The idea behind the first CDA was simple enough. Cyberspace doesn't provide an easy means to keep porn away from kids. So Congress passed a law that said that anyone uttering "indecent" speech in cyberspace must make sure that kids couldn't hear. The Supreme Court found this statute both too burdensome and too broad: the category "indecent" covered too much speech; and the burden of complying with the statute (through adult IDs or credit cards) was too great.

This time around, Congress has narrowed the class of speech that it is regulating (to cover speech "harmful to minors") and it has narrowed the range of speakers covered (only commercial providers on the web). But it has kept the same mechanisms for verifying that an adult is an adult. Under CDA 2.0, one must either purchase an adult ID (most likely, one that will work on only some adult sites) or one must present a credit card to an adult speech vendor, on the assumption that all adults—and only adults—carry credit cards.

Both methods for segregating kids are burdensome. Many companies offer adult Ids, and different sites may contract with different companies, so to have access to any particular adult site, one would have to carry many IDs. And while consumer fears about using credit cards on the Internet are generally overblown, certainly the last people in the world that one should give credit card numbers to are pornographers.

The constitutional test, however, is whether the methods are "too burdensome," and that begs the question, "relative to what?" And here we believe a clear alternative exists that Congress has not considered, but which would satisfy its objectives at a far lower cost.

The idea is to tag kids rather than adults. Rather than requiring that adults carry IDs, enable parents and schools to install "kid-IDs" and require sites to block out such IDs. The IDs we have in mind are not adult passwords, entered voluntarily when someone accesses a site. The IDs that we have in mind are "digital certificates" that would be bound to a particular user on a particular browser, and identify the user as a child. Only the fact that a kid is a kid would be transmitted; not his name, or age, or address. Adult sites on this model could simply check for such an ID; if one was present, the user would be denied access.

There are many details in this proposal that would need to be filled in. And there are features of the technology that would need better development. The system must assure that the tag doesn't attract predators, though again, the tag would not reveal any identifying information other than age. There are a number of advantages to this approach. One is that the vast majority of users would not be burdened by the regulation. Instead the regulation would target only those who need protection. The tag might also be used for another child-protective proposal: banning the gathering of personal data from kids by commercial sites. The government, to minimize the burden on parents, could subsidize or even give away such certificates.

Kid IDs are much easier to implement, and far less speech-restrictive, than some of the content filtering software that is already being developed. Instead, each site itself would decide whether its material was "harmful to minors," and then if it so decided, it would take steps to block those who came marked as a kid. The tags would encourage parental control rather than decisions made by the government or third parties. Tagging kids rather than parents also removes the specter of a government-compelled "national ID card" for all citizens. Finally, this law would be relatively easy to enforce. The FBI need only browse with a kid-certificate, and then see whether they could nonetheless access information that was harmful to minors.

As we said, we believe this proposal would be less burdensome than the one that Congress is now considering. To be certain would require study, and careful review. But Congress apparently has no time properly to consider these alternatives. Rather than putting off a decision on CDA 2.0 until it has a better sense of the facts, Congress is rushing to enact legislation that—especially in light of this alternative—should be held unconstitutional.

Both those (like us) who oppose any new legislation, and those who want legislation to address the perceived problem of content harmful to minors, have a strong interest in making sure that the law that passes does the least possible damage to free speech values. For speech advocates, the interest is obvious. For advocates of regulation, the Supreme Court decision last year in ACLU v. Reno should make it clear that haste will get you nowhere.