Last Updated April 23, 2008
This project arose in response to one of the great myths or the Internet age, i.e. the coming of cyberspace heralds the end physical constraints which will eventually lead to the death of cities. In fact, the exact opposite is occurring. The largest concentrations of Internet users and producers are located in urban areas and many of the most innovative firms in the Internet space are housed in downtowns. There should be nothing surprising about this since, cities have always been the primary source of innovation and will continue to play this role in the future.
Although the power of the Internet does opens up new possibilities for long-range collaboration and even new spaces of interaction within cyberspace it also exhibits much of the traditional unevenness that has characterized urban and economic development throughout history. The fact that information can be easily and widely distributed is often mistaken for an indication that the production of this information is also diffused. In fact, there is a much more complicated dynamic involving the connection of specific places to global networks resulting in a system of production that is both place-rooted and networked at the same time.
One of the greatest challenges facing any research project involving the Internet is finding reliable and practical indicators to support ones theories. In particular, assigning geographical locations to what takes place on the "spaceless" Internet is especially difficult. With this specific problem in mind, the Domain Name research project is an attempt to map the physical geography of one indicator of the Internet, i.e. domain names such as nokia.fi or nytimes.com.
In many ways domain names are one of the most basic building blocks of the commercial Internet. Although actual data packets are routed by computers according to IP addresses, these numbers, e.g., 184.108.40.206, are hard for human beings to remember. The domain name system was developed so that users could use the Internet address www-dcrp.ced.berkeley.edu rather than its numeric equivalent. This system, which was originally designed as a convenience for a small number of computer specialists and academics, has now become the ubiquitous means of brand identification within the Internet content business. Although registering a domain name has become relatively easy and inexpensive, it nevertheless represents a conscious decision to use the Internet in a more sophisticated manner. In short, domain names are the best indicators for the location of Internet activity.
Domain names also have the advantage of containing contact information for the person or entity which registered it. Although there is no guarantee that the registration addresses for a domain name and the location of Internet content production is the same, an analysis using the CorpTech database shows a strong correlation between the two. The CorpTech database contains accurate and up to date contact information on 50,000 high technology firms in the United States. For 84 percent of these firms, the Zip code obtained from its ".com" domain name registration matched the Zip code in the CorpTech database at the 3-digit level (roughly equivalent to a geographical area the size of a small to mid-sized city) and 73 percent of these firms match at the 5 digit Zip code level (roughly equivalent to a neighborhood within a city). While this is a small sample of total domains, it does strongly support the use of domain names for determining the location of Internet content production.
US and Global Geography of the Internet
The use of domain names to indicate the geography of the Internet in the United States is relatively straightforward because the principle top level domains in use are com, org, net, and edu (CONE) which are centrally administered.  A complication in using domain names as an indicator of the global Internet is that in addition to CONE top level domains, country code (CC) domains such as ".de" for Germany and ".jp" for Japan are also in wide use in countries outside of the United States.  The ".com" domains are by far the most prized and many companies add ".com" to their name in an effort to increase their share prices and some companies have chosen their name based on the availability of a ".com" domain name.
Country code domain counts and the number of hosts or computers associated with these domains, have often been used as proxies for statistics on Internet use by various countries. However, these counts consistently underestimate the total number of domains in countries because they exclude CONE domains. Although it is true that most of the CONE domains historically were registered in the United States, the trend over the past several years has been towards dispersal. For example, MIDS reports that in January 1997, 83 percent of all ".com" domains were located in the US and, the top three countries, the US, Canada and the UK, accounted for over 90 percent of the ".com" domains. In July 1999, only 70 percent of the ".com" domains were located in the United States and the same top three countries accounted for just 78.6 percent of total come domains world wide. This makes it extremely important that both CONE and CC domains are included when conducting a global comparison.
List of Papers
To date, I have written a series of articles and conference papers outlining my findings and I will continue to post my research here.
I also strongly recommend Martin Dodge's Atlas of cyberspace http://www.geog.ucl.ac.uk/casa/martin/atlas/atlas.html for similar research.
 This analysis has purposively excluded the top-level domains ".gov" used by the US Federal government, ".mil" used by the US military, and ".us" largely used by US state and local governments because of data availability problems. Moreover, these top-level-domains do not reflect the current trend of Internet use, i.e. private individuals and firms using the Internet for personal and commercial purposes.
 CONE domains are centrally administered by a monopoly delegated to Network Solutions by the US's National Science Foundation (NSF) which was originally responsible for registrations. Country code domains are generally regulated by a central institution determined by each country. The exact nature of this institution and the rules governing country code domain name registration vary depending on the country. Moreover, as the Internet has grown, this simple structure has rapidly become much more complicated than the original design. For example, some country codes such as the Pacific island nation Tuvalu's have emerged not as a symbol of the country's Internet presence, but as a potential battleground for television networks hoping to capture the potentially important and lucrative ".tv" brand.