New gTLDs — auctions
I'm looking for comments so that I can understand the arguments about auctions better.
Concerns are often expressed about auctioning off new gTLDs — strong concerns — because of worries about what will be auctioned and what process will be followed to get to the point of auctioning.
Same concerns with lotteries (which I suppose would be like auctions except without a money element).
I know that the OECD made a proposal favoring auctions here and Profs. Milton Mueller and Lee McKnight made a proposal here. CircleID ran some articles here. Are there other materials on one side or another of this specific issue? What are the best arguments in this area?
Thanks for comments.
Comments
20 Responses to “New gTLDs — auctions”
Got something to say?
My concerns around auctions all center on the idea that the DNS is a resource for the entire world and money is not equally distributed throughout this world. If new TLDs are distributed solely by auction, you're virtually guaranteeing (a) that all registries will be operated by companies based in the developed world, and (b) the developing world will continue to be underserved by TLD registries and meaningful TLD strings. So I think “affirmative action”-style considerations or the “maximization of the social good” ought to be taken into the consideration as well (with the full understanding that this takes us out of bright-line, objective tests for TLD selection).
I also wonder about groups like IODesign (with which I'm not affiliated) who proposed something years ago. They'd likely get outbid for .WEB. Is that fair? I'm sure Google, with it's local service and Google Earth software, might like .GEO and could outbid SRI, which proposed it in 2000. How do we deal with those issues? These are problems that can be solved, I'm sure, but you are going to have circumstances in which someone who is not the highest bigger claims to have a superior right to the highest bidder in the idea or text string.
– Bret
I have to agree with Bret (and thank him for bringing us up). We proposed .Web in 1995, were asked by Jon Postel to build out the registry and prove the concept in 1996, waited until the process wound through three administrations until 2000 when we pay $50,000 to apply to ICANN and were told, in no uncertain terms, “You're not turned-down, you're just not approved yet. Next round.”
So we're still waiting, some ten years later.
Issues around an auction notwithstanding, to hold an auction and require those who paid the $50,000 application fee in year 2000 and are still waiting to compete would hardly be fair.
Of the applicants in 2000, there are some 40 or so remaining. I suspect that of those, only 10 still want in. That's a guess. It might be 15. Let's address those first and “clear the books” before we create a whole new process.
Thanks, Chris, for reminding us about the 2000 round.
It's an interesting thought that there might be only 10-15 of that round who are still interested. Is there any way to get those people to comment in the GNSO process? I know, I know, you've been commenting for a long time. But it seems to me that it would be helpful to know who is still interested.
Thanks, Bret.
not easy. Someone else pointed me as well to these materials:
1. Please see “An Economic Analysis of Domain Name
Policy” by Karl M. Manheim and Lawrence B. Solum,
starting around page 103. The full document may be
downloaded from this page:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=410640
2. Solum debates Auerbach at
http://lsolum.blogspot.com/archives/2003_04_01_lsolum_archive.html#200102145
3. The Cross-Constituency Position (BC-IPC-ISPCP) in
opposition to auctions at
http://www.bizconst.org/positions/WPnewgTLDsfinal.doc
Indeed, it probably would be. There are two reasons that just asking might not have results. The first, and it's simply speculation, is that some of the applicants might be hesitant to speak up for fear that someone else will see that the TLD is still desired and, sensing a chance to inject themself into the process, might speak up and claim it as well. I suppose that's one benefit to being continuously-involved: nobody has had a chance to claim we've given up.
The second reason is more probable, I think - most applicants don't want to give the issue much thought until there is a process in place about which they can feel comfortable, for fear of this being yet another instance of Lucy inviting Charlie Brown to take a run at the football.
That said, if the ICANN board wants to go down this path, I would gladly volunteer to contact each and every applicant personally, speak with them, and settle the speculation, if it would help. Some kind of indication from ICANN that we applicants from 2000 have not been forgotten (at best) or dismissed (at worst) would likely get things moving.
And maybe we'll get to kick that football this time.
The proposals favouring auctions fail to understand that commercial exploitation is not necessarily the thing to be maximized in [all] new TLDs. For example, would you give .redcross to a firm in Japan or Canada rather than to the Red Cross, just because the former can put more money on the table?
In case strings have strong semantic meanings, people expect them to be managed by a “legitimate” entity in a broad and inclusive way, rather than to be given to someone to squeeze as much money as possible out of them. What is “legitimate” is another matter, but certainly it is not money that creates legitimacy to steward a global public resource such as a gTLD.
I forgot to add that the ALAC is collecting comments on new gTLD processes through a wiki interface at http://www.icannwiki.org/ALAC_on_New_TLDs - you might want to examine that as well, if you like. Anyone's contributions are welcome.
Thanks for those links. Both Solum and Manheim seem to support “mixed regimes” of auctions+(something else) to distribute TLD resources more equitably worldwide. I too want to read more about this. Unlike U.S. frequency auctions, which take place purely in one market and under one currency, TLD auctions will span conventional markets and currencies. Do we have a precedent for auctioning off a worldwide resource? I'd be very interested in other areas.
– Bret
Well, if you want us to talk about lotteries, you should use a different header.
Lotteries have a much more egalitarian prospect to them. You can structure a lottery so that only qualified entities can enter: ones that can show that they can run a TLD with at least six geographically distributed nameservers that have enough bandwidth and so on.
So, what do you put up for lottery? Ask the people who want them. Qualify at least a hundred organizations that want to run a new TLD. Charge them some entry fee that is used only to do the vetting; it should only be around $1000. Quite frankly, you might get a thousand organizations.
Ask each qualified organization to list of up to ten TLDs for which they would want to own the monopoly. Hold lotteries for the 40 TLDs with the highest number of desiring potential owners. Only let those who said they were interested (as one of their ten choices) into the lottery for a particular TLD; you don't want the Red Cross, who really only wanted .redcross and .emergency, to be in the lottery for .pr0n or .potato or .pólvora.
Also have lotteries for 10 names in which there were exactly two organizations that wanted them. Each of these lotteries would be a coin toss. This would be a good way to measure if lesser-desired TLDs are actually also worthwhile.
Unless there is a terrible result (that is, more terrible than typical ICANN blunders), repeat the process every year until there is no interest.
Much though I hate to disagree with my buddy Paul, lotteries to allocate “desirable” names are unlikely to produce a good result. Turn back the clock 25 years and you may remember that the FCC had lotteries to hand out cellular phone licenses. For the B licenses there was a lottery among all the phone companies in the territory, while for the A licenses there was a lottery for anyone who met the qualifications, which was a whole lot of people since one could buy a kit of qualification materials. What invariably happened was that unless the lottery winner happened to be a large telecom company that could afford to build a cell system, the winner immediately sold the license to the highest bidder, thereby turning the lottery into an auction where someone else got to keep the money. I see no way to keep that from happening here, since a no-sale rule will be both unenforcable and have perverse consequences (”oh, that wasn't a sale, that was a merger”)
I do have a lot of sympathy for the hybrid model of auctions to allocate commercially desirable names and directed allocation of socially desirable ones. If there's competition for a socially desirable one, pull out the thesaurus and make everyone a winner. If the applicants don't like that, they can merge their applications at any time.
So, what's the problem with that? The Internet still gets 50 new TLDs a year, ICANN doesn't get fatter, and the folks who register in them get to learn that there are risks in life, particularly when you're trying to get something you don't understand for cheap. There are no significant technical downsides. I don't hope that the lottery-becomes-auction happens, but if it is inevitable, why not let it happen?
The problem with lottery as auction is that it gives away the value of the domains, making the lottery tickets excessively like the kind you buy at the 7-11. One question I haven't seen addressed in auction proposals is what happens to the money. It could just go into ICANN's budget, but it seems to me the last thing the world needs is more money flowing into ICANN with no accountability. If it were up to me I'd dedicate the auction lucre to an outreach fund used for virtuous purposes like paying for plane tickets for network users in LDCs to go to the wacky places where ICANN has its meetings.
I am not sure how many of the original 47 or so applicants from 2000 round are still waiting to bid again but it is worth noting that at least 5 of the current sTLD round were previous applicants, even if in some cases as successors to the original applicants from 2000, who had waited diligently for 4 or 5 years .post,.travel,.mobi .Tel (telnic) and , of course, .xxx.
Like Chris, after these,my guess at the number of remaining would-be applicants is likely to be around 10.
It's interesting that the numbers of waiting and anxious applicants may be relatively low. If we set up a process that reliably created 10 (or fewer!) gTLDs a year, we might soon exhaust the energy for new applicants and things could go on peacefully at a low level for years. I guess there's no way to test whether that's actually true, but it would be helpful to have this information as part of the gTLD GNSO process.
Susan
If there's a private auction that follows the lottery, and that is only open to accredited registries (who meet certain requirements, like registrars have to), why would ICANN necessarily participate in the funds flowing from the auction (above, say, some set fee)? Just wondering. This is like the secondary market in domain names question.
These questions are posed in my individual capacity, and not on behalf of anyone else. I want to make sure everyone understands that I'm looking for information here. Susan
As I said above, I don't think many applicants from 2000 would bother to speak up until there's a process in place.
Or, alternatively, if ICANN were to ask. Put out a simple call - “How many applicants still-pending from 2000 still wish to have their applications considered?”
That simple.
Paul and I worked up some longer comments and posted them in
my blog and his blog.
Thanks - reading with interest.
My own comments on how to allocate new TLDs is found in my blog
entry of Jan 26, 2006 -
How Top Level Domains (TLDs) Should Be Allocated
The main difference in what I propose from what has been suggested
by Paul H. and John L. are these:
- I propose the allocation of “slots” rather than names - it being
none of ICANN's business what character string is selected as the name
for the slot, as long as it is unique - ICANN simply is not an
appropriate body to enact, much less enforce a home-brew law of
trademark over domain names.
- I don't require that the applicant have worldwide facilities,
rather they should merely have what is required to meet the
anticipated demand. (I realize that setting up a worldwide ring of
servers isn't really that expensive, something less than perhaps
$100,000 per server to cover hardware, installation, a years worth of
bandwidth, etc, but if a new TLD anticipates only a few thousand
queries a day then why should it have to pay to build an
infrastructure on par with that supporting .com?)
Anyway, take a look at my note
How Top Level Domains (TLDs) Should Be Allocated
thanks so much for the pointer, Karl.