The Right Way to Tax DAT
[This article does not concern software, not directly. It
concerns a parallel issue about sharing copies of music.]
[The article was first published in Wired magazine in 1992; the
text has not been changed; instead, I have added notes, in square
brackets and with italics or other emphasis.]
[The original article addressed the (then hypothetical) issue
of sharing music using (then just appearing) digital audio tape
recorders, since that is what the proposed US tax law was supposed to
address. Nowadays it could be applied to Internet file
[Another approach developed by the late Francis Muguet with my
assistance, which includes some of these ideas, is called
the Global Patronage
system (in French, Mécénat Global). I support both
solutions; that is to say, I favor adopting either one.]
Record company magnates don't like the digital audio tape recorder
(DAT), which can make
perfect copies of musical recordings. They fear that customers will
copy music themselves, and stop buying prerecorded music.
Threatening lawsuits, they have obtained from the manufacturers of
DATs an agreement to pay a fee for each DAT unit and each DAT tape
sold to consumers. This fee is to be divided among various
participants in the music business: musicians, composers, music
publishers and record companies. In addition, DAT manufacturers have
agreed to cripple DAT units so that they cannot make a copy of a copy
of a prerecorded piece.
Now the record companies have asked Congress to enact a law turning
this fee into a tax and prohibiting manufacture of DAT tapedecks that
function without imposed limitations.
The stated purpose of the tax is to “compensate”
musicians for copying done by individuals using DATs. However, 57
percent of the funds collected would go to record companies and music
publishers—leaving less than half to the people who participate
in the creative process. Most of these remaining funds would go to
musical superstars, and thus would do little to encourage musical
creativity. Meanwhile, DAT users would be unable to make full use of
the power of DAT technology.
Here is a proposal for a different system for taxing DATs and DAT
tape—one designed to support music rather than cater to vested
- Collect funds with a tax on DAT machines and DAT tapes, as the
current proposal provides.
- Use a survey system to measure the extent of copying of each
- Distribute these funds entirely to the people who create
- Adjust each contributor's share so that it increases more slowly
per copy as it gets larger. This spreads the funds more widely to
support a larger number of musicians adequately.
- Make no restrictions on the functioning of DATs.
What is the purpose of copyright?
The record industry presents its proposal as a way to
“compensate” musicians, assuming that they are entitled to
be paid for any copy made. Many Americans believe that copyright law
reflects a natural right of authors or musicians—that these are
entitled to special consideration from public policy. However, any
lawyer specializing in the field knows this is a misunderstanding, a
view rejected by the American legal system.
The stated purpose of copyright, given in the U.S. Constitution, is
to “promote the progress of science and the useful arts”.
Progress in music means new and varied music for the public to enjoy:
copyright is supposed to promote a public good, not a private one.
Yet copyright is often thought of as a natural right by laymen and
politicians, which often leads to wrong decisions about copyright
policy. Even courts, defining the details of the copyright system,
often let this thought creep back implicitly even though it is
supposed to be excluded. This is a conceptual error because it
mistakes a means (copyright) to a larger end (progress) for an end in
Promoting progress in the arts does not inherently justify the idea
that authors are entitled to any particular sort of copyright, or even
that copyright should exist at all. Copyright is justified if the
benefits of progress exceeds the burden that copyright imposes on
everyone except the copyright holder.
How do we make this cost/benefit comparison? It depends partly on
facts (how does a particular law affect musical activity and music
users) and partly on our value judgements about those results.
Let's assume that it is worth paying a DAT tax if the result is a
significant increase in musical activity, and investigate how we
should arrange the details of this tax in order to maximize the
benefit. But first, let's review basic principles and facts which
have a bearing on the inquiry.
The law of diminishing returns is a general principle of economics.
It states that each additional increment of efforts or funds spent on
a given goal typically produces a smaller and smaller increment in the
results. There are exceptions to this law, but they are local; if you
keep on increasing the inputs, you eventually leave the exceptions
For example, you can make traffic flow more smoothly by improving
roads. Adding one lane to 20 miles of congested roads in a city might
increase the average traffic speed by 15 miles an hour. Adding a
second lane to those roads will not give the same improvement; this
might increase the average speed by only 5 more miles an hour. The
next additional lane might make no noticeable difference if the
traffic jams are already gone. Yet each successive lane will cause
greater dislocation as more and more buildings must be torn down to
When applied to the activities of musicians, diminishing returns
tells us that each successive increase in the income of musicians will
have a smaller effect on the amount of creativity in music.
Diminishing returns is the first reason to reject the idea that any
use of music “should” be covered by copyright. There is
nothing to gain by trying to guarantee owners control of every
possible aspect of the use of music or to give them a financial stake
in every possible aftermarket. Extending copyright can only
“promote progress” up to a certain point. Further
extensions merely increase what the public pays to the owners for what
they will do anyway. Extending copyright beyond that point is
Those with a vested interest in extending copyright start the
discussion by claiming that copyright “should” be extended
as far as it can go. But the principle of diminishing returns renders
this claim implausible. So they fall back on the position that
copyright should be extended to maximize the rate of progress. But
this too is wrong, because it ignores the existence of other
trade-offs. Copyright imposes costs and burdens on the public, like
any other government project. The benefit may not be worth the
Government fills many important functions, but few would say that
any one of these functions should be expanded to maximize output. For
example, governments build roads, and this is very useful. But few
leaders would advocate building every road that could be built. Road
construction is expensive, and citizens have other uses for their
money. Too much concentration on building roads means that other
social and individual needs will be unmet.
The same considerations apply to individual decisions. By spending
more money, you can buy a bigger and fancier house. Most people would
prefer the more expensive house, all else being equal. But given
finite resources, at some point spending more on a house becomes a
poor allocation of them.
Copyright does not directly spend public funds, but it does impose
a cost—a loss of freedom—on every citizen. The wider the
scope of copyright, the more freedom we pay. We might prefer to
exercise some of our freedoms rather than trade them away. We must
judge any decision in copyright policy by comparing the benefits with
“Incentive” is the wrong concept
The idea of providing a monetary incentive for making music is
based on a misunderstanding. Musicians hope primarily for other kinds
of reward; they must. Very few musicians get rich from their music; a
talented person whose primary goal is wealth would seek it in other
In fact, psychological studies show that the desire for an
extrinsic reward (such as profit) generally hampers creative
activities such as writing music. The people who can do them well are
usually those who do them mostly for their own sake.
This is not to say that musicians don't care about being paid.
Most hope to make a living from music so they will be free to devote
their time to it. As long as they earn enough to live, they will make
music as best they can. We might wish them to earn somewhat more than
just enough, so they can live as well as most Americans. But to offer
them wealth beyond this gains the public little—it is a matter
of diminishing returns.
With this understanding, let's consider how a tax on DAT tape could
be designed to serve the intended purpose of copyright.
Who should get the funds
If the purpose of the DAT tax is to better reward musicians and
composers, then all the money collected should go to them—not
just 43 percent. The musicians and composers are the ones who truly
create the music. In principle, we could do without record companies
Record companies do provide a useful service: they distribute
prerecorded copies of music, usually of high quality. This service is
widely used, and will probably remain so. And it is right that the
purchasers of prerecorded copies should pay for this service. But
listeners making copies for themselves or their friends do not consume
this service; they use only the work of the musicians and composers.
The record companies contribute only incidentally and their role is
Dividing the funds
What share of the tax revenues should each musician or composer
get? The record company proposal would divide the money in proportion
to record sales.
It makes sense to distribute the funds based on how much that
musician's work is copied, more or less. But strict proportionality
is not the best apportionment. If each musician gets a share in
strict proportion to the amount of copying of his or her music, then a
large share will go to make a few superstars even richer than they are
now. This won't do much to promote musical culture or diversity.
We can promote music more effectively by making any one musician's
share of the tax revenues taper off as copies increase. For example,
we could calculate an “adjusted number of copies” which,
beyond a certain point, increases more slowly than the actual
The effect of tapering off will be to spread the money more widely,
supporting more musicians at an adequate standard of living. This
encourages diversity, which is what copyright is supposed to do.
The US government has already established a program to fund
diversity in the arts: the
However, NEA grants involve discretionary power, which makes them a
center for controversy, sometimes because a few members of the public
strongly dislike the work, and sometimes because hardly anyone
particularly likes it. Spreading out DAT tax revenues will also have
the effect of supporting less popular musicians. However, it will not
support musicians whose work nobody likes. In addition, since it
involves no discretion, no arbitrary decisions, there is little room
for objection on account of any particular case.
[I was later asked an interesting question: what organization would
“manage” the distribution of these funds. Since this is
tax money, a government agency should collect the tax and distribute
the funds. Private organizations should not be involved.]
Encouraging home copying
The record company proposal includes a requirement to make it
difficult for home listeners to make copies. Specifically, it
requires that consumer DAT machines refuse to copy a copy that was
made on a consumer DAT machine. The argument for this requirement is
based on the assumption that home copying is somehow unfair.
In the past, many people have considered it unfair, because it
reduced the income of musicians. The DAT tax makes this reason
obsolete. Once home copying does contribute to the income of
musicians, through the DAT tax, the reason to discourage home copying
Therefore, if a DAT tax is adopted, the ability to copy DAT tapes
should not be restricted. Home copying is more efficient than record
companies and record stores; music lovers should be encouraged to use
home copying as much as possible.
Measuring the use of each piece of music
Today, nearly all the recorded music in the United States is
purchased in record stores; home copying is but a small fraction.
This will probably remain true for a long time, because record stores
offer a place where a person can go to find a particular piece or to
browse a wide selection. While this remains true, we can usually
estimate the audience of a given piece fairly well by counting record
Eventually, home copying may become so widespread that estimating
its extent from sales figures may be unsatisfactory. This is already
unsatisfactory for musicians who distribute independently without the
help of record companies; and if any musicians need additional
support, these are the ones. We need another way to estimate usage of
any given piece, in order to distribute the tax funds.
We can make these estimates by survey. From time to time, survey
staff would ask randomly chosen members of the public to show what
copies they have made of copyrighted music. The citizens asked would
not be required to answer. But no penalty and no guilt would attach
to having made copies, so most people will be glad to participate.
Fans will hope to be chosen so that they can contribute to the count
for their favorite musical groups.
To make the survey more efficient and broader-based (and thus more
accurate), it could be automated. The survey bureau could mail
read-write memory cards to the chosen participants, who would connect
them momentarily to their DAT units and then mail them back. With
proper design, the survey bureau would have no way of knowing who had
sent in any particular card, and thus no information about who had
copied what, but they would still have an accurate total.
The record companies have proposed an excellent scheme for taxing
the public to increase their own income, but this isn't a legitimate
purpose of copyright. Through due attention to the ends of copyright
rather than past means, we can design a system which supports
musicians while giving citizens full freedom to copy music as they
What You Can Do
[This section is no longer applicable today; it is too late,
because the DAT tax bill was adopted in 1992—and DAT
recorders are obsolete nowadays. However, the same method can support
musicians and other artists in a world where sharing copies on the
Internet has been legalized.]
Record company lobbyists are working hard to pass their form of DAT
tax. There is little organized opposition, and little public debate.
Their bill has already been sent out of committee in the Senate.
This article proposes an alternative to the record company plan.
In order for this alternative, or any alternative, to have a chance,
we must first prevent the hasty adoption of the record company plan.
To help accomplish this, please write letters to:
Congressman Barney Frank
437 Cherry St
West Newton, MA 02165
United States Senate
Washington, DC 20510
House Subcommittee on Intellectual Property
House of Representatives
Washington, DC 20515
Urge Congress to reject the record company bill so that this and
other alternatives can be properly considered. It takes just a few
minutes to write a short letter, but in combination with other
people's letters it can do a great deal of good.
If you know any musicians, composers, or songwriters, give them
copies of this article. Many musicians prefer this alternative to the
record company tax plan, and they are strongly motivated to act on