A discussion of hybrid open source and proprietary licensing models.

[Marshall Van Alstyne <marshall AT MIT.EDU>]

Ryan, here are several thoughts based on the questions you raised. 
Apologies for a delay, I've been preparing a tenure package and that has
dominated all other tasks.    Your set of issues around
NonCommercial and Commercial licenses raise multiple follow up
ideas!

At 03:16 PM 4/17/2007, you wrote:
> Let’s say I make a derivative
> work using someone else work that’s under the “NonCommericial” license
> option and I want to commercialize it to make money.  Under this
> license I have to approach the original licensor to work out a commercial
> licensing agreement to establish payout percentages.  I run the
> risk, however, that the original licensor will not want to come to a fair
> and balanced assessment of each other’s contribution.  With this
> potential risk loaming in the future, I am less likely to even start a
> derivative work in the first place­there’s less initial incentive. 

I think you've nailed one of the biggest problems.  In fact, it
seems there are related two sub-problems.  One is hold-up from a
single owner -- in this case the (c) holder -- who will want to negotiate
for a chunk of value.  The other is the cleverness of the idea can
be lost if you try to negotiate and accidentally disclose it.

Either way, I agree that these 2 kinds of risk discourage a person from
making any kind of investment beforehand.

> What if, however, there was a
> New Licensing option­somewhere between ‘Attribution Non-commercial’ and
> ‘Attribution No Derivatives’­that stated that you are free to pursue
> commercial gain without consent, but if the derivative work marks money,
> the payout percentages are assessed by the community and not left up to
> the individuals and their lawyers to determine?  

One possibility that I like a lot is to use a default menu.  That
is, suppose the license has an offer of the following form:

If you make $x amount of money, then contribute y% of revenue to the
original work.
If you make $(x+d) amount of money, then contribute (y+n)% of revenue to
the original work.
etc.

Alternatively,

If you keep your piece closed for x years, then contribute y% of revenue
to the original work.
If you keep your piece closed for (x+d) years, then contribute (y+n)% of
revenue to the original work.
etc.

This set of default options lets the person writing the derivative work
choose based on a fairly clear set of terms.

> So in other words, under this
> agreement the assessment of value contribution is determined by the
> community and not the individual parties.  In this case I can have a
> little more confidence using someone else work initially, knowing that if
> my derivative work does indeed make money, that the community’s
> assessment will mostly likely grant a fair evaluation. 

This idea is pretty interesting.  I see at two possible
concerns.  (1) Is that there is still some uncertainty over what the
final % will be.  I agree it seems that the "community,"
however defined, is likely to be more neutral in assessing value but
there is still some lack of clarity beforehand. (2) Have you thought
about how large a community you need?  If it's very small, there may
be too few voters to establish a meaningful baseline so there might be a
need to build the community first.

>  
> What is the community in this case?  Well that could take many forms
> depending on the media and/or mode of distribution.  In the music
> industry, diehard fans could determine, through some type of ranking
> system, how each artist should be paid out.  If, however, asking the
> public to audit contribution becomes too overwhelming, there could be a
> pool of editors that get paid to make these assessments.   This
> system of editors, however, would have to be anonymous in a sense, to
> avoid the possibility of bribery or kickbacks.  An additional
> safeguarding mechanism could be utilized were the editors in turn rank
> each other.  Editors that show an unusual high level of bias would
> be flagged by the community of editors and thus facilitate a higher level
> of consistency.

Overall, this idea sounds quite useful.  A good reputation system
can be used to "rate the rater" and help ensure integrity in
the review process.

This "information revelation" mechanism looks like it should
require a subsidy i.e. some means of compensating the raters for their
time and effort.  If done well, it could be covered out of the
revenues generated by the pool of derivative works themselves.

Another form of information disclosure mechanism could be to have people
who actually download the derivative work themselves state what % of
value the new piece adds.  This is harder to "fake" since
someone must actually pay for the good (so if you're going to stuff a
ballot box it's going to be costly).  It's also unbiased because the
person doing the buying is not independent of the person doing the
rating.

>  
> Although I’m sure there’s ways to establish a fair assessment of value,
> this licensing system would require a transparency in accounting as
> well­such that contributors would be assured that all the revenue is
> accounted for.  Transparent accounting might be a tall order, but if
> artists demand it, producers will find a way too oblige. 

This is also an excellent point.  It has long been an established
tenet of economics that more complete information leads to greater
efficiency in production.  Such a system should be as transparent as
possible.

Note also that it should also align incentives the right way e.g. no one
should have a reason to misrepresent their rating, and the raters' votes
should correspond closely to the consumers' reasons for buying.

>  
> Anyways, Just a little food for thought.  Let me know if there’s any
> major hurdles I’m not considering in implementing a system like this…
> just thinking out loud here.

Personally, I think the questions you raise are very interesting. 
Plenty of thought munchies here! :-)

>  
> Is there any current discussion out there that addresses this idea? 
> I could not find any. If you could forward, I’d be much appreciative.

There is some good discussion on the Free Software Business (fsb)
discussion group but aside from that, I've found very little on this
topic.  Thanks for raising some interesting issues.

MVA

>  
>  
> Thanks  
>  
>  
>  
>  
>  
>  
>  
> _______________________________________________
> Cc-bizcom mailing list
> Cc-bizcom AT lists.ibiblio.org
> http://lists.ibiblio.org/mailman/listinfo/cc-bizcom