May 22, 2005 to May 29, 2005
Regulating the Internet
Blogging, Spam, and the Taxation of Internet Transactions'Posner
The rise of the Internet has created a host of social, economic, political, and regulatory issues, a few of which we address in this week's postings. (Another, at present under consideration by the Supreme Court in the Grokster case, is copyright infringement by means of file sharing.) One that is naturally near to our heart concerns proposals to regulate blogging, either formally or through voluntary adoption of ethical standards. Many blogs are electronic counterparts of newspapers, magazines, and other advertiser-supported "mainstream media," and the argument is that since the mainstream media have adopted ethical standards concerning such matters as reliance on anonymous sources and retraction of errors (with electronic media such as television stations subject to formal regulation), so should bloggers.
The idea of parity among media is attractive, since exempting the producer of a close substitute of a taxed or regulated product from taxation or regulation tends to promote inefficiency; the exemption operates as a subsidy. The parity issue is starkly presented by the question whether to tax Internet transactions, discussed below. With respect to blogs, the contention is that exempting them from ethical or other informal (or formal) regulation subsidizes their competition with the mainstream media. Not that they are or would be totally exempt from controls over content; there is no legal exemption for a blog that defames someone, invades the person's right of privacy, exhibits child pornography, reveals classified information, infringes copyright, or otherwise violates generally applicable laws, though in many cases the bloggers will not have sufficient resources to make suing them for money damages an attractive course of action. But there is no compulsion on bloggers to comply with the ethical standards applicable to the conventional media. Moreover, they face less market pressure to comply with ethical standards than the conventional media, because they generally are not supported by advertising revenues (though this is changing) and thus don't have to worry about offending advertisers'or for that matter viewers, since bloggers do not charge for visiting their sites.
Nevertheless I think this "exemption" of blogging from the ethical standards applicable to the mainstream media makes good economic sense because of economic and technological differences between those media and the "blogosphere." There are vastly more bloggers than there are newspapers, magazines, and radio and television stations. In fact, there are some 10 million blogs, though most of them are personal rather than oriented to news and opinion'yet there are doubtless thousands of the latter. The large number is related to the fact that there is no set-up or operating expense for blogging except telecommunications line charges and, of course, the opportunity cost of the blogger's time. As a result not only of the large number or blogs but of the speed of transmission and the fact that many bloggers are far more specialized than journalists, the blogosphere pools information with extraordinary completeness and rapidity, in a speeded-up version of Friedrich Hayek's well-attested model of how the economic market pools information efficiently despite its decentralized character, its lack of a master coordinator. The blogosphere is a larger and faster-paced network than even the global marketplace. This means that errors in a blog, to the extent they concern matters of public interest and concern, are corrected almost instantaneously. Not only by other bloggers, but by the bloggers" readers, who post comments to the blog; the comments become a further part of the information network.
There is also greater political diversity in the blogosphere than in the mainstream media, because a conventional journalist's career appeals disproportionately to liberals
The self-correcting machinery of the blogosphere is more efficient than the internal fact-checking departments of conventional media enterprises. This is not only because many more people (not only the bloggers, but also, as I have just noted, their audience, which can communicate with them instanteously by means of the comment feature that most blogs enable) are watching out for mistakes; it is also because corrections are disseminated virtually instantaneously throughout the network. In contrast, even when the mainstream media catch mistakes, it may, especially in the case of the print media, take days or weeks to communicate a retraction to the public. The process is especially deficient in the case of newspaper retractions, which are printed inconspicuously and, in all likelihood, rarely read.
Given these differences between blogging and the mainstream media, the case for imposing ethical standards on bloggers is weak. Moreover, there is no way in which thousands or millions of bloggers could agree to adhere to a set of standards, whereas such (benign) collusion may be feasible in a highly concentrated media industry, such as the newspaper industry.
A criticism of blogging that has some merit is that there is less advance filtering than in the case of the mainstream media, which because of fear of offending advertisers (more precisely, the advertisers" customers and so derivatively the advertiers) engage in a degree of self-censorship, much of it desirable'censoring out hate speech, wild rumors, and fantastic conjectures and misinformation. Blogging is less constrained and there is a valid concern that it encourages and reinforces antisocial tendencies. The problem is not limited to blogging but includes other uses of the Internet, such as chat rooms. It is not, however, a problem amenable to solution or even alleviation by a program of promoting the adoption of voluntary ethical standards, and there are practical as well as legal obstacles to official censorship. There may also be an actual social value in allowing antisocial elements to blow off stream'and by doing so to identify themselves to law enforcement and intelligence agencies, which can monitor blogs and chat rooms for dangerous movements. What is more, self-censorship motivated simply by a concern with avoiding offense may impair the marketplace of ideas by excluding heterodox ideas and perpetuating comfortable myths.
A genuine externality created by the Internet that may call for some kind of government regulation is the phenomenon of "spam." Spam is advertising (broadly understood to include the various scams that seem to occupy a significant segment of spam space) that is emailed to people's computers. Because the cost of spamming is very low, enormous volumes of spam are emailed despite a very low response rate, the explanation being that the lower the cost of advertising, the lower the break-even response rate is. Something like 75 percent of all email is spam, and this figure is expected to rise to 95 percent by next year in the absence of regulation.
Spam imposes costs (without offsetting benefits) of two kinds. First, most of it is of no interest whatsoever to recipients and some of it is downright offensive; hence receipt imposes a cost. (An earlier example of the first cost was faxed advertisements, which cost the recipient the paper the fax was printed on.) That in itself does not differentiate spam sharply from many other forms of advertising, such as junk mail, but the bother of discarding junk mail is trivial compared to having to pick through a flood of spam to find the emails one wants to read. In the case of much other advertising, moreover, such as the advertising one finds in newspapers and magazines and on television, the recipient is compensated for having to encounter unwanted advertising, in the form of a lower price for a tied product that he wants to consumer, such as "free" television (i.e., paid for by advertisers). Second, the cost of filtering out spam (the demand for such filtering being further evidence that spam imposes net costs on most of the people who receive it) to the computer industry, and of "binning" in in hard drives and servers, is already in the billions of dollars a year, for which the spammers don't pay.
Spam thus creates a negative externality, a form of pollution. True, spam is not completely worthless'if (setting aside the scams) it generated no sales at all, it would be abandoned as an advertising medium. But the worthwhile spam can be preserved by a system in which spam is allowed to be sent to people who subscribe to the receipt of either all spam or particular categories of spam.
A different kind of externality has been created by the federal law (the Internet Tax Freedom Act, enacted in 1998) that bars state and local taxation of sales made over the Internet. The effect is to divert sales from conventional retail outlets to Internet sellers, such as Amazon.com. The exemption of e-commerce from normal taxes operates as a subsidy of that commerce. The most frequently heard arguments for the exemption are, first, that it is necessary to encourage the "infant industry" of e-commerce, and second that it is necessary to prevent a further growth of government. I find neither argument convincing. An infant-industry argument may make a little bit of sense for a country that has a promising industrial future but cannot finance the start-up costs of industry other than by preventing import competition, i.e., forcing consumers to finance those costs by paying supracompetitive prices. (Just a little bit, at most, because the global financial market is huge and highly efficient and therefore should be able to finance any promising commercial venture.) But there has never been difficulty in financing Internet-related ventures, even after the bursting of the dot-com bubble. And the "feeding the beast" argument is unconvincing because state and local taxation, unlike federal taxation, is effectively constrained by competition, since businesses and individuals can move with relative ease from high-tax to low-tax states. If states obtained substantial revenues from taxing Internet sales, there would be pressure to reduce tax rates. Banning state and local taxation of e-commerce seems a gratuitous blow to federalism.
In summary, I see no pressing need for imposing ethical standards on bloggers, but controls over spamming, and a repeal of the Internet Tax Freedom Act, deserve serious consideration.
Comment on Regulating the Internet-BECKER
The internet is one of the most remarkable and important innovation of modern times. Although initially encouraged by the American government, it has mainly grown worldwide in an unregulated, unsubsidized, and decentralized fashion. Its impact has been huge on communication and information, and to a much lesser extent as yet, on sales, and its development during the next couple of decades is likely to be just as important. The present and future revolutionary impact of the internet, and my skepticism toward government regulation and involvement, colors my attitude to informal or formal regulation of blogging and spam, and extends to taxation of internet transactions.
I agree with Posner that additional regulation of blogging and other internet postings is undesirable and unnecessary. Robert Merton, the late outstanding sociologist of science, demonstrated that the main way plagiarism and dishonesty are "policed" in research is through the incentives provided other researchers to discover and expose such malfeasance. These incentives are even more powerful in blogging and other internet activities, where many thousands of individuals seek to discover serious errors committed by bloggers, business leaders, and politicians. I have been impressed by the extent of the information revealed in comments on our blog, far more than in the responses per column from readers during the almost 20 years I wrote for Business Week magazine.
The case for taxation of internet transaction is to level the playing field with conventional retail and other outlets. A couple of years ago over 100 economists signed a petition to Congress to allow taxation of internet transactions for precisely this reason. Evidence by my colleague Austin Goolsbee does indicate that internet purchases are higher in states with bigger sales taxes.
Although this argument about equal treatment of different type of sales has merit, it is not enough in my opinion to overcome the case against internet taxation. I oppose taxation of the internet not because it is an "infant industry" that needs artificial stimulation to grow, nor because sales taxes of a few per cent alone would destroy this industry. Rather, my reluctance to interfere with the dynamics of the growth of the internet largely explains my opposition to taxation of transactions and other activities on the internet. I fear that the additional regulation of the internet that would inevitably accompany efforts to enforce taxation of transactions by either American states or the federal government would have a negative effect on internet growth in the United States.
Any significant sales tax on internet transactions would induce sellers and buyers to find ways to evade paying the tax. That includes setting up offices outside the United States, perhaps while shipping from places within the country, false invoicing, and still other methods from creative minds intent on evasion. All taxes induce avoidance and evasive actions, but internet transactions are particularly difficult to police, as seen, for example, from the proliferation of internet pornography. Hence attempts to collect taxes is likely to lead to substantial regulations that would slow down the so-far remarkable rate of innovation on the internet.
There are a few good other arguments against taxation of internet transactions. But the most important, I now believe, is its effects on further regulation of this dynamic media.
Spam does create real problems since it wastes time of recipients, and discourages use of email and other parts of the internet. As Posner indicates, spam has grown rapidly, and is commanding a larger share of all emails sent and received. The main factor behind its rapid growth is the low cost of sending spam in large volumes. A report by students in the Graduate School of Business of the University of Chicago collects very useful information about the spamming industry. Their study clearly shows how much cheaper spam is to send than either junk mail or telemarketing. About 150 spam operators appear to control the industry, and they collect many millions of names by scanning the internet. They then sell lists of names for tiny amounts per name.
The vast majority of spam is unwanted by recipients, but spam is costly either to block, designate as spam on recipient email accounts, or read and then send to trash. In other words, spam creates what economists call negative externalities or harm to recipients, which sometimes calls for regulation or taxation to discourage the activity producing the harm.
However, Guity Nashat (my wife) has indicated to me that the negative effect of spam is mitigated to the extent it substitutes for more costly intrusions on the time of recipients, such as telemarketing and junk mail. The total harm caused by spam is net of this reduction in other intrusions by sellers on recipients" time. The harm from spam is probably still positive, but less than when evaluated in isolation.
To the extent that the net effect of spam is a serious problem, there would be a case for simple and effective methods to reduce spam. By simple I refer to my concern expressed above over inducing additional regulations of the internet, in this case to control efforts to avoid anti-spamming approaches. A 2004 Federal Act requires an opt-out notice with a valid return email address, and places restrictions on unsolicited commercial email. Europe instead has a presumably more effective opt-in provision, but apparently does not provide any opportunity for companies or consumers to sue for damages. The effectiveness of either of these laws appears to be very limited, partly because many spamming companies have moved their operations to countries like Taiwan and China that do not yet regulate spam.
Other proposed solutions to the spam problem include Bayesian filters, challenge-response systems, blacklists, and digital signatures. Perhaps some combination of these systems will prove effective against spam without overly regulating legitimate email. As yet this has not been the case. So while it could be advisable to experiment with various approaches, it would be a mistake for governments to move aggressively against spam until more evidence is available on which anti-spamming devices are effective without imposing major costs on legitimate uses of the internet system. Otherwise, the attempt to regulate spam is likely to end up causing more harm to the internet and to society at large than any benefits that accrue from reduced spamming.
Blogging--A Response by Posner to a Comment
I want to respond to a single comment because it reflects interesting misunderstandings of the blogging phenomenon. The comment takes Becker and me to task for not responding to all the comments (on our postings) that are criticial of us. By thus not responding, we are said by the commenter to be shutting off debate.
We are not shutting off debate, and this for three reasons. First, a failure to respond to a criticism may end a debate in the sense of leaving nothing more to be said, but it does not "stifle" debate or silence the critic--on the contrary, it leaves the critic with the last word. Second, the comments are public--they are accessible to readers of the blog, at no cost, with a click. We "enable" (as the blogging expression has it) all comments; we engage in no censorship. Someone who reads a comment critical of a posting of mine or Becker's and observes that we have not responded to it will be more inclined to agree with the critic than if we did respond.
And third, and most important, bloggers have no "market power" that might enable them to limit debate. Not only are there 10 million blogs, but because it is costless (except for opportunity costs of time--though these can of course be significant and are one reason why Becker and I do not respond to all the comments on our postings) to create or post to a blog, and as any posting is diffused throughout the "blogosphere" essentially instantaneously, there is no way in which inaccuracies in a blog can be insulated from prompt correction. There is nothing to prevent the commenter fronm creating his own "anti-Becker-Posner" blog devoted to correcting our mistakes and omissions!
There is another point worth noting. Inaccuracies in blogs are less pernicious than inaccuracies in the mainstream media even apart from the superior opportunity for prompt correction of bloggers' errors. The reason is that bloggers are known not to employ fact checkers or editors; there is no pretense that they have the resources to eliminate all errors in their postings. The mainstream media, in contrast, represent to their public that they endeavor assiduously to prevent errors from finding their way into articles and broadcasts. They ask the public to repose trust in them. Bloggers do not. That is why serious errors by the mainstream media are played as scandals; they are not merely mistakes--they are breaches of trust.