Demography of the Tech 100
Technology in general and a desire for high-tech business in particular are shaping the U.S. cultural landscape and simultaneously building a new landscape: the American blogosphere, an imagined community composed of millions of individuals writing online. The link-based social hierarchy of weblogs allows us to identify a number of popular bloggers who have disproportionate influence on a national level. These online opinion leaders represent the some of the loudest voices of the technocratic elite, and knowing who and where they are can help us understand the relationship between the blogosphere and the real world.
The blog itself is notoriously difficult to define, but Jill Walker’s definition, “a frequently updated website consisting of dated entries arranged in reverse chronological order so that the reader sees the most recent post first,” covers most instances of the genre.[i] As of November 2004, more than 8 million Americans had created weblogs, and twenty-seven percent of American Internet users, roughly 32 million people, claimed to be blog readers.[ii] Both blog readers and creators are often young, male, and well-educated.[iii] Seventy percent of blog creators have broadband access at home, and 82% have been on the Internet since 1998 or earlier.[iv] This is a genre for people deeply entrenched in the Information Age.
There is considerable controversy as to how to determine which are the most popular blogs. In May 2005 the site Blogebrity, participating in a competition to see who could create the most virulent Internet meme without advertising, claimed to be creating a gossip magazine focusing on bloggers[v]. Borrowing entertainment journalist James Ulmer’s term for the most bankable actors, Blogebrity created an “A-List” of popular bloggers (as well as a B-List and C-List) and attracted more incoming links than any site in the contest – however, as it lacked an objective determining metric, the Blogebrity A-List was fundamentally arbitrary. Still, the lack of a clear measurement has not killed the idea that there are elite “A-List” bloggers.[vi] After all, both readership and incoming links follow a power law distribution, so that some sites are dramatically more popular than others.[vii] Though the boundaries of any study of the most popular bloggers are regrettably vague, the influence of the A-List on American blog readers makes it a group worth examination.
For the purposes of this essay, Technorati.com, a website that tracks links at more than 22.4 million blogs all over the world, was used to determine which blogs were the most popular. The Technorati Top 100 Blogs list ranks “the biggest blogs in the blogosphere, as measured by unique links in the last six months,”[viii] links being more valuable than hits or visits here because they measure popularity within the blogosphere as determined by the bloggers themselves, and thus closely approximate the informed insider perspective of an A-List. Unique links also provide a good approximate measure of blog readership and influence: some of the most popular sites don’t provide a public record of how many visitors they’ve received, but readers are more likely to find commonly linked-to blogs and the ideas at those blogs are more likely to spread.
Since the Technorati 100 only takes into account links found in the last six months, some blogs on the list drop in rank and disappear over time as they become less popular. I confined my study to the November 12, 2005 Technorati 100 and the subset of those blogs which were also on the list on November 12, 2004.[ix]
Of the 100 blogs ranked by Technorati, 19 were in a foreign language, most often Japanese, and were assumed to be foreign-based, and 14 were English-language blogs whose primary author or editor[x] was based outside of the United States. The latter category included Michael Yon, a freelance reporter from Massachusetts who blogs on-location in Iraq. Three of the Top 100 blogs had no discernable location: a gossip site named The Superficial, the Web Standards blog, and Hack-a-Day. While their focus and vocabulary seemed to indicate that these three blogs were U.S.-based, they were excluded from the survey for lack of clear evidence. Finally, the collaborative site Boing Boing, with five authors, was only counted as 3/5 of a blog: Cory Doctorow was ignored because he lives in the United Kingdom, and John Battelle was excluded on the grounds that, as sole author of the 98th-ranked Searchblog, his demographic information was already being taken into account.
It was immediately clear that the remaining 63.6 blogs were very different from the average blog. Even after excluding LiveJournal and other online diary websites, Herring’s study of randomly selected blogs found that blogs are “overwhelmingly of the personal journal type (70.4%), in which authors report on their lives and inner thoughts and feelings”.[xi] In contrast, only three (4.7%) of the top U.S. blogs were personal journal-style sites run by a single individual. Thirty-three U.S. blogs, more than half, were maintained by more than one individual, whereas less than ten percent of blogs in the Herring survey had multiple authors[xii]. Only five (7.8%) of the top U.S. blogs could be considered personal rather than topical. These percentages were slightly higher among the 21 long-lasting Tech 100 blogs, which included two diary websites (dooce.com and kottke.org), but in general, people link to these blogs because they entertaining or informative, not because they have an emotional investment in the lives of strangers.
The most popular bloggers tend to be well-educated and male to a much greater extent than the rest of the blogosphere. Four of the top U.S. blogs are mouthpieces for companies or organizations and have no clear author or editor; these were excluded from this analysis. Of the 59.6 U.S. blog authors or editors remaining, 53.4, or 89.5%, were male. Among bloggers generally, males are only slightly more prevalent: the percentage of male bloggers was 54.2% in the Herring survey[xiii] and 57% in the Pew Internet Life poll.[xiv] Even more dramatic is the education level of top U.S. bloggers. In the American blogosphere as a whole, 39% of bloggers have college or graduate degrees[xv]. Of the 56.6 top U.S. bloggers for whom education levels could be determined[xvi], 55.6 (98.2%) had at least a bachelor’s degree, 21.2 (37.4%) had at least a master’s, seven (12.3%) had J.D. degrees and five (8.8%) were Ph.Ds. In contrast, Census 2000 data puts the percentage of Americans 25-and-older with at least a bachelor’s degree at 24.4%.[xvii] Despite the low level of technical knowledge required to set up and maintain a blog, well-educated male Americans hold most of the top spots, a trend already seen in the blogosphere as a whole and amplified in the Technorati 100.
For the purposes of this study, the 63.6 blogs were divided into six categories: personal (6), political (21), technology (19.6), pop culture (9), humor (3), and corporate/organization mouthpieces (5), and assigned to Consolidated Metropolitan Statistical Areas as defined by the Federal Office of Management and Budget in 1999 and used in the 2000 Census. A best subsets regression was done to fit the number of Technorati 100 blogs in general and in each category to the following independent variables, separately and in concert: CMSA population, the five indexes devised by Florida and Gates, the Metropolitan New Economy Index score, MNEI broadband access score, MNEI weighted education score, Milken Tech-Pole rating, and Milken Tech-Growth Rating. Index data was unavailable for four smaller MSAs (1 tech, 2 political, 1 pop culture), and they were removed from multivariate analysis, resulting in a slightly higher significance (r2 = 73.0 rather than 71.1) for population. As expected, the number of blogs in a given metropolitan area was significantly influenced by the population; in fact, population was the most significant factor in determining the number of blogs for all types except humor and corporate, which correlated best with broadband capacity and MNEI respectively. None of these results are especially surprising – broadband access was also a significant factor for the Technorati 100 generally and the top American corp/org blogs are run by some of the top technology organizations, a finding which merely reaffirms the value of the MNEI.
One central question was whether the most successful bloggers shared the preferences of the amenity-seeking urban professionals Florida and Gates describe. After calculating the best combination of factors as judged by significance and a low Mallows C-P, I determined that A-List blogs in general don’t follow any of the Composite Diversity Index factors. Population is easily the most important factor, along with the city’s MNEI score and access to broadband. Together they can account for r2 = 82.4 of the blog distribution among CMSAs. What didn’t have influence came as more of a surprise: expressed as either the MNEI education score or the percentage of residents with a college degree, education is a relatively poor predicator of blog distribution in general.
Education only becomes important for blogs that emphasize opinions rather than utility or entertainment[xviii]: all the top U.S. bloggers with Ph.D.s or J.D.s maintain political blogs, and their level of education as well as a considerable amount of biographical information is usually readily available on their websites. It’s their graduate degrees that give these bloggers much of their cachet – there’s no shortage of political opinion on the web, but these blogs offer a high level of consistency and authority. A pundit’s blog postings can also translate into increased real-world speaking opportunities, which can make him still more popular in the blogosphere. Their level of education suggests that these bloggers are also older than average, and more established in their careers. Conservative bloggers were slightly more likely to locate in or near cities with a lower population, and for political bloggers in general there was a less significant relationship (r2 = 47.6) between a metropolitan area’s population and its number of bloggers than there was for top bloggers generally (r2 = 73.0). The political orientation of the top blogs is roughly even – 12 liberal, 9 conservative – but taking into account differences in linking patterns between conservative and liberal bloggers[xix] the conservative blogs are overrepresented in the Technorati 100. It’s also important to remember that these blogs represent not one unified community of interlinking bloggers but rather at least two separate communities. As Adamic and Glance have observed, “91% of the links originating within either conservative or liberal community stay within that community.”[xx] This separation is only one example of a larger Internet linking trend of “like links to like.”
Political affiliation has little to no influence on the linking decisions of technology and pop culture bloggers, the next most common types. These categories are necessarily broad: technology includes both analog do-it-yourself sites and digital hacker fare, and the pop culture category, though mainly concerned with nonpolitical gossip, also includes Mark Cuban’s sports-oriented Blog Maverick site. The most popular Technorati site by a fairly large margin was the group technology site Boing Boing, with 67,893 links from 17,719 websites. It was also the most popular blog in 2004. Like many successful collaborative sites, Boing Boing is difficult to classify: the site tagline is “A Directory of Wonderful Things” and the editors will often post links to odd news or products they find interesting, but the main focus is on various types of technology. Technology blogs are influenced by the same factors as the Technorati 100 in general; adding the Milken Tech-Pole rank to the aforementioned parameters adds 0.17 to the correlation and results in an r2 of 68.3, and adding the Milken Tech-Growth rank provides another small jump, but there’s no clear place to stop adding in parameters. Though one would expect all of the diversity factors used by Florida and Gates to correlate with the technology jobs these two Milken ratings are ostensibly mapping and thus with the presence of A-List technology blogs, adding in the six other independent variables results in an r2 increase of only 2.3.
For the eight included pop culture blogs, however, one of the Florida and Gates diversity measures became one of the most significant independent variables. Population, broadband access, and the city’s Bohemian Index ranking yielded an r2 of 61.4 and a near-zero Mallows C-P. Moreover, Lexington, Kentucky, which hosts the most popular pop culture blog (Fark.com) but could not be included in this data set, ranked 13th out of 68 other 250,000-500,000 resident Canadian and United States cities on a Bohemian Index in another Florida and Gates paper.[xxi] Unlike most other top bloggers, the pop culture bloggers do seem to be amenity seeking – they live in places like Seattle, New York, and Los Angeles, the top three cities on the Bohemian Index. Some, like the L.A.-based Defamer.com and Manhattan’s Gawker.com, were intentionally located in their respective cities by a parent company, with the expectation that there was a large audience for local gossip from someone with local access. However, most A-List pop-culture blogs, especially those located outside of those three major cities, are collaborative efforts with a single editor. In cases like these the connection is less obvious; perhaps the kind of people who want to edit culture-oriented sites are also drawn to bohemian centers of creativity offline, or vice versa.
As noted above, the Technorati 100 contains a far lower percentage of diary/personal blogs than the Internet in general. The six U.S. based diaries in the Technorati 100 represent several different personal webpage types: there are the expected single-author Internet diaries so common elsewhere; two sites, Postsecret.com and Overheard in New York, that incorporate the personal thoughts or experiences of hundreds of users, and finally the New Journalism-style Survival of New Orleans Weblog, which records the personal experiences of a group of people participating in a major news story. The two diary-blogs also popular in 2004 were the traditional type – the aggregated group diaries that appear on this list are still a relatively new form for most weblog readers, and sites like The Survival of New Orleans or the election blogs of 2004 speak to a momentary Zeitgeist. Unlike these other diarists, the writers of more traditional personal weblogs were Internet veterans with at least four years experience blogging. That no newer bloggers of the same type joined their ranks in 2005 is evidence of both the limited popularity of the web diaries, which tend to pursue nanoaudiences of readers acquainted with the writer, and the difficulty of achieving a significant following when millions, rather than thousands, are doing the same thing. It’s here that one finds a trend likely to continue – when so many bloggers have already established themselves, it’s difficult to become very popular without filling an increasingly specific niche. As popular websites remain popular, the A-List stabilizes. Though the low number of A-List diarists makes any conclusions speculative, these bloggers were influenced by the same factors as Technorati 100 bloggers in general: population, access to broadband, and a city’s Metropolitan New Economy Index score, when considered alongside a city’s Milken Tech Pole rating, accounted for an r2 of 62.7 (population alone had an r2 of 47.6).
Tech-savvy early adopters whose webpages were still novel in their day – even relative latecomer dooce.com offered readers a unique hook: she was one of the first people to be fired because of her blog, hence the Internet neologism “dooced” – the A-List diarists have much in common with the writers of technology weblogs. Indeed, it’s these two groups that most closely resemble the Technorati 100 as a whole; population, MNEI index score, and access to broadband were the key variables for these bloggers, while measures of education and diversity were poor indicators of their presence.
Broadband, then, emerges as a surprisingly important amenity – while it’s safe to say that most metropolitan areas offer some form of Internet access, the popular bloggers tend to emerge in areas where fast Internet access is widely available, that is, where Internet use is a popular and important part of the local culture. Ensuring that a city is well “wired” is a good way to encourage this. Though their indexes (most strikingly the bohemian index) can predict the presence of some blogs, Florida and Gates’ greater contribution to this study is the notion that “low entry barriers to human capital help attract talent.” Wide availability of broadband increases awareness of an Internet culture – and aside from that perhaps purely psychological effect, competition for broadband customers, of which the broadband index used in this paper is largely a measure, has pushed down prices and lowered economic as well as psychological barriers to entry in highly-ranked cites.[xxii] It will be interesting to see if Technorati 100 representation increases next year for those cities offering or planning to offer free wireless service to their citizens.
One final observation: human capital doesn’t just move between cities, it moves on- and off-line. A future study might look at factors in the online environment analogous to the indexes used by Florida and Gates. Does the presence of popular, transparently local weblogs create a “blogfather” effect, encouraging others to start blogging? What influences a blogger to take up a local, rather than broadly national identity? Does a visible online GLBT or immigrant community increase the rate of growth of the local blogosphere? These and other questions require us to take a few more steps into the blogosphere. I’ve attempted to find the relations between the online and the offline world, and perhaps there are variables yet-unaccounted-for that would explain the distribution of top bloggers even better than the ones I chose – but clearly the parallel world of the blogosphere can’t be explained just in terms of “meatspace”[xxiii] – it has its own geography and demographics, and no doubt its own influences on the offline world of which it is ostensibly a reflection.
[i] Routledge Encyclopedia of Narrative Theory, s.v. “Weblog.”
[ii] Lee Rainie, “The state of blogging,” Pew Internet & American Life Project Report: Blogosphere, http://www.pewinternet.org/PPF/r/144/report_display.asp
[iii] Ibid.
[iv] Ibid.
[v] Joshua Glenn, “Almost Famous,” The Boston Globe, June 5, 2005. http://www.boston.com/news/globe/ideas/articles/2005/06/05/almost_famous/
[vi] Alex Halavais, “Blogging in the Plural,” http://alex.halavais.net/?p=1281
[vii] Clay Shirky, “Power Laws, Weblogs, and Inequality,” http://www.shirky.com/writings/powerlaw_weblog.html. Shirky also emphasizes the arbitrary nature of the A-List mentality, noting that “The largest step function in a power law is between the #1 and #2 positions, by definition. There is no A-list that is qualitatively different from their nearest neighbors, so any line separating more and less trafficked blogs is arbitrary.”
[viii] http://www.technorati.com/pop/blogs/
[ix] This leaves a six-month gap from November 2004 to May 2005 unaccounted for, but the popularity of these sites during two separate non-consecutive six month periods should still suffice as evidence of their consistent popularity.
[x] A number of blogs have numerous contributors, but except in rare cases (Boing Boing, Powerline) where there was no obvious editorial control and a composite person was created, the blog author is the person who has final editorial control over content, excluding copy editors and spellcheckers.
[xi] Susan
Herring, et. al. “Bridging the gap: A genre analysis of weblogs.” (Proceedings
of
the 37th Hawaii International Conference on System Sciences, 2004), 6.
[xii] Ibid, 5.
[xiii] Ibid, 5.
[xiv] Lee Rainie, “The state of blogging,” 2.
[xv] Ibid, 2.
[xvi] Whatever their education level, the indeterminate three (Microsoft employee Scott Isaacs, actor Wil Wheaton, and Survival of New Orleans blogger Michael Barnett) are well advanced in their careers. The sole top blogger definitively without a degree, Theron Parlin of Thought Mechanics is currently in college.
[xvii] Census 2000 Demographic Profile Highlights, http://factfinder.census.gov/servlet/SAFFFacts?_sse=on
[xviii] Few blogs have a single purpose like “entertainment” or “utility,” and the most successful are frequently entertaining regardless of their focus. The pseudonymous political blogger Billmon (Technorati #94) gets links despite his lack of credentials precisely because he delivers his liberal political views with an especially entertaining writing voice.
[xix] Adamic, Lada A. and Natalie Glance, “The Political Blogosphere and the 2004 Election: Divided They Blog,” 2005, 3.
[xx] Ibid, 3.
[xxi] Gertler, Meric, et al. “Competing on Creativity: Placing Ontario’s Cities in North American Context,” 2002.
[xxii] “Competition slashing costs of broadband.” USA Today. 7/14/2005.
[xxiii] “Meatspace” http://www.wordspy.com/words/meatspace.asp