Flaherty (1994) in relating his experiences with privacy protection and GIS in a national and provincial context has said that GIS technology is “not personal data intensive” and that information privacy issues are always solvable by applying fair information practices.

GIS has the power to integrate diverse information from multiple sources. Some of the data are of a personal nature where individuals may be identified or identifiable while others are of a spatial nature that may be used to locate individuals through geocoded data such as a home address. The privacy threat is from the new inferences that may be made by correlating geographic information with personal information.

But what is privacy? Professor Arthur Miller of the Beckman Center for Internet and Society, Harvard Law School has described privacy as an intensely, perhaps uniquely, personal value. The word stems from a Latin root privare which means “to separate”. To want privacy is to want to be separate, to be individual. Another meaning of the Latin is “to deprive”, and privacy also means leaving something behind (See Beckman Center for Internet and Society and in particular the course syllabus for ‘Privacy in Cyberspace’ at http://eon.law. harvard.edu/privacy99/ syllabus.html).

Indeed Roger Clarke (1999) has identified four privacy interests – information, territorial, bodily and communications privacy. Information privacy relates to the interests of the individual in controlling the information held by others about that person while territorial privacy is the interest in controlling intrusion into aspects of behaviour especially in regards to sensitive issues. Bodily privacy is the freedom from interference with one’s person while communications privacy is the freedom from surveillance and privacy of communications.

Location-based services (LBS) rely on the key ingredients of time and space. LBS may be considered to be no different from geodemographics, an information technology that enables marketers to predict behavioural responses of consumers based on statistical models of identity and residential location (Goss 1994, 1995).

LBS have become commonplace because of the use of geocodes and GPS and other mobile communication and tracking technologies. LBS inferentially involves the tracking of people through the use of credit card data that may result in profiling exercises, statistical modelling, and pattern analysis. More generally, GI Science using such technologies may give the game away as to who we are, where we are, and what we have been doing either by way of speech, purchases or simply being at a location. There is one view that without legislation to curb the (mis)use of such data there would be chaos. However, an equally compelling but opposing view is that there should be no legislation but rather just self-regulation by industry itself (Westin 1967, 1971).

Private sector commercial applications of both GIS and LBS are perhaps the fastest growing areas of business and this has fed the need for more data. For example, in order to maintain a competitive edge, marketers need good databases to make their decisions while simultaneously handling geographical data efficiently. Patterns, relationships, and trends become clearer when depicted visually in graphs, charts, and maps rather than just columns of numbers or text.

A database called EQUIS, developed by the US National Decision Systems, “maintains a database of financial information for over 100 million Americans on more than 340 characteristics including age, marital status, residential relocation history, credit card activity, buying activity, credit relationships (by number and type), bankruptcies, and liens. This information is updated continuously at a rate of over 15 million changes per day,” (Curry 1992, p. 264).

In 1993, Equifax and National Decision Systems announced Infomark-GIS – a fully integrated GIS specifically designed for marketing applications and decisionmaking (Equifax National Decision Systems 1993). There are also other U.S. companies are also engaged in the collection, processing and storage of data pertaining to individuals (A search of those firms classified under the Standard Industrial Classification (SIC) code 7374 for companies engaged in marketing and business research services yielded approximately 50 companies). These firms obtain consumer information from credit bureaus, public records, telephone records, professional directories, surveys, customer lists and other data aggregators.

In view of the commercial market for data, databases and data aggregation services, Curry (1994) has suggested that the use of geospatial technologies will produce multifaceted problems that would similarly require multi-dimensional solutions. The concerns raised include the fact that the technologies consist of and promote the widespread availability of unregulated data. This leads to the difficulty of regulating data matching that must take place if the geospatial tools are to produce meaningful results. Further, geospatial technology is inherently visual but this strength also exposes a major weakness in that it may produce map inferences that may be both statistically and ecologically fallacious. Finally, there is the altered expectation of privacy rights because the case law may promote an erosion of those aspects of life where a person can feel safe, secure from search and surveillance, and most importantly information kept private.

Geospatial technologies are said to do best at the intersection of location, time, and content. But each of these elements of location, time, and content tends to produce tensions of their own. For example, in regard to letting people know where they may be and to keep this fact hidden because they may not wish to be found. Here we have a technology that is employing the power of the ‘place’. GISbased geostatistical models using locations and space have been used to study the home range of animals, for instance. In an analogous way, geographical profiling may be applied to the “home range” of predatory humans who may have inadvertently patterned habitual routes when stalking potential victims.

Leipnik et al. (2001) reported on the important role of geospatial technologies in investigating and gathering evidence of a locational nature in order to convict a serial killer. When the serial killer Robert Lee Yates was arrested in April 2000 in Spokane County, Washington State, he is reported to have told his wife to “destroy the GPS receiver.” This was because there were incriminating data on it showing the 72 waypoints associated with several journeys that had been used in disposing the bodies of the victims. Indeed, this example demonstrates the extent to which GIS and GPS technologies are permeating society and their use – both by law enforcement agencies as well as criminals. However, a court appeal could challenge the validity and use of the GPS data. The argument could be that using the GPS to track suspects without their knowledge might involve an invasion of privacy rights as well as not meet the legal test of finding the “least obtrusive means” for police to gather information about a suspect. But, in counter argument, it may be submitted that sometimes consent needs to be conspicuously absent in cases where suspicions of the ‘quarry’ are not to be aroused prematurely.