ICANN Reform - a personal view

Note: This is not the view of any body, organization or entity that I sometimes represent. It is my personal attempt to organize thoughts that can form the basis of saying something about how ICANN should be organized.

This page has changed over time. See the end of the note for a changelog.

What ICANN was designed to do

ICANN, as designed, was supposed to carry out a few tasks:

The clerical part of these functions can occupy a full-time person. Making sure the information about those changes and modifications are visible to the world at large throuh a web service can occupy another.

The rest of ICANN is concerned with one matter only:

Who gives those two people their instructions?

What ICANN was designed to prevent

There are a few disaster scenarios that ICANN was created in order to make sure they did NOT happen. They deserve mention.

Again, these are a matter of looking at who gives instructions - but the instructions may have to be given to people outside ICANN's direct control - someone apart from the 2 people mentioned above.

Other things that ICANN might as well do....?

A number of things concerning the DNS also cry out for some "serious management":

The current ICANN has attempted to chart a course between the scylla of doing nothing and the charybdis of doing everything; all its detractors seem to agree that the course is not right, but many of them disagree about which way the rudder should be tilted.

The point to be brought away is not that ICANN has to go to an extreme one way or the other. It is that there is a long, slippery slope here, and that the required size of ICANN is a direct function of how far down the slope we decide to slide.

1) Note to armchair lawyers: I know, the UDRP is not an arbitration mechanism in the legal sense of the word, and should not be (IMHO). But I could not find a better word in my English vocabulary. 

Appropriate and inappropriate control methods

There are a few well known methods for getting people to do the Right Thing, and not to do the Wrong Thing.

Some control methods that have been tried, but are probably inappropriate:

Some things to be avoided:

The Stuart Lynn proposal

This proposal envisions.

Boards come in multiple fashions - one fashion is the rubber-stamp board, which accepts the officers' policy proposals; another is an activist board, which actually tries to set organizational direction.
The desired type is not stated here, but is presumably "activist". (The alternate is that the staff decides, and the board approves...something that may not be appropriate....)

 The interesting thing about this proposal is its deep detachment from the concerns listed above. It appears orthogonal.

It tries to avoid a couple of the traps - in particular unscaleable "have to listen to everyone before deciding" traps. But it does not seem to do much to avoid the problems with inappropriate decisions - this is left to the wisdom of the Board, whose selection processes are left carefully unspecified.

A majority of 8 trustees can (probably - bylaws vary) force anything. This means that if the nomcom comes up with 5 tech-heads, and the governments come up with 5 social-engineering types, they can easily override the advisory committee representatives, who presumably represent the "business of providing registratoin services" - which are dominant economic actors in the current ICANN.

The various proposals for elected representatives

The big problem with most proposals for elected representatives is this:

If a self-selected group of 20 persons claiming to represent 6 billion people is undemocratic.....what makes a self-selected group of 20.000 people claiming to represent 6 billion people more democratic??????

And if not so - what criteria can we use to select the group that will be allowed to select the group that makes the decision?

Some proposals - like restrictions to domain name owners - may cause curious results.

A hundred years ago (up to 1884 in Norway, for instance), it was common for "democracies" to allow only landowners to vote. The workaround was to buy tracts of completely useless land (marshes, for instance), and parcel it out to those who you wanted to have the vote; those were called "marsh men"..... with domain names and the Internet, fabrication is far easier.

A compromise that roughly translates to "prove that you can receive postal mail" was tried in the ICANN At Large election; the experiences from that shows, among other things, a terrifyingly high failure rate in the postal system.

We can thus group voting proposals into three groups:

There seems to be no easy way out of this. Either we accept decisions by people who cannot claim to represent "we the people", we leverage existing governments and accept that we have no more control over the outcome than we have over UN-based organizations like ITU, or we give up and go home.

An alternate approach: A smaller ICANN

NOTE: In the feedback on the previous version of this note, this paragraph was the one that attracted the most comments. In fact, very few people commented on the rest. Some comments are reflected under "weaknesses" below - my main reason for putting this in is to try to make people think outside the existing axis of "current ICANN", "ICANN + At Large", "Stuart Lynn's ICANN". People who only criticize and have no alternative should shut up; thus, I wanted to try floating a proposal.

This is "trying to think differently", not a complete proposal. Food for thought....

Keep the 2 staff members actually updating the files.

Add four more staff members who keep track of public input into the process, function as secretariats for working groups and so on. They do not make decisions or write proposals.

Add a triumvirate of directors with veto power over any proposal:

Add supporting directors if required. Advisory councils, if required. The GAC may be required to select the government representative.

Add working groups, with members selected by the directors. Each WG to have a public mailing list for discussion, in addition to its own work. No WG member gets paid.

Sign MOUs, not contracts.

Total cost: A bit lower. Total income: Unknown.

Ability to get into trouble: Quite a bit smaller.

Weaknesses of this plan

There are two glaring points open:

It is also still unclear what sources of income this organization has or how many lawyers it will require to protect itself against liability issues - and this little sketch says NOTHING about how to decide policies for new TLDs.

And, of course, we still don't know how to conduct business on open mailing lists in a climate of mutual hostility and distrust.

These things are left as an exercise for the reader....

Addendum: Changelog

To record the history of the page, and why some people may have different readings of what was here...