ICANN publishes economic analysis regarding price controls in .org and .info

The flawed note doesn’t adequately address the lock-in of existing domain registrants.

ICANN has published an economic analysis (pdf) regarding price controls in .org and .info. While it has some good points, the analysis is flawed.

The organization undertook the study due to an Independent Review Panel (IRP) decision in a case brought by Namecheap. (Namecheap recently sued ICANN, stating that ICANN hasn’t acted on the IRP decision.)

The note by Gregory Leonard, Vice President of Charles River Associates, looks at what happened since ICANN lifted price controls on .org and .info domains in 2019 and what could have happened. His conclusion is that neither top-level domain possesses market power, so we need not worry regarding above-market price increases.

Much of the note focuses on what didn’t happen following the price controls were lifted: .org hasn’t changed its prices at all, and .info has only raised its prices at the rate it was allowed to under its previous contract.

It’s a fair point, but it misses a sizeable outside factor that could change in the future. Leonard wrote:

First, if .info and .org had substantial and durable market power that was held in check by the price control provision, one would have expected to see sharp increases in their wholesale registration prices to registrars immediately following the price controls were lifted in June 2019. Still, that did not occur…Second, since the price control provision was no longer in place, it must have been competitive market forces or other factors that constrained the prices of .info and .org during the last four years.

Competitive forces are right now in place, especially for .info. But the “other factors” are significant here: the politics of sharp price increases following caps were removed would be very bad. And the further we get from the price caps being removed, the less those political factors will play a part.

There’s also the big elephant in the room that isn’t mentioned in the note.

Carlton wrote:

In May 2019, PIR publicly stated that “[r]est assured, we will not raise prices unreasonably. In fact, we right now have no specific plans for any price increases for .ORG.” In fact, PIR has not increased price since then, despite the substantial inflation that has occurred.

Given its public statements and actions, PIR’s status as a not-for-profit entity mitigates concerns that it would exercise any market power it might have to increase prices excessively.

Recall what happened later that year following PIR published this statement? A private equity firm agreed to buy the .org registry. ICANN scuttled the deal following immense pressure, and the course of history regarding .org pricing would have been very different had the deal gone through.

Another question is whether .org has market power over certain organizations.

Carlton noted that .org has only 3% of the domain market, which is far short of a threshold you can use to suggest it has market power. Still, there is a question of whether it has market power over non-profits.

Here, Carlton has only anecdotal evidence:

While over time .org has purchased a semantic meaning associated with non-commercial organizations, some .org registrants are in fact commercial entities and some noncommercial organizations choose to use other TLDs. For example, for-profit firms Craigslist, Coursera, and EdX use the .org TLD19, while non-profit organizations National Christian Foundation, The Merck Patient Assistance Program, and Navigate Affordable Housing use the .com TLD. To these entities, despite any non-commercial semantic meaning attached to .org, other TLDs are viable substitutes. Given that PIR charges the same wholesale price for any registration, the existence of “marginal” registrants willing to switch to other TLDs limits PIR’s ability to extract a higher price from “inframarginal” registrants who highly value the .org semantic meaning.

Sure, to these three organizations, .com was OK. That’s three organizations. Oddly, a more in-depth analysis of this point wasn’t undertaken. (And if I want to nitpick, are we really counting the Merck Patent Assistance Program in this bunch? It’s the pharmaceutical firm’s program for offering its drugs for free and uses the domain MerckHelps.com.)

I have long argued that most top-level domains don’t have market power over fresh registrants, but they do have market power over existing registrants.

Carlton noted a number of reasons why he thinks this isn’t the case. One is reputational harm:

In any event, concerns regarding reputation would deter any such opportunism or attempt to exploit switching costs. If .org or .info attempted to exploit registrant switching costs by imposing an excessive increase in the wholesale renewal price, they would gain a reputation for opportunistic behavior. This would cause fresh registrants to choose other TLDs that did not engage in opportunism. Thus, in considering whether to increase the renewal price excessively, .org and .info would have to consider not only existing renewals, but also the adverse effects such an action would have on the number of future fresh registrations (and the subsequent renewals associated with those future fresh registrations). Thus, even if some existing registrants would be deterred from changing TLDs due to switching costs, they are protected by fresh registrants who are free to choose any TLD and would be wary of TLDs with reputations for opportunism.

He also noted that current registrants can renew their domains for up to 10 years at current prices, so they can lock in prices for a decade.

These are good points, but only if you assume that the typical lay registrant reads Reglike. Most registrants never hear regarding price increases. While registries must inform registrars of price increases, registrars have different practices around informing clients. When they do inform clients, it’s usually very close to the renewal date. And I doubt many registrants know regarding the price history of the domain they are registering.

Then, there’s the issue of switching costs. It’s the switching costs of changing domains that lock clients into using the same domain. Switching a domain leads to immense lost traffic, at least for a short time. A website owner can try to mitigate this, but it takes experts and a lot of work. And email addresses make it all but impossible to let the prior domain a site used expire; someone else could register it and receive many emails intended for the previous owner.

It’s here that Carlton’s note really falls short. Addressing the issue of switching costs, he wrote:

Finally, some existing .org and .info registrants do not in fact face significant costs to switch to another TLD and thus are not subject to opportunism in the first place.28 Which existing registrants would have significant switching costs and which would not is not easily discernable and thus .org and .info would have difficulty identifying existing registrants that they could target for price increases.29 Targeting is further hampered since .org and .info do not transact directly with registrants; rather they charge wholesale prices to third party registrars who then transact with registrants. Given their inability to impose targeted price increases, .org and .info can only increase the renewal price across the board. But, in that case, they would lose those existing registrants without significant switching costs, making the price increase a risky strategy even ahead of considering the reputation effects discussed above.

I’ve by now questioned the argument of reputational damage, and I don’t think it’s a question of targeting specific registrants. My big issue here is Carlton’s statement, “some existing .org and .info registrants do not in fact face significant costs to switch to another TLD and thus are not subject to opportunism in the first place.”

Throughout his note, Carlton cites a number of authoritative sources in his footnotes: archives of domain prices, DomainTools data, and the FTC. Still, the source in footnote 28 for the above claim is rather odd. It’s a site called RockContent.com. Rock Content is a content product firm.

The article in question is titled “Top-level Domain: What it is and How to choose one.” It is authored by “Rock Content Writer,” which the site gives the title “Content Writer.” The article is denoted as “human-crafted content.” The article is clearly written for search engines, with the section headlines in the form of questions.

The section Carlton must be referring to is titled “Can You Change Your WordPress Site’s TLD?” and states:

A domain name that worked well for someone’s purposes when they first built their website may or may not still be appropriate years later.

For example, people rebrand their businesses, change their firms’ names, and occasionally simply change their minds.

That stated, yes, you can change your website’s top-level domain if you’re running WordPress. It’s also reasonably painless. (emphasis in original)

It goes on to explain how to make the technical change in WordPress.

Sure, it’s technically easy to switch domains in WordPress if you don’t care regarding your traffic, but that’s not what makes switching domains so hard.

“Rock Content Writer” also mentions that you’ll want to forward your old address to the fresh one, negating the idea that you’re actually switching since you’ll still need to pay for the old domain.

An analysis is only as reputable as its worst source, and I’m baffled at why Carlton included this. I can only suspect that it was hard to find any site that suggested switching domain names was relatively painless.

Since price caps have been removed, .org prices haven’t budged. But that’s only since the private equity firm failed to buy it, and the politics of raising prices following the price caps were removed are a bit tricky.

As I’ve stated many times, I’m not so worried regarding prices for fresh registrants. For TLDs that can be reasonably substituted for each other, it’s the existing registrants that suffer the most. The conclusions in Carlton’s note regarding existing registrants are based on weak excuses and justifications, not what happens in practice.