The views of South African marketers concerning consumer privacy

As part of the transition to democracy, the South African government has expressed its intention to table the Open Democracy Bill, which addresses the issue of the collection and dissemination of personal information. This bill will. in its present form. severely damage the direct marketing industry in South Africa. A literature review of consumer privacy has been completed. A survey of 246 marketers was undertaken. The results show that there are two distinct clusters. a pro-privacy group and an anti-privacy one. In this article some conclusions are developed that are of interest to policy makers and managers.


Introduction
In the infonnation age of computer databases, direct marketing and relationship marketing, the exchange of personal infonnation about individuals has become an integral part of the world economy.The search for competithe advantage has resulted in companies seeking individual level infonnation about their consumers.Consumers in tum have become increasingly disturbed by privacy invasions, particularly where infonnation is sought using scurrilous and unethical means.Phelps, Gozebach & Johnson ( 1994) report that press coverage of privacy issues in the five major American newspapers increased from ten articles during 1984 to I 05 during 1991.Privacy was one of the ten areas for marketing and public policy research mentioned by Mazis ( 1997).The American Direct Marketing Association recently listed consumer concern over privacy as the most important issue facing direct marketers today (Phelps, Gozenbach & Johnson, 1994: IO).The privacy issue is often linked to direct marketers, but is of relevance to any organization engaging in market research.
Because few legal restrictions apply to the gathering use of personal infonnation in the USA or in Canada, privacy is an ethical issue as well as a business strategy issue (Campbell 1979).According to Nowak & Phelps (1995), laws and regulations have produced more symbolic than real privacy protection.They call for professional and industry self-regulation to alleviate consumer doubts and concerns.
In a comprehensive report on privacy online in the USA, the Federal Trade Commission ( 1998) has recognised certain core principles of fair infonnation practice.These core principles require that consumers be given notice of an entity's infonnation practices; that consumers be given choice with respect to the use and dissemination of infonnation collected from or about them; that consumers be given access to information about them collected and stored by an entity; and that the data collector take appropriate steps to ensure the security and integrity of any infonnation collected.
As part of ongoing refonns in South Africa, the government has expressed its intention to table the Open Democracy Bill which addresses inter alia the collection and dissemination of personal infonnation.Specifically, the Bill requires the consent of a respondent in order to use infonnation for any purpose other than that for which it was obtained.Any sale of infonnation in the fonn of lists will be prohibited without the consent of each individual on the list.In its present fonn, the Open Democracy Bill will effectively destroy the direct marketing industry in South Africa, and seriously curtail any fonn of database marketing (The Star, May 4, 1996).There is a delicate balance between consumer privacy rights and industry interests, and care should be taken in developing and changing laws (Thomas & Maurer, 1997).
Despite the importance of the privacy issue, and its potential for ethical conflict, there has been little academic research addressing the topic, particularly from the marketer's viewpoint.As is the case with most marketing issues with ethical content, marketers would rather self-regulate than comply with stringent regulations.Without some research in this area however, South African business will be hard pressed to justify opposition to impending privacy regulations.It is unknown for example, whether South African marketers support privacy protection measures such as opt-out clauses, or would be willing to adhere to industry codes governing regulating consumer privacy.The aim of this research was to investigate the perceptions of South African marketers regarding consumer privacy and business infonnation practices.

concept of consumer privacy
Defining the ambit of the tenn .'co_nsumer privacy' is .rtant, particularly when cons1dermg whether or not ll11po .
f .h b •vacy has been breached.The subJect o privacy as een :died in the behavioural and social sciences, the legal ofession, and more recently, in an infonnation context.This :ultiple focus has resulted in various definitions depending on the perspective.Bauer & Miglautsch ( 1992: 8) call for a definition of consumer privacy 'to distinguish the concept from other similar concepts and clearly define what is meant by the tenn'.
Westin defines infonnation privacy as 'the claim of individuals, groups or institutions to detennine for themselves when, how, and to what extent infonnation about them is communicated to others ' (1967: 277).
Although this definition is a cornerstone of American privacy legislation, it has been criticised from two angles.Firstly, Westin's use of value-loaded tenns such as 'claim' makes it difficult to apply the definition impartially.Secondly, Westin's definition is concerned solely with infonnation privacy, whereas consumer privacy as a concept includes both physical and infonnation privacy.This dual nature of consumer privacy means for example, that the privacy invasion of a telephone call should be distinguished from the computer infonnation which resulted from, or led to, the telephone call (Goodwin, 1991 ).Goodwin's (1991) approach narrows the realm of privacy to a consumer perspective, and recognises the dual dimensions of consumer privacy (namely actions taken to secure information, and the use of the infonnation itself).She examines privacy ' ... in tenns of two dimensions of control.The first dimension includes the control of unwanted telephone, mail, or personal intrusion in the consumers' environment, whilst the second is concerned with the control of information about the consumer' (Goodwin, 1995: 10).Culnan's (1995) definition was the only one found to meet the criteria defined above.It examines consumer privacy from an information perspective, and takes into account the dimensions referred to by Goodwin ( 1991 ).The definition states that: 'privacy is the individual's ability to control the tenns under which personal infonnation -infonnation identifiable to an individual -is acquired and used' (Culnan, 1995: 29) The most visible shortcoming of the definition is that it does not refer to consumer privacy per se, but rather to individual privacy.For the purposes of the current research, the following definition of consumer privacy will therefore be employed: 'consumer privacy refers to a consumer's ability to control the terms under which personal infonnationinfonnation specific to that consumer -is acquired and used' -(derived from Culnan, 1995: 29).

Marketing activities and breach of consumer privacy
A fundamental aim of marketing activities is to satisfy customers.Yet, in trying to understand how customer needs can best be satisfied or directly targeting customers with focused product offerings, marketers (particularly those engaged in market research and direct marketing activities) may well be infringing on consumer privacy.The effective marketer must strive to balance the benefits received from customer information with the costs of invading customers' privacy to obtain and use such information.
In order to debate actions which constitute invasions into consumers' privacy it is useful to examine literature which explores dimensions of privacy in general.A model developed by Altman (1974) illustrates how individuals choose to either participate in or withdraw from social interaction to create an optimal level of privacy.Altman (1974) observes that privacy is bi-directional; an individual's privacy or lack thereof occurs as a result of the input of others and the output (or interaction) of the individual.From a marketing perspective, this implies that both consumers and companies determine the boundaries of invasions of privacy -individuals by their willingness to provide information, and companies by their methods of obtaining information and their usage of information provided.Altman's (1974) notion of an optimal level of privacy is also relevant when considering consumer privacy.Many consumers welcome relevant information received as a result of activities such as direct marketing, despite the accompanying invasions of privacy.One aspect of the model that cannot always be extended to consumer privacy is the assumption that individuals can choose to participate in a social action, or they can elect to withdraw.Unless the consumer has total control over the circumstances in which their details are used, many instances (such as telemarketing) render consumers relatively powerless to avoid intrusions without taking extreme measures (such as taking their telephones off the hook).Prosser ( 1960) observes that privacy can be invaded on four counts: -intrusion (physical invasion of a person's seclusion or solitude) -disclosure (public disclosure of private facts) -false light (incorrect representation of a person) appropriation ( use of a person's image or identity without pennission) An extensive evaluation of these dimensions with reference to consumer privacy was conducted by Nowak & Ph~lps (1992).Although all four actions are likely to pre~nt ethical dilemmas in tenns of the definition of consumer privacy provided above, Nowak & Phelps' (1992) evaluati~n w~ conducted from a legal perspective.The first d1mens10~ . of intrusion encapsulates marketing activities s~c~.as ~elev1S1on arch and tele-marketing in that both act1v1t1es ... reprerese 1• d • t sent an intrusion upon a person's seclusion or so ltu e or m o private affairs' (O'Brien, 1976: 7).Nowak & Phelps (1992) argue however that in most cases consu_me~ wo~ld be unable to legally argue a breach of priva~y usmg mtrus1on as an argument since television research is voluntary, and ~at telemarketing is not a great enough intrusion to quahfy as an offence (under US law).The use of surveillance cameras could however be ruled as an intrusion.Prosser's (1960) second dimension of disclosure may be violated by marketers who disclose confidential informati~n about customers.Laws in the United States prevent certain categories of information such as sexual relations or i_llness, and the use of confidential data from doctors from bemg revealed.In most democracies however, free speech laws (such as the First Amendment) tend to supersede attempts to restrict the use of individual data (Nowak & Phelps, 1992).
From a marketing perspective, Prosser's (1960) third dimension, false light, is closely linked to the correctness of data that companies hold on individuals.From an American legal perspective, plaintiffs who sue credit bureaux guilty of disseminating incorrect information about them must prove that the company acted with 'malice or reckless disregard for the truth' (McWhirter, cited in Nowak & Phelps, 1992: 50).Given that most companies are likely to be considered negligent, rather than malicious, most claims would be rejected.
Marketers who copy competitors• product and promotional strategies may be regarded as violating privacy based on Prosser's (1960) fourth dimension of appropriation; however this is a commercial issue rather than a consumer privacy issue.The appropriation by a marketer of a consumer's identity or image is an unlikely scenario.
A major drawback of Prosser's model is its consideration of privacy from a legal perspective.Not only does the model rely heavily on US law, but it ignores relational, behavioural and ethical aspects of privacy (such as those considered by Altman, 1974).Given the low level of legal remedies available, consumers who wish to protect their privacy are dependent on the self-regulation of marketers.
Integrative Social Contract Theory (JSCT) allows consumer privacy to be examined from an ethical, rather than a legal, perspective.ISCT suggests that consumer and marketing behaviour concerning privacy is governed by two differing contracts.The first agreement is that of a theoretical macro-social contract between marketers and consumers which describes a normative hypothetical contract -that is what should be.Within this first macro-social contract, parties form micro-social contracts between themselves which are highly context-specific.In certain situations, consumers will be happy to release more or less information, and marketers will be more or less likely to collect, use or sell data (Culnan, 1995;Cespedes & Smith, 1993).Culnan ( 1995) suggests that the macro-social contract is derived from three principles coined by Republican Edward J. Markey: knowledge, notice and no.According to these principles, consumers must be aware that their personal information is being collected (knowledge), that it is being shared (notice) and that they must be able to restrict sharing if they choose to (no).
Building on the foundation of ISTC, Milne & Gordon ( 1993) propose a model which essentially explains the formation of the micro-social contract.The authors suggest that consumers perform cost-benefit analyses when exchanging personal information for some benefit.Where benefits (information received) exceed costs (loss of privacy) consumers will willingly surrender personal information.Milne & Gordon (l 993) suggest that in addition to the macro-social con-S.Afr.J.Bus.Manage.2000,31( 4 ) tract and government legislation, consumers consider four factors: -targeting (is the consumer likely to receive mail of per.sonal interest?)volume (is the expected volume of mail to be received satisfactory?)-permission (will the company contact the consumer to request permission each time the information is used?) -compensation (does the company offer an incentive, either monetary or other, to consumers who surrender personal information?)Nearly all frameworks and models exploring consumer privacy have been developed for application in the United States.Given differing marketing information practices and legal environments, the needs of consumers and businesses can vary from country to country.Milne, Beckman & Taubman (1996) have formulated a conceptual framework to examine privacy in countries outside the United States, particularly those with less developed marketing information systems.Their model suggests that individual rights, demographics and an individual's experiences with direct marketing (which in tum is a function of the country's overall direct marketing infrastructure and technology) all affect consumers' attitudes to privacy.Individual rights refer to the norms that govern the macro-social contract suggested by Milne & Gordon ( 1993 ).
Testing of the model occurred in Argentina, where it was found that consumers' attitudes to privacy varied in terms of demographic variables such as age, education and income (Milne et al., 1996).The authors also report that the percep• tions of Argentinian consumers regarding their privacy rights, as well as their level of understanding of business information practices, contrasted greatly with their American counterparts.Given the similarities between South Africa and Argen• tina in that both countries have recently emerged from repressive regimes that cared little for individual rights and privacy, Milne et al. 's (1996) hypothesis that political repres• sion leads consumers to pervasively distrust government and information collection may well be applicable to South Af• rica.
In summary, the issue of determining whether or not a con• sumer's privacy has been breached is a complex one.Inva• sion of privacy depends upon whether a judgement is being made purely in terms of law, or whether it incorporates be• havioural and ethical aspects as well.Marketer's actions are less likely to be judged infringements of privacy when adopt• ing a legal perspective (Nowak & Phelps, 1995) than when viewing the action from an ethical perspective such as the In• tegrative Social Contract Theory (Culnan, 1995;Donaldson & Dunfee, 1994).Norms governing consumer privacy are in• fluenced by factors such as a society's direct marketing infra• structure and technology, and its political and legal systems (Milne et al., 1996).Apart from a societal normative ap• proach to what constitutes an invasion of privacy, consumer• specific factors affect the nature of the micro-social contract (Donaldson & Dunfee, 1994) and include consumer percep• tions of the costs and benefits of the invasion of privacy (Milne & Gordon, 1993) as well as demographics and experi• ence with marketing practices (Milne et al., 1996).Consumer attitudes vary depending on the reputation of the company collecting the infonnation.and the consumer's perceived relevance of the infonnation (Wang & Petrison. 1993).

Marketing's response to consumer concerns
Empirical research conducted shows that American consumers are becoming increasingly concerned about privacy invasions (Katz & Tassone, 1980: Equifax Harris, 1995).Despite the level of concern, many consumers have little knowledge of what information is collected, used and sold by business (Nowak & Phelps, 1992).Given the high levels of consumer concerns, and consumers' reliance on marketers to protect them from invasions of privacy.there is a paucity of research investigating the responses of marketers to the issue of consumer privacy.
Toe major study investigating this topic was conducted by Taylor, Vassar & Vaught ( 1995) who examined the privacy beliefs of marketing professionals and found that marketers' privacy beliefs differed across business sectors, and that marketers who engaged in data collecting activities were less sensitive to privacy issues.Marketers are generally against government regulation, and believe that where companies employ safeguards.consumer privacy is protected (Taylor et al., 1995).
In an attempt to alleviate consumers' privacy fears, and stave off privacy legislation.many companies and industry associations have formulated strategic responses to consumer privacy.The three most common of these are name removal systems, opt-out clauses and information practice codes.
Name removal systems are normally maintained by industry associations: in South Africa the Direct Marketing Association (OMA) plays this role.Consumers who do not wish to receive marketing communications can request their names to be placed on a database which is circulated among the association's members who remove the names should they appear on any lists.Given consumers' lack of awareness of such remedies (Nowak & Phelps, 1995), the system is not widely used, and even when in use, has its problems.
In a discussion on the failure of the system from the marketer's perspective, Posch notes that: ' ... the OMA cannot speak for all individual direct m~keters ... The ... service cannot mitigate the compilation of dossier-type information from government records.Finally, even when there is success in remov-in~ a name, that success is temporary, for the name will be back on other lists after subsequent purchases ' (1994: 64).Opt~out clauses allow consumers to prescribe that their personal information may only be used for the purpose for which it was collected.Although such a clause allows consumers to exert some control over their personal infonn;i.tion,it restricts o~ly the transfer of information.Primary colkction of data without the .
• .consumer s knowledue may sttll be unprotected.ln response t h . .: _further method of regulation occurs where companies an ~dustry bodies develop codes oovernino fair information ~ractice.Although codes mav be a~ import~nt starting point in respect' • mg consumer desires for privacy, the success of the co?e depends upon its implementation.Carusone, Brown & ~1se ( 1991) researched the response of tele-marketers to their mdustry's_ code of ethics (which included regulation of consumer.pnvacy), and found that 'tele-marketers are not as committed to the industry standard of consumer privacy as the framers of the code feel they should be' (Carusone Brown & Wise, 1991: 12).'

Research propositions
Th~ purpose of this study is to identify and evaluate the be_hefs of South African marketers regarding consumer pnvacy and business information practices.
. ~i~en tha_t experiences of repressive regimes that suppress md~v1dual nghts leads to a pervasive distrust of government (Milne et al., 1996), and that marketers are against government involvement because they are more aware than consumers of the safeguards legitimate companies employ to protect consumer privacy (Taylor et al., 1995), proposition one suggests that: Pl: South African marketers are against government legislation protecting consumer privacy.Most of the press coverage given to the consumer privacy issue in South Africa has resulted from the South African government's intention to table the Open Democracy Bill.Proposition two suggests that: P2: South African marketers view consumer privacy as a government-generated issue, and not one of prime importance to consumers.An alternative to legislation of consumer privacy is for marketers to engage in self-regulation.This calls for the implementation of systems such as opt-out clauses, name removal systems, and information clauses.The latter options must be implemented by a controlling body, but individual companies can implement opt-out clauses autonomously.Proposition three asserts that: P3: South African marketers are willing to accept methods of self-regulation such as opt-out clauses.
Given the lack of current regulation addressing consumer privacy (either statutory or by self-regulating bodies) and that it has only received attention in recent years (both locally and abroad) proposition four states that: P4: South African marketers do not view privacy as an integral part of infonnation gathering and use.Taylor et al. (1995) suggest that marketers' privacy beliefs differ across industry associations, and that marketers who engage in data collecting activities are relatively less sensitive to privacy issues.Propositions five and six address these findings from the South African perspective: P5: Consumer privacy beliefs of South African marketers differ across business sectors.P6: South African marketers who engage in activities which may infringe on consumer privacy are less concerned than those marketers who do not engage in such activities.

Methodology
This study consisted of a mail survey aimed at South African marketers.

Sample
The population consists of South African marketing professionals.The sample frame was based on the Direct Marketing Association's (OMA) mailing list.The DMA's list includes a cross-section of South African marketing professionals who use individual level data.This is not representative of the entire South African marketing professionals population, but Taylor, Vasser & Vaught ( t 995) found that marketers who did not use individual infonnation had little understanding of the issue and responded to issues as consumers and not marketers.The DMA list is not restricted to direct marketers, but is concentrated towards marketing professionals who use individual level consumer data.
Since the Open Democracy Bill is aimed at marketers who collect individual level data, it is more useful to examine marketers that do collect than those that do not.The DMA mailing list covers a broad range of industries, size of businesses and extends across the entire geographic region of South Africa.There were 595 marketers on this list.

Research instrument
The questionnaire was divided into two parts.Section A classified the respondent into a categorical variable according to what business sector they worked in.It also collected infonnation on various data activities perfonned by marketers according to previous studies such as that of Milne, Beckman & Taubman (1996).Section B included 14 statements that dealt with legislative responses to protect consumer privacy, fair information practices 'embodied in self-regulation, consumer privacy and its affect on South African businesses and general privacy beliefs.These statements were drawn from the Equifax Harris Consumer Privacy Survey as adapted by Milne, Beckman & Taubman (1996) and Taylor, Vassar & Vaught (1995) as well as from Nowak & Phelps (1992).An addressed returned envelope was included with all the questionnaires.The questionnaire was pre-tested in a pilot study.The sample of this study comprised three academics and five marketing professionals.Meetings of the DMA's Privacy Task Group were also attended to improve the researchers' understanding of local issues.As a result of these meetings, two statements were added to the questionnaire.They were, firstly, South African consumers are concerned about threats to their personal privacy and secondly, a consumer's name and address is public domain data.The fourteen statements are illustrated in Table 2.

Data analysis
Section A of the questionnaire used categorical scales and section B used a five-point Likert scale as this was the scale used in most privacy studies (Katz & Tassane, 1990).The scale ranged from 5 = strongly agree, to I = strongly disagree.
The data was captured twice using a well-known statistical software package to ensure accuracy.Non-response variables were entered for missing responses.Analysis of Variance (Anova) was run to test for differences between different industries and different forms of data activities.Where differences were found, multiple comparison tests were administered to discover where the differences lay.Since the descriptive statistics indicated two diametric groups, a cluster analysis was performed to better understand this division.

Respondents
Of the 595 questionnaires mailed, 246 were returned by the cut-off date, representing a response rate of 41 %.Three questionnaires were returned unanswered and three incomplete questionnaires were rejected.There was a good spread of business sectors with the majority falling into the direct marketing and banking sectors.These industries are the major users of individual level data.

Privacy beliefs of South African marketers
Table 2 shows the beliefs of South African marketers regarding privacy issues.
The results reveal that there is a large spread of views across the frequencies indicating little similarity between the respondents' beliefs on many statements.Statements that dealt with self-regulation such as opt-out clauses and infonning consumers of information use enjoyed the most homogenous responses.Government and legislative statements also attracted more homogenous responses and were mainly against government intervention.Despite this concentrated 'yes' to privacy planning, respondents were severely divided on whether consumer privacy was actually a threat to profits.Broad privacy issues such as whether a name and address was in the public domain also drew mixed responses.Overall, the broad range of responses indicates some form of conflict The government should limit the amount and type of consumer information that can 103 60 28 28 16 2.12 be collected Scale: 5 = Strongly agree; 4 = Agree; 3 = Neither agree or; 2 = Disagree; I = strongly disagree between opposing camps, rather than a homogenous response from marketing professionals.
The statements that generated the most positive responses were the two self-regulation issues, the inclusion of an optout clause, and the right of consumers to access their data.The acceptance of these statements by respondents indicates broad support for the principles of knowledge, notice and no (Culnan, 1995).The opt-out clause is the consumer's chance to say 'no' and scored a mean of 4.33.Notice-based statements such as being told about information use and sharing also receive strong support.The principle of knowledge is met by allowing consumers to access their data and this statement had a mean of 4.29.All these statements were significantly positively correlated with each other at the alpha of 0.05.Correlations were .24 between giving consumers an opportunity to opt-out and their right to access their data.
Telling consumers their information is going to be shared and telling them how their personal information will be used had a correlation of .60.
Responses to general privacy issues lacked the definite nature of self-regulation issues.In the case of consumer privacy rights, South African marketing professionals disagreed greatly.The statement that South African consumers are concerned about threats to their personal privacy was negatively correlated with the four statements: a consumer's name and address is public domain data; it is acceptable for direct marketers to buy and sell names and addresses of consumers without their knowledge; consumer privacy rights are adequately protected by law and business practices; and consumer privacy is a threat to profits.All correlations were significant at the alpha of0.05.
The most negative responses were received for statements dealing with government intervention and legislation to pro-tect consumer privacy.Marketers felt strongly that the government should not do more to protect consumer privacy, not appoint a privacy commission and most definitely not limit the collection or use of consumer data.All these statements were strongly positively correlated with each other.Marketers were close to unanimous in their rejection of government controls and efforts to protect consumer privacy.

Differences between business sectors
Analysis of variance was used to investigate whether marketer's privacy beliefs differed according to what sector they worked in.This is shown in Table 3.
The Anova revealed that there were significant differences between industries for the following statements.It is acceptable for direct marketers to buy and sell names and addresses of consumers without their knowledge (p = 0.147); the government should limit the amount and type of consumer information that can be collected (p = 0.002); companies should have to get written permission from consumers before they sell or rent personal information {p = 0.007); and the government should do more to protect the privacy rights of consumers (p = 0.239).Scheffe's method of multiple comparisons as well as Least Squares Difference (LSD) tests were run to test where the differences lay.Banking and insurance marketers considered the buying and selling of individual level information as less acceptable than direct marketers.This may be due to the sensitive financial data managed in the industry.Marketers from the industrial sector considered the buying and selling of consumer's names and addresses as less acceptable than any other sector.Industrial marketers were more inclined to accept government limits on the use and collection of

Table 3 Analysis of variance between business sectors
South African consumers are concerned about threats to their personal privacy Companies should consider consumer privacy issues when planning information gathering and use A consumer's name and address is public domain data It is acceptable for direct marketers to buy and sell names and addresses of consumers without their knowledge Consumers should be given a legitimate opportunity to opt-out of subsequent uses of data The government should limit amount and type of consumer information that can be collected Consumers should be told if their information is going to be shared with third parties Companies should have to get written permission from consumers before they sell or rent personal information The government should do more to protect the privacy rights of consumers Consumers should be told how their personal information will be used There should be a government commission or system to protect consumer's privacy Consumer privacy rights are adequately protected by law and business practices Consumers should have the right to access.verify and correct information about themselves Consumer privacy is a threat to profits individual data than any other sector.The publishing and media sector also considered government limits in a more favourable light.More than any other sector, consultants were in favour of companies having to get written permission from consumers before they sell or buy data.This difference is limited by the small number of consultants in the survey and cannot be interpreted too broadly.Marketers working in the nonprofit sector considered consumer privacy rights to be adequately protected by law and business practices.This belief may be linked to non-profit firms' inability to afford the expenses of opt-out clauses and privacy protection (Posch, 1994).

Differences between firms that engage in data activities and those that do not
Analysis of variance was used to test for differences between marketers who engage in various data activities and those that do not.Activities included the primary collection of data, purchase of data, using data for developing marketing strategies.for advertising and finally for direct selling such as mail order.Respondents' activities were categorised as binary, that is, they either participated in the activity or they did not and respondents were coded as such.The summary statistics for the Anova are shown in Table 4.
There were no significant differences at the alpha level of 0.05 between marketers who collected primary data and those that did not.There were also no significant differences between marketers who did or did not use individual data for advertising.Significant differences existed between marketers who purchased consumer data, sold it, used it for developing marketing strategies.direct selling and those that did not participate in these activities.

Cluster analysis
Our analysis has suggested that respondents fell into two groups, one pro-privacy and the other anti-privacy.A cluster analysis using K-Means clustering was run and two distinct clusters emerged.Table 5 lists the actual means for each cluster.
Cluster one was named anti-privacy.It consists of 133 respondents who resist any form of privacy legislation.They view information practices such as buying and selling information as legitimate.Any restrictions such as limiting data collection or the government protecting privacy rights are rejected.Despite this, anti-privacy marketers are willing to provide an opt-out clause and to allow consumers access to their data.
Although they do not really plan for consumer privacy problems, they are open minded to the problem and willing to compromise on some issues, as long as it does not allow for a blanket restriction on their activities.Cluster two was named privacy advocates.This cluster, consisting of 91 respondents, strongly believe that South African consumers are concerned about their privacy and that their companies should plan for privacy problems.Although luke-warm to legislation to re-Table 5 Cluster means 163 strict data collection and use, they believe that consumers should be allowed to opt-out of future data use, be informed of that use, and have access to the data.Despite these pro-privacy beliefs, they do not see privacy as a threat to their profits.
A cross-tabulation was performed on business sector versus clusters.This is shown in Table 6.
Over 72% of direct marketers fall into the anti-privacy cluster.Banking and insurance respondents were more divided and only 57% fell into the anti-privacy cluster.This may be a reflection of their greater sensitivity to consumer privacy because of the delicate nature of financial data.Respondents from industries that make little use of consumer information such as the industrial sector fell almost exclusively into the pro-privacy cluster.
A cross-tabulation was run against the different data activities, and is shown in Table 7.
This revealed a general grouping of marketers who engaged in different activities who fell into the anti-privacy cluster.This suggests that marketers who engage in activities that use individual level data are less sensitive to consumer privacy than marketers who do not.

Discussion
The findings of this study will be related to the propositions formulated.
P1: South African marketers are against government legislation aimed at protecting consumer privacy Findings indicate that South African marketers as a whole are against government legislation to protect privacy.This finding is supported by similar results in the Taylor, Vasser & Vaught ( 1995) study.It also supports Milne, Beckman & Taubman 's ( 1996) conceptualisation that government attempts to protect consumer privacy is resisted in countries that have experienced repressive regimes in the past.Thus, proposition one is supported.P2: South African marketers are willing to accept methods of self-regulation such as opt-out clauses Self-regulation issues such as opt-out clauses received overwhelming support.What was more important was that the support was near universal among all industries and there were very few exceptions.Therefore, proposition two is supported.

P3: South African marketers' privacy beliefs differ across business sectors
South African marketers' privacy beliefs do differ according to what industry they work in.This supports Taylor, Vasser & Vaught's (1995) findings that American marketers differ.Thus proposition three is supported.
P4: South African marketers who engage in activities which may infringe on consumer privacy are less concerned about consumer privacy than those marketers who do not engage in such activities Beliefs do differ, but to a far more limited extent than in different _industries.f?ifferences were confined to purchasing data, sellmg data, usmg data to develop marketing strategies and direct selling.Differences were further confined to certain statements.Differences were concentrated on government and self-regulation issues.Essentially, marketers who S.Afr.J. Bus.Manage.2000,Jl( 4 ) did engage in data activities were less sensitive to consumer privacy and most fell into the anti-privacy cluster.Wher there were significant differences, marketers who use indivi~ dual level data, did not agree with attempts to protect consumer privacy as much as marketers who do not use data.Thus, proposition four is supported.PS: South African marketers view consumer privacy as a government generated issue and not one of prime importance to consumers South African marketers were unsure whether privacy was of prime concern to consumers.Nevertheless respondents who fell into the privacy advocate cluster strongly believed that consumers were concerned.These respondents tended to work outside the direct marketing industry and engage in few data activities.Research into consumer beliefs of privacy would be better able to substantiate this proposition.The vague and unsure nature of marketer's beliefs regarding this proposition does not allow the proposition to be supported or rejected.Further research is needed.
P6: South African marketers who engage in activtties which may infringe on consumer privacy are less concerned than those marketers who do not engage in such activities Marketers believe that privacy is integral to planning infonn• ation gathering and use.This belief is contradictory because marketers do not see consumer privacy as a threat to profits.Marketers may be expressing a desired state of affairs (such as the macro-social contract (Donaldson & Dunfee, 1994) or marketers may view meeting consumer privacy fears as a corporate social responsibility exercise.Nevertheless, marketers do view consumer privacy as an integral part of planning information gathering and use, and so proposition six is not supported.

Conclusion
This research aimed to investigate the privacy beliefs of South African marketers.Generally, it was found that South African marketers were not overly concerned about consumer privacy and did not see it as a threat to profits.However, they were concerned about any form of government intervention to protect consumer privacy.Forms of self-regulation such as opt-out clauses and allowing consumers to access their data were widely supported.This finding is similar to Taylor, Vasser & Vaught's ( 1995) study of American marketers.Marketers were particularly unsure about general privacy issues and whether consumer rights were adequately protected by the law and business practices.
Marketers were not particularly homogenous in their re• sponses to consumer privacy.Their privacy beliefs differed according to their business sector.This reflects the context• specific nature of consumer privacy (Cespedes & Smith, 1993) as marketers in the banking and insurance sector were more sensitive to privacy concerns than direct marketers.
Marketers who participated in various data activities were also found to be less concerned about privacy issues than marketers that did not use personal data.
This disparity between marketers is further reflected through the cluster analysis which identified two clusters -one tenned anti-privacy, the other privacy-advocates.The clusters may represent the personal fears of threats to the privacy of individual marketers.rather than their business concerns.Overall, South African marketers do not view consumer privacy as a major threat.but the Open Democracy Bill has raised fears of government intervention.
impending South African Open Democracv 8 ~_the ~MA has proposed~the implementation of legislatio~ mh mg It mandatory to cite an opt-out clause in all cases w ere consumer data is collected. S.Afr.J.Bus.Manage.2000,31(4) consumer privacy issues when planning informalion gathering and use A consumer's name and address 1s public domain data II is acceptable for direct marketers lo buy and sell names and addresses of consumers without their knowledge Consumers should be given a lcgillmalc opportunily In opt-oul of subsequent uses of data The government should limn amounl and t} pc of consumer information that can be collected Consumers should be told ifthc1r information 1s going to be shared with third parties • b f they sell or rent personal information Companies should have to get \Hillen perm1~sion from consumers e ore The government should do more to protect the privacy rights of consumers Consumers should be told how their personal information will be used There should be a government comm1ss1on or system to protect consumer's privacy Consumer privacy rights are adcqt1atcly protected by law and business pracuces Consumers should have the right to access.vent) and correct information about themselves Consumer privacy 1, a threal to profits Scale: 5 Strongly agree.I -Strong!

Table 1
Table l indicates the profile of respondents by business sector and by data activities.Profile of respondents by business sector and data activities

Table 2
Privacy beliefs of South African marketers

Table 4
Analysis of variance between firms that engage in data activities and those that do not

Table 6
Cross tabulation of business sector versus

Table 7
Cross tabulation of activities versus clusters