What is the digital divide?

The premise of the "digital divide" is that there is differential access to computers, the Internet, telecommunications, and information in society. This differential access is typically characterized along the variables of income/poverty, education, race, age, ethnicity, disability and geography. It also includes unequal access to knowledge, training, resources, job opportunites and so on, or the practices of the information economy. To characterise the digital divide solely as a technological divide is therefore a (common) misconception.

What is the geography of the digital divide?

The geography of the digital divide combines two concepts: understanding and sensitivity to spatial difference, ie., spatial justice, and the new round of unequal access described above. To some extent it is true that the digital divide is a new name for an old phenomenon: uneven spatial distribution of resources. On the other hand, there are some new aspects and difficulties to overcome, especially because the resources are abstract (information, knowledge) and apparently do not carry the same obstacles to distribution as physical resources.

The geography of the digital divide can be looked at across scale. For example, the issues at the global scale are different from those at the regional, and again at the local/citywide.

New! Census 2000 Supplemental Data for Georgia (income, poverty rates, housing characteristics). Available here.

Example: Uneven access to Internet at global scale: 2000
Example: Global domain name density (Internet Geography Project) Source: Matthew Zook
Example: How well does access correlate with income? About R2 of .76.
Example: Regional Map of Access, Africa. While many countries may have some access, this may be restricted to the primary city only. Source: Mike Jensen

For the most part, the work described here is at the citywide scale (sub-county). The reason for this is simple, there has not been much work done at this scale, and the local is where interventions begin and have their effect.

The digital divide in Atlanta

The work to be carried out in Atlanta will use GIS to analyze the geography of the digital divide across the city.

1. To create a probability surface of access across Atlanta. This will involve the combination of relevant variables (income, education, race, etc.) to answer the question of who has access, and where. This surface is a model, rather than a record of who actually has access. The disadvantage of this is that it may not be right. The advantages are that it can be applied to other cities, or other time periods (past or future) once it has been calibrated. In addition, actual Internet access data is expensive and often proprietary. Finally, the model can be checked against samples to test its accuracy.

Baseline Data: Census 2000, Percent Black shows a strong "tale of two cities".
Example: Computer Ownership by county for Metro Atlanta, 2000
Example: Dot-Com Domain Density, SE (links to clickable Shockwave). Note concentration in Atlanta. Source: Matthew Zook
Example: Online by Zip, Metro Atlanta. This map shows the density of online access by zip code across Atlanta.
Example: Population Density by Census Tract, Atlanta, 2000. This shows urban areas (over 3200/square mi) and suburban areas (320-3200/square mile). How does it compare with the online map? Source: US Census Bureau (American FactFinder)

2. Develop a location-allocation model to determine the best sites to locate resources to bridge the digital divide. These resources may include community technology centers (CTCs), mobile computing centers (eg., a GIS bus) or other efforts (mobile GIS deployed through wireless PDAs). Location-allocation is a particular spatial technique developed to resolve the question of how many, and where best to locate centers in relation to the surrounding spatial pattern of access (eg., roads, public transportation routes) and population (those with the lowest Internet access in this case). It is at this stage that a comparison can be made to actual remediation efforts, such as that by the Mayor's Office of Community Technology (MOCT), which is installing 15 cybercenters across Atlanta by end 2001. Are they optimally located, are they close enough to users, are there enough/too many?

Example: A surface of Atlanta welfare (TANF) recipients, 1998 Source: The Atlanta Project
Example: Atlanta's five cybercenters. Ten more are planned over the next year. Shown here are two mile buffer zones around each center.
Example: The five current cybercenters overlaid on a poverty surface (% in poverty by census tract, 1990).

3. Use Internet GIS capabilities to provide distributed, locally relevant content. Internet GIS, such as these initial GSU projects, can provide neighborhood oriented content through several kinds of interactive mapping. Most exciting however, is the idea that these servers are not one-way (users downloading) but two-way; whereby users, or really participants can build up the site themselves through contributions, map corrections or updates, safety reports, etc. This is known as “distributed human computing” or more simply the power of the community.

Links to related work

AAG 2001 New York City: There were two sessions explicitly about the geography of the digital divide. You can find the abstracts of the nine papers presented here. Here is the complete Powerpoint presentation I gave (5.6MB). It includes a brief rundown of the concept and status of the digital divide, before outlining the situation in Atlanta.

Digital video local fieldwork. These DV clips show the main CTC in Atlanta. They also feature interviews with several of the people behind the Atlanta cybercenters. Coming soon!

Working paper, by Jeremy W. Crampton. This 6-page document was written to briefly describe the GIS project for Atlanta which is sketched above. Unfortunately I cannot make this publicly accessible, but email me if you would like to see it.