Thursday, 4 December 2003
Mechanics of the CAN-SPAM registry
There have been many questions about how a do-not-spam registry should be implemented. This proposal suggests a regime for funding for the registry and the highest level logical operation of its database. My plan would allow consumers to choose (through market forces) an opt-in system while still adhering to the overall opt-out structure of the CAN-SPAM Act. For that reason, I believe it solves some of the nagging First Amendment problems that come with a government-madated opt-in system.
If you have not already seen my summary of the CAN-SPAM Act, I suggest you check it out before reading this.
The registry should not necessarily be funded by taxes, because that would require people without email accounts to share the burden a system that carries no direct benefit for them. ISPs stand to benefit the most (in financial terms, at least), because a successful registry will reduce their bandwidth and other costs substantially. I would hesitate to levy mandatory fees on ISPs because they would look too much like the fees imposed on bell companies to fund rural telephone lines and the 911 system. I would prefer to leave ISPs as unregulated as possible while still having them share in the cost of the registry. I would not be averse to paying a few dollars to get myself into the registry, but ISPs should not have a free ride while consumers fund the entire thing.
My proposal is to make ISPs intermediaries between the FTC, which would manage the registry, and consumers, who will have ultimate control over the status of their addresses.
First, charge ISPs a monthly fee for having their domains listed in the registry. This fee would be assessed according to the number of email addresses at each domain, and those addresses would be automatically opted out of receiving spam. If a user wants to change that status, he would ask his ISP, which would relay the request to the FTC. An ISP would be charged a small transaction fee for each username it changes from its default status, as an incentive to "guess" what most customers will prefer. Individuals whose ISPs do not list their domains in the registry would have the option of opting out individually, paying the same transaction fee directly to the FTC. This option would be available to anyone in the U.S. with an email address, even those who maintain email addresses at their own personal domains and do not use an email address provided by an ISP.
To keep the size of the database's output manageable, it would need to spit out three separate lists. The first list would contain all the domains listed in the registry. The second list would contain all the individual email addresses that have requested opt-out status. Any email address covered by these two lists would be off-limits to spam. The final list would contain the addresses of ISP customers who have decided to switch away from their ISPs' default opt-out status. Addresses on list #3 are fair game for spam.
My plan would require some taxpayer funding for startup costs, although these could be recouped over the first few years by charging slightly higher fees during that time. After that, the monthly fees for listing domains and the per-user transaction fees would cover operational costs. ISPs will inevitably pass some of those costs on to consumers. However, there is harsh competition among ISPs, so the market would quickly allocate those costs efficiently. I believe this is more equitable than a program funded wholly by taxes. The recently-implemented do-not-call registry is funded by taxes because telephone penetration is nearly 100% in this country. However, many fewer people have email accounts than telephones, so full funding by tax dollars seems unfair to me.
The system is national in scope, so it will be large enough that the fees per domain and and per user can be small. Only a few indigent people and organizations could legitimately complain about the cost, and these might be exempted from paying fees. To start, the exemptions might be granted to educational institutions, 501(c)(3) organizations, and individuals below the poverty line. I have little experience in this area of social policy, so I would leave it to others to work out those details.
This structure would allow the market to demonstrate once and for all whether the public really favors an opt-in or an opt-out system. Many people have speculated on this question, but the truth is that nobody knows for sure. We may see a surge of subscriber defections away from ISPs that choose to be listed, or we may see a surge of individuals listing their own addresses. The point is that consumers, not the government and not spammers, would finally have direct control over the marketing they receive.