try to enforce terms on
a contract I didn't sign?
In addition, you get even worse problems when the merchant is also
anonymous. Suppose I post through anonymous remailers and a mail2news
gateway that I have a copy of your manuscript available for anyone who
sends me the price of a copy in anonymous cash. I don't know to whom I'm
selling, and they don't know from whom they are buying. How are you going
to prosecute me THEN?
> IP law is merely a societal contract under which the government is delegated authority to enforce the
> individual contract rights of all of its citizens. It does not matter if the government is a monopoly state
> or a competetive PPL. The principle is the same.
I disagree. Normally I cannot enforce the terms on a contract upon
someone with whom I have not contracted; only through special IP laws is
this possible.
> I read that, which was what got me started on this whole thread. You still treat patents as monopolies,
> while I treat them as competetive advantages in a competetive market. If you stop treating patents as
> monopolies, then your argument goes away.
I think on some level one must cede this definition, at least on the point
that a copyright holder has a monopoly over copies of that particular
piece of information.
> With electronic duplication and transmission, the costs are almost nil.
Yes, but it's that "almost" that would keep the market going, at much
larger quantities than before, at a far far lower price. Water is almost
free, yet it sells regardless. Since copies without IP laws are widgets,
the analogy is apt.
> But I can go buy the Red Hat CD and reproduce it if I want. I have a CDR deck on my PC here, and I can get
> CD blanks for less than $2.00. Where is Red Hat gonna be when I can sell copies of their CD for 10% of what
> they sell it for, or even 20%? I could offer it on my web site for free transmission. Obviously Red Hat
> has put some added value into their Linux offering, so how do they expect to recoup their investment if
> everyone can freely reproduce their product?
Beats me, but they do it anyway, and people DO put it up on their
web sites. Trouble is, it's rather large; so large that it's often more
efficient to earn $50 and buy a CD than it is to spend your time
downloading Linux. But nonetheless, it's available online, and yet the
CDs still sell.
Weird, isn't it?
> Even if it slowly advances, since there is no longer sufficient incentive for most people to invent, since
> they have to worry about feeding their families first, invention will again become merely the province of
> rich bachelors who can't get laid.
Whenever the free market rations anything, it rations it to the rich and
to the exceptionally devoted. Fortunately for us, the exceptionally
devoted are also in many ways the most productive. So it's a trade off.
At any rate, if we gain economic efficiency as a result of this trade off,
it's a good thing.
> Its actually rather simple, though tedious. Its merely a measure of growth of sales of products bearing
> chinese patent numbers. Just as today's economic growth here in the US is almost entirely in the high tech
> sector. Kill IP in the US and you will have economic stagnation here.
I think you misunderstood my point. Clearly, some of the decrease in
invention came from the lack of IP, but I think you'd be hard pressed to
argue that the planned economy had nothing to do with it. I was asking
how one would go about discovering what part of the lost patents were due
to a planned economy and what part was due to a lack of IP.
> Yes, this is rather indicative, isn't it??? A lack of ability to forecast productivity increases due to
> technological innovation seems to be a rather huge gaping hole in a model which claims to be so accurate,
> doesnt' it? Of course, they only give average accuracy numbers. I'd like to see how the accuracy numbers
> have changed over time. Even Greenspan has noted how current economic models have totally failed to predict
> the present economy, which is almost entirely due to increased industrial productivity derived from the
> high rate of technological advancement. I'll bet that the accuracy of the model is far less now than it was
> 20 years ago. I note that it is also restricted to a forecasting ability of only 4 years.
Economic forecasting is not entirely unlike weather forecasting to that
extent. However, fortunately for my argument, these are all macroecnomic
forecasts, not microeconomic forecasts. My argument from microeconomic
efficiency is not affected by the failures of macroeconomic models.
> How do you expect to demonstrate their relative sizes without knowing how large either shape is??? If you
> can do that its you that should get the nobel.
The men who have done so have ALREADY won Nobel prizes; I stand upon the
shoulders of giants when I follow their work. As to how I can determine
relative sizes without knowing the size of either shape, we use
calculus and the supply and demand graph to tell us a lot about economic
efficiency; for example, since monopolists represent the entire market,
their marginal revenue curve decreases twice as fast as the demand curve.
Since firms will produce up until the point where marginal revenue equals
marginal cost, the quantity produced by a monopolist will be less than
that produced in competition.
Unclear? I've made graphs.
http://pantheon.yale.edu/~dgf4/monop.html
I've just demonstrated that the monopolist's quantity will be less than
that produced in competition, despite the fact that I don't know how many
would be produced in competition or how many the monopolist would produce,
verifying my earlier claim that I could find relative sizes without
knowing absolute sizes.
> SUch schemes are not currently used as the cost is currently greater than the cost of IP enforcement at
> this time, due to the sufficiently widespread respect for IP protection. If this were to change, then such
> schemes would then become cost effective. One such scheme takes advantage of passive boot sector virus/worm
> technology to monitor tampering. If write attempts are made to specific hidden read only files, the virus
> or worm built into the software will corrupt enough data to foil the tampering attempt, making the software
> unusable.
A priori, I cannot show that it would be cost effective to protect
software in a secure way; I can do so using market numbers, however.
According to the Software Publishers Association, "Few other industries
lose as much revenue to theft, estimated at $13.2 billion; nearly 1 out of
2 copies of software is illegally installed."* In light of this, I find it
hard to believe that eliminating piracy would be unprofitable, even if it
is thanks to piracy in other countries.
On the contrary, the reason software companies don't use this particular
scheme is that it wouldn't work: Just set up a copier that emulates a
computer, that records the program rather than executes it, and you've got
a fully functioning warez machine. Nintendo game copiers operate on
precisely this principle; if you weren't aware, there is a thriving warez
market for pirated Nintendo and Super Nintendo games, both of which come
on ROM chips soldered into the game cartridges. A "tamper-proof"
cartridge would do you no good: All I have to do is read the game off of
the cart, and I'm golden.
* http://www.spa.org/piracy/cpcfacts.htm
> However, you can predict the change in incentive to invent, and determine the utility value of invention
> for inventors. This gives you the invention supply curve.
Beg your pardon? By precisely how much will the incentive to invent
change? My argument has been qualitative, arguing that the incentive to
invent would decrease, but not specifying absolutely by how much. In
order to find an actual supply curve, we'd have to know absolutely how
much the incentive to invest would decrease. How exactly do you propose
to do that?
> Then correlating rates of patents being issued
> with productivity growth in an economy, all other factors being equal, you can predict how change in the
> supply of inventions will change the productivity growth.
The whole point of my argument is that all other factors would NOT be
equal: the market for copies would greatly expand if copyright were
eliminated, and not having to pay for enforcement would lead to a
non-trival income effect. You can't just leave these out of your
calculations.
> While I assert that current IP is merely a matter of monopoly government acting as a PPL in protecting the
> IP licenses of its contracted (i.e. patents granted) citizens. If I were in a libertarian society, I would
> have a free choice of PPL agencies to contract the protection of my IP with, and would only engage in IP
> trade with indviduals that were customers of the same PPL or with competetive PPL's with which my PPL had
> IP agreements with.
A PPA which tries to enforce contracts on those who have not agreed to
their terms will probably not be economically efficient; similarly, it
will be impossbile to catch your intended victims over the Internet when
they are selling anonymously.
> I find it difficult to follow your charts due to the problems with lining up the characters. If you can
> produce either graphics or excel compatible charts I would appreciate it. Feel free to send these via
> personal email at retroman@together.net. In any event, I see something which you may not. If a competetive
> business only has to lower its price half as much to increase its market share as a monopoly does to
> increase the entire market, then obviously it is more efficient to have a monopoly provide a product. If it
> wants to increase its market size it must lower its price twice as much than if it were a competetive
> business. That tells me that the comsumers benefit twice as much from a monopoly than from a competetive
> business. I for one don't beleive that.
Do you still have misgivings after looking at the analysis on my web site?
> If the inventor has to recoup his research, development, and production refinement costs and his
> competetors do not, then he is at a competetive disadvantage to invent. Better to let some other schmuck
> take those costs and steal his inventions from him. In such a situation, the incentive to invent will be
> negative, and invention will stagnate.
Again wrong. Competitive disadvantage only takes place when MARGINAL
costs are higher, not when sunk costs are higher. For an analysis of
this, look at: http://pantheon.yale.edu/~dgf4/incent.html
The cost of invention is sunk, which means that it doesn't change the
shape of the supply curve. Competitive disadvantage only takes place when
your supply curve is higher than the other guy's supply curve, so this is
clearly not the case. Instead, we see from the graphs that invention is
only profitable if the increased revenues from invention outweigh the sunk
cost of invention; this incentive is smaller than that under a monopoly,
but not zero.
> If it were so, then why are all of the economists totally flabbergasted at today's economy? None of their
> models have predicted the huge increases in productivity which are what is driving our economic growth.
MACRO, not micro. Give me an example of where MICRO has failed us.
> If the person who bought my land from me signs my covenant, one clause in my covenant is that the buyer
> must agree that anyone he sells the land to is bound by the same covenant. Thus while you cannot reproduce
> the land (except as seen below), you also cannot change the minimum terms of sale to your customers, and
> their customers, etc. I can add additional terms, but the origianl terms cannot be recinded.
True, but again, say you license to Alice who illegally sells a copy to
Bob. Bob has broken the law, but now suppose he sells a copy to Carol. Is
Carol liable under the terms of a contract she's never signed? Now what
if she sells a copy to me? As you can see, you quickly reach a point
where it would be absurd to prosecute me under the terms of your contract:
I've never seen it, have never agreed to it, and am not obligated to do
anything to help you recover your loss; unless there are special IP laws
which require me to do so.
Similarly, if Alice sells me a copy anonymously over the Internet, and I
start reselling it anonymously, how will you ever catch me?
> While anonymous money may make a significant impact on electronic commerce, I don't think that anonymous
> production of IP will make any significant headway.
Not production, but RESELLING. I expect you see the potential NOW.
> No, but they have contracts with their PPLs, and I have contracts with my PPL. Any reasonable PPL will
> recognise the validity of a claim upon an individual who has received stolen goods, even if the original
> perpetrator cannot be found. As long as the stolen goods can be proven to posess a significant congruence
> with the goods of the claimant which are documented prior to the receipt date of the stolen goods, then a
> case and claim can be made.
If I am correct about the economic efficiency of IP laws, then it is more
economically efficient NOT to recognize such a claim than it is to do so:
it would be costly to prosecute all of those people, espeically when they
use cryptography, and it would also destroy a thriving copy market without
adding enough to the incentive to invent to make this worthwhile. And if
Friedman is right about PPAs not enforcing economically inefficient laws,
then PPAs won't hear such claims.
> Then you cede to the existing law. As IP law is prosecuted as civil offenses, rather than criminal offenses
> (except when it involves transportation across state or national borders), you agree to most of current IP
> law as practiced.
Er, no: I don't agree that IP law should be enforced on resellers, but
rather on those who broke the license in the first place; if this has
happened at all, I have not heard of it.
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 14:49:08 MST