Richard A. Epstein: No need to fight yesterday’s wars
>Thomas W. Hazlett: The ‘Noam index’

Media concentration and its potential impact on politics and culture has become a big topic around the world, from Italy to Australia, and reaching even Iceland. In America, Congress last year received almost a million angry letters from across the political spectrum when the Federal Communications Commission tried unsuccessfully to loosen ownership restrictions.

Both opponents and supporters of ownership relaxation conducted the debate with remarkable self-righteousness, perhaps because both are in a way correct. Data from my forthcoming book, Media Ownership and Concentration in America, show that US mass media have, in the aggregate, indeed steadily increased in concentration since 1988. But they also show that the concentration on a national basis is usually fairly low by the standards of US antitrust. Here, the official US government guidelines define an unconcentrated industry as having a “Hirschmann-Herfindahl Index” of less than 1,000. The HHI is the sum of the squares of the market shares, or Σs². For example, if a local radio station market consisted of two companies with 40 per cent each, and of two companies with 10 per cent each, it would have an HHI of 3,400. Markets with an HHI above 1,800 are defined as highly concentrated, while markets with an HHI below 1,000 are deemed to be unconcentrated.

By that antitrust standard, many media industries are unconcentrated. Our study shows that even radio, the poster child for growing media concentration, had in 2002 a national HHI of only 469, which is quite low. For TV station ownership it was 152; for newspapers, 254; for film distribution, 1,072; and for cable TV, 1,380. (Concentration is much higher locally, given the smaller markets. For radio, it is 2,400, and for newspapers and cable it has long been above 7,000. Rarely is there more than one local newspaper in a city. But that type of concentration is usually not affected by national mergers, and is a separate question we will reach later).

Compare these numbers with those of other industries. Just for example, video game consoles have an HHI of 4,494; for cellular handsets the number is 1,967 - much higher without unleashing anything like the same passion about concentration.

The reason for this discrepancy is that most people desire a greater diversity in their media sources than in their computer hardware. They want more choices for themselves, and for the political process. Therefore, the question arises of whether the traditional antitrust measure of the HHI is appropriate for media.

The issue is partly whether the concentration threshold for media should be lower, and also whether the HHI methodology itself accounts sufficiently for media pluralism. For example, in the radio example above, if the two smaller stations were replaced by 20 stations, each with 1 per cent of the market, the HHI would decline only slightly, from 3400 to 3220. Yet the diversity of the local radio market would clearly be significantly increased by the presence of 18 additional radio station providers.

To deal with that issue, the FCC introduced a new “Diversity Index”, which counted each media outlet as a “voice”. But this approach disregarded the size of the stations or newspapers. It equated the New York Times with the Poughkeepsie Gazette, since both happen to be in the same media market. That approach was repudiated in June by a federal court, which declined to accord the agency even the usual deference to its expertise. This leaves the question of how to measure concentration and pluralism in media wide open.

Pluralism is important. But there is no conceptual, practical or legal way to officially define and measure the vigour of a marketplace in ideas. The best one can do is to count voices, and assume that in a competitive system, diversity of information increases with the number of its sources.

Yet market power is also undeniably important. The antitrust HHI is a pretty good litmus test for market power, but it does not make allowance for pluralism. As a radio listener, I am better off with another 20 stations on the dial or another newspaper sold at the news kiosk, even if few people listen to or read them. Their availability provides an option that carries value even if it is unexercised by most readers or authors.

The conclusion is that one should not have to choose between a measure of market power (the HHI) or of pluralism (the number of voices) but ought to incorporate both. I therefore propose that one replace the FCC Diversity Index and the HHI with another diversity index. Such an index would take the HHI as a measure of market power, and divide it by the square root of the number of voices. The index would be Σs² /√n. Thus, the less concentrated and the more numerically diverse a market is, the lower the index. To keep the index practical there should be a cut-off of a minimal size for a voice. One per cent seems a reasonable floor: small but not trivial.

This leaves the question of what the concentration thresholds ought to be. This is a matter of policy, taste and market size. To some people, one voice is plenty so long as it agrees with them. Others would want to assure almost every perspective its voice, regardless of how minor. For an order of magnitude, an index of 500 would be the result of about 7.5 equal voices, or of four voices with a market share of 20 per cent each, with the remainder shared between seven other small players. Today, radio, TV stations and newspapers are below that number nationally, while cable TV is close to it, and TV networks are above. The music industry’s recently approved merger between BMG and Sony exceeds that threshold. As a national figure, 500 seems in the ballpark for a threshold. At 300, there would be about ten equally sized companies, or a lot of small ones. The range between 300 and 500 is one of moderate concentration.

The formula, slightly modified, could also be used for the concentration of all media in the aggregate, since a company might have no special market power in any particular medium but be involved in several media so that overall it would hold significant power, especially if it were to have multiple holdings in one city.

The desired media concentration threshold would depend on the size of the market. For a local market, or for a smaller country, it would be different than for the giant American national market. As an illustration of the concept, such a function could be: Diversity Threshold = 10,000 x (population size of the market) ֿ·² . Such a formula would set the threshold, in a small local market of 100,000 population, at an index of 1,000, which translates to about 4.6 independent voices. In a market of a million people, the formula would result in 6.3 equal-sized voices. And for the US as a whole, the overall threshold would be an index of 200, the equivalent of 13.5 voices if they are equal-sized, or for example 4 companies of 15 per cent each, 4 companies of 5 per cent each, and 17 companies of 1.2 per cent each.

Some people will oppose this approach. They might suspect darkly that it means a loosening - or tightening - of the existing rules. But that is a question of where the thresholds would be set, not of the methodology itself. Others might argue that no intervention at all is warranted, since markets will generate competitive entry and diversity. In that case, fine, there would be no need to ever use any media diversity test. But suppose that market power does emerge? What then? After all, the economics of media and information, with their high fixed costs and low reproduction costs, create strong economies of scale that often favour concentration. In any event, such a laisser-faire approach is highly unlikely in the real world, considering the FCC’s debacle in Congress and the courts in trying to loosen the rules even a bit.

To still others, any numerical test is suspect as mechanistic. They would prefer a case-by-case consideration of many factors relevant to a media market. But this would leave a judgment call over media ownership to government officials able to reward friends and punish enemies, or enable powerful media companies to thwart unfavourable decisions - both undesirable options given the inherently adversarial relationship of government and media. This argues for a relatively clear-cut test, with a relatively clear-cut methodology. And to allow for truly special circumstances, it would be what lawyers call a rebuttable presumption, not an inflexible rule.

Given the contentiousness of the issue, it would be best to create such a system in advance rather than to do so ad hoc, ad hominem and ad infinitum.

The writer is professor of economics and finance at Columbia University and director of its Columbia Institute for Tele-Information

............................................................................................................................................

Richard A. Epstein: No need to fight yesterday’s wars

>

As Eli Noam reports, the issue of media concentration has been a focal point for intense squabbles worldwide. The most notable of these was the American extravaganza in which hordes of outraged citizens (and frightened competitors) beat back the effort to liberalise the ownership restrictions on the number of radio and TV outlets that major players could have in key markets.

Prof Noam is quite correct to note that the debate functions on two dimensions. The first deals with the level of economic concentration in any given market, which is generally measured by the Hirschmann-Herfindahl index. To that he adds a measure of the number of distinct voices in the market (by using the snazzy 1/√n) to generate some combined number that corrects the HHI by giving better scores to markets with more distinct voices. The HHI thus joins forces with the Noam Voice Index, or NVI.

Any exercise of this sort gives rise to the nasty question of defending a particular, unavoidably arbitrary, breakpoint between an acceptable and unacceptable levels of concentration. In those cases where the boundaries of the market seem firm, and the dangers of monopoly large, it may well be worth deciding which markets are unconcentrated and which ones not. But why go through all this with respect to broadcast outlets, where technology is overrunning all the traditional boundaries in the communication market?

It is noteworthy that this column, like our entire FT Tech series, is published online. The ease of entry into the online market is too well established to be denied. Right now the ubiquitous bloggers continue to exert a powerful influence on the shape of political discourse in the United States and around the world. You do not have to be a reader of Glenn Reynold’s InstaPundit.com or the Volekh Conspiracy to be subject to their influence. You cannot but note that they often tend to be further to the right than the dominant political media.

But right or left, these guys are ubiquitous. The bloggers are there first; they do not have to worry about publication deadlines; and they do not have to clear their broadsides with an editorial staff. You may not read them at all, but you can be sure that the political writers for the major media outlets rely on the bloggers, just as the bloggers rely on the established media, in a warily symbiotic arrangement. And if you want to track them down, it takes about five seconds to find them on Google from a dead start.

So the question is whether these folks count as part of the media market? I think that they do. They may have smaller readerships than the large daily papers or the broadcast world. But by the same token, that readership is likely to be more connected and influential that those indifferent souls who are willing to wait 24 hours to get the latest scoop. So if they are in the game to stay, then the HHI and the NVI have to be recalibrated to take into account their role. The number of real players is larger than Prof Noam’s calculations suppose. Ditto the number of voices. It is like the good cholesterol has swamped the bad. If so, then why fight the futile wars of the last millennium in the FTC and the Halls of Congress, both of which are showing their age. Just beat the whole process by adding your voice to the chorus of folks who are already online. Like Prof Noam. And me.

The writer is the James Parker Hall Distinguished Service professor of law at the University of Chicago and Peter and Kirsten Bedford Senior Fellow at the Hoover Institution

Back to top

............................................................................................................................................

Thomas W. Hazlett: The ‘Noam index’

>

The “media diversity” hurricane that has swept the US and other countries has been almost entirely unpredictable. Eli Noam here attempts to inject some measure of calm. His intentions are noble.

“Media diversity” has been discovered as an issue of great public salience - on an extremely superficial level. In many markets where choices are becoming greater with the emergence of competitive entry, debate focuses not on the improvement in consumer options but on the structure of markets. Of course, if this focus were sophisticated, its misplaced aim might not amount to much of a detour. But the public debate is stunningly shallow, even by contemporary standards. Hence, when a Ted Turner exploits opportunities created by the deregulation of cable television to build an empire challenging old-line broadcasting, journals of news and opinion report that regulation opened these doors - and Ted Turner himself headlines the allegations. In articles in the Washington Post and Washington Monthly, the media billionaire claims that only through government protection was he able to launch his cable TV enterprise, when in fact government protection of broadcasters kept such competition from popping out years earlier.

Eli Noam’s contribution here is twofold. First, he focuses attention on the facts of the market. Have media markets become more concentrated over time? He believes that, in many cases, they have. Are these markets dangerously monopolistic? In general, he finds that they are not.

Second, he advises policy makers to take into account not only standard measures of industry concentration such as the Herfindahl-Hirschman index but also the raw number of programme choices available to viewers, readers, or listeners. Options are important, even if few people choose those outliers far from the popular core.

This is a reasonable approach within the context of the ongoing regulatory process and may well improve the result. But, alas, the broader picture is that the very media diversity analysis Prof Noam hopes to uplift is mired in regulatory failure. The contrast between print and broadcast markets is stark. In the former, the First Amendment has been interpreted to prohibit government controls (beyond antitrust regulation). In the latter, government licensing and content regulation - equal time rules, fairness doctrines - are allowed. Despite market power in newspapers, magazines and book publishing, the public is treated to a robust diversity of print opinion far beyond what is made available via regulated airwaves.

The challenge for modern advocates of free speech is to protect our cherished freedoms even as they move to digital transmission. When American judges started applying the First Amendment based on whether the information was distributed via printing press or by electromagnetic spectrum, they quarantined free speech protections within a traditional technology. That tradition is now becoming ancient, as newspapers migrate to “new media.” Will our laisser faire First Amendment continue to shield the New York Times from government editorial review when its median reader uses a wireless Blackberry to download its news? The Noam index may assist the analysis by including a measure of the numerous digital choices available to readers. But the better answer is categorical extension of the First Amendment beyond yesterday’s “print” to today’s “communications.”

The writer is a senior fellow at the Manhattan Institute for Policy Research

Back to top