MANAGERIAL ACCESS TO INFORMATION by MICHAEL NEWMAN B.Sc , University of London, 1969 M.Sc, University of London, 1974 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Faculty of Commerce and Business Administration) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA December 1980 © Michael Newman, 1980 I n p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f t h e r e q u i r e m e n t s f o r an a d v a n c e d d e g r e e a t t h e U n i v e r s i t y o f B r i t i s h C o l u m b i a , I a g r e e t h a t t h e L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e a n d s t u d y . I f u r t h e r a g r e e t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y p u r p o s e s may be g r a n t e d by t h e he a d o f my d e p a r t m e n t o r by h i s o r h e r r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l n o t be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . D e p a r t m e n t o f /jF7/k t^"t&yL£. The U n i v e r s i t y o f B r i t i s h C o l u m b i a 2075 Wesbrook P l a c e V a n c o u v e r , Canada V6T 1W5 DK-fi (2/79} - i i -ABSTRACT The purpose of this study was to examine managerial access to i n -formation within organizations and i t s relationship to certain organiza-tional variables. The study had three objectives: 1. to develop and use a simple framework within which the l i t e r a t u r e on access to information could be integrated; 2. to test s p e c i f i c hypotheses which l i n k important organiza-tional variables with access to information; 3. to suggest prescriptions, based on the findings, for im-proving managerial access to information which can be used by organizational and information system designers. A framework was developed to describe how access to information in organizations i s controlled d i r e c t l y , by imposing rules, and i n d i r e c t l y , by erecting barriers to the retr i e v a l and use of information by organiza-tional members. Data on the regulation of access, organizational v a r i -ables, and managers' characteristics were collected by means of a struc-tured questionnaire from 170 middle managers in B r i t i s h Columbian organ-izat i o n s . In addition, f i f t y - t h r e e interviews with middle managers were conducted in the Vancouver area. Fourteen hypotheses l i n k i n g access to organizational variables were derived on the basis of the framework. Additionally, other r e l a t i o n -ships concerning the direct and indirect regulation of access were pro-posed. The study showed that access to work-related information i s regulated in organizations largely i n d i r e c t l y whereas access to personnel information i s often governed by elaborate rules. Additionally, the per-ceptions of access were found to be inversely related to the inconveniences of retrieving information but was almost independent of the problems of using information. Several of the relationships hypothesized between the organizational variables and access to information were supported. Access was somewhat poorer in larger organizations, as hypothesized. Contrary to expecta-t ions, however, access was found to be s ignif icant ly better in organiza-tions with more levels of authority. In organizations where an attitude of trust and openness is prevalent, s ignif icant ly better access to infor-mation was found. This was also found to be the case in organizations where sharing information is an accepted "norm". The use of computer technology was associated with greater barriers of access to information managers do not need for their jobs. The findings were used to suggest prescriptions for improving man-agerial access to information. For example, i t was suggested that as access is governed more by retrieval problems than by problems in using information, the systems designer should concentrate on minimizing the inconveniences of retrieving information. It was further suggested that because of the greater problems of access associated with s ize , managerial access could be improved in larger organizations by providing more effec-tive f a c i l i t i e s to promote access to information. The study concluded with a discussion of the research methodology, pointing out i ts advantages and l imitat ions, and with suggestions for further research into managerial access to information. - iv -TABLE OF CONTENTS Page Abstract i i Table of Contents iv L ist of Tables vi L ist of Figures v i i Acknowledgements v i i i Introduction 1 Chapter One: A Framework for Studying Access to Information. . 6 I. Direct Regulation of Access to Information 7 II. Indirect Regulation of Access 27 Chapter Two: Determinants of Access 42 I. External Regulation of'Information 43 II. Structure 46 III. Attitude to Data Sharing • . 56 IV. Technology of Access 60 V. Other Determinants 74 Chapter Three: The Study 79 I. The Hypotheses 80 II. Data Collection 87 Chapter Four: The Findings 97 I. The Sample 97 II. Profiles of Access and Authority . 100 III. Constituents of Access 115 IV. The Determinants of Access 123 V. Fac i l i t ies and Strategies to Promote Access to Information 157 Chapter Five: Prescriptions for Design and Future Research. . . 166 I. Designing Organizations and Information Systems for Improved Managerial Access 166 II. Prescriptions for Future Research . . 184 Bibliography 188 Appendix I: Questionnaire and Covering Letters 197 - V -Page Appendix II: Barriers to Access 210 Appendix III: Intercorrelations of Barriers for Case 3 211 Appendix IV: Kendall's Coefficient of Concordance 212 Appendix V: Access and the Barriers to Access 213 Appendix VI: Techniques to Prevent Access 215 - vi -LIST OF TABLES Page Table Number T i t le I . . .Composition of Total Responses by Industry 98 II . . .Composition of Total Responses by Department. . . . 98 III . . .Composition of Interview Responses by Industry. . . 99 IV . . .Composition of Interview Responses by Department 104 V . . .Summary of Techniques to Prevent Access 124 VI . . .Summary of Results 124 VII . . .Designing Organizations and Information Systems for Improved Managerial Access - Summary 167 - v i i -LIST OF FIGURES Page Figure Number T i t l e 1 . . . Information Categories 9 - v i i i -ACKNOWLEDGEMENTS Acknowledgements are due to - My committee, who tolerated being bothered by details and squeezed to deadlines so graciously - Colleen Colclough, who made sense of the f inal draft on the typewri ter - my supervisor, Ilan Vertinsky, an animated source of inspiration throughout hours of discussion, and to whom I always had access - my wife, G i l l , without whose academic assistance, personal support and domestic talents the work would have been more pain-ful and, indeed, possibly incomplete. The shortcomings of the f inal work, however, I acknowledge as my own. - 1 -INTRODUCTION The aim of this study is to explore managerial access to informa-tion within organizations and its relationship to specif ic organizational variables. The study does not attempt to deal with access to al l re-sources in the organization''': the concern here is with access to a specif ic commodity - information. Although much of the discussion of access to information is applicable to al l levels of employees, the empirical work concentrates on middle management only, focussing atten-tion on their access to internal information. The equally important questions concerning access to organizational information by those out-side the organization are not direct ly covered, although i t is hoped that several of the results will find application in this area. Forrester is one of the few writers to alert us to the importance of access to information. He suggests that "organizations can be seriously handicapped by the loss of energy consumed in the struggle for information", and proposes that, as a general pr inciple, new organiza-tions should allow wider and more convenient access to information than is normally practiced (Forrester, 1965). Other writers, when discussing increased interdepartmental access to information, have offered a counter argument, claiming that widespread access to information can be det r i -mental to individuals and organizations (Ackoff, 1967; Argyris, 1971). While there is no agreement on whether access to information improves performance or not, there is a consensus that access to information plays Mechanic (1962) discusses access to information, persons, and instru-mentalities by lower-level participants. - 2 -an important role in shaping organizational behaviour. It is clear that a lack of access to information needed to perform a task wil l result in ineffective performance. However, while access to information is an impor-tant variable in determining organizational behaviour, l i t t l e is known about the factors which influence access and the mix of strategies that organizations use to control or promote access. The major objective of this study is to investigate these access patterns and their determinants within Canadian organizations. Organizations offer formal authority to individuals to access organ-izational resources, including information (Mechanic, 1962). Associated with a given position in an organization is the authority to access cer-tain types of information and the lack of authority to access other types of information. In order to allow managers to perform their tasks effect ively , the organization must at least recognize their need for formal authority to secure access to essential information (Barnard, 1938:175). This authorization is referred to in this study as the direct regulation of access to information. Forrester, writing about the problems caused by a lack of access authority, comments, "Most persons in most organiza-tions feel that they do not have access to al l of the information they need. Sometimes they lack the information speci f ica l ly needed to accom-plish their duties" (Forrester, 1965). In their findings on access to personal information, Westin and Baker (1972:430-431) use the expression "rights of access" to refer to the authority given to individuals to access their personnel records. Formal access, however, is not a suff ic ient condition. Managers must be capable of both retrieving and using information (Dhalakia and Sternthal, 1977). For example, managers may be authorized to receive - 3 -accounting reports but, because the reports are'held in another c i ty , access to this information is made extremely inconvenient (Simon et a l . , 1954:61). In this example geographic location was an effective barrier to access. Gerstberger and Allen (1968) use the expression "channel accessibi l i ty" in a similar study of data choice among engineers. A l -though the term is not clearly defined, i t would seem from the context that i t too is a measure of geographic location. Additionally, other barriers to access have been identif ied such as timing and incompatibility of information sources (Simon et a l . , 1954:61-63) and, in a more recent study, an individual 's ignorance of the existence of his own personal record (Westin and Baker, 1972:431). These and other restrictions or barriers are the ways an organization regulates access to information in -direct ly . Thus, although there may not be a rule (direct regulation) governing access to information, the organization may choose to regulate access to this information by manipulating costs associated with its retrieval and use ( i . e . , constructing effective barriers to access). An organization's direct and indirect regulation of access to in -formation provides us with a structure by which access can be studied. However, although this structure would be a contribution to the l i terature on access (represented largely by the references given above), more impor-tant contributions can be made by applying the structure to different areas of study. F i r s t l y , the study of managerial access to information gives us a way of evaluating management information systems. By analys-ing managerial access to information, i t should be possible to assess the effectiveness of the current information system as viewed from a manager's perspective. If common barriers to access are found in organizations then general recommendations can be made to systems designers allowing - 4 -them to concentrate on minimizing those barriers that form the greatest inconveniences. In addition to measuring a manager's access to an organ-izat ion's information, the study also examines the effects of technology of access on the manager's view of access. For instance, does computer technology reduce some barriers to access while increasing others? The relationship between the maturity of the information system and access to data is also studied. F ina l ly , the study measures managerial access to personnel records from the perspective of both direct and indirect regula-tion of access. This could corroborate the view that general secrecy is reported to surround middle managers' grades and salaries (Forrester, 1965). The same data could also reveal what rights of access exist in different organizations (Westin and Baker, 1972). By measuring access to personnel information we have a more useful perspective than that of privacy, a term that has always been d i f f i cu l t to define (Sieghart, 1976:12f). Secondly, the study relates organizational variables to access and provides us with insights into organizational theory and behaviour. Several authors have identif ied information as one of the key ingredients of organizational effectiveness (Steers, 1977; Galbraith, 1973). How-ever, information is worth l i t t l e i f i t is not made available to the ind i -viduals who need i t to perform their duties, and so access to informa-tion is a better indicator of the effectivness of an organization than the sheer volume of information available CGalbraith, 1973; Forrester, 1965). Speci f ica l ly , some of the more important organizational variables are used in this study of access. As an example, the relationship between the size of the organization and access to information is explored: do large organizations have special problems in providing effective access - 5 -to data? Are there inherent barriers to access associated with size and how can organizational designers minimize the effects of these barriers? Thirdly, the study of access has implications for policy analysis. It is expected that the authorization and convenience of access to par t i -cular types of information varies greatly across companies. Should the current policies be changed to ref lect a more open attitude to information access or is access to information inadequately regulated in some organ-izations? An attempt will be made to apply the results to answering these and other questions by suggesting general policies that organiza-tions can adopt. The study proceeds as follows: Chapter One describes in detail the structure of the access model used in the rest of the study. The chapter examines the l i terature from several areas as i t applies to direct and indirect regulation of access to information. Chapter Two looks at several organizational and other variables and their relationship to access. Each independent variable is discussed using the l iterature to suggest possible relations to access. Chapter Three details the study: f i r s t l y , the hypotheses derived from Chapter Two are expl ic i t l y stated, then the data collection methods are discussed. These include the develop-ment of a questionnaire to measure access and the independent variables, and a description of the interview methodology. Included in this chapter is a description of the sample selection procedures. The findings are presented in Chapter Four. Chapter Five concludes the study by providing a normative view of access policies and an outline of future research possibi1it ies. - 6 -CHAPTER ONE A FRAMEWORK FOR STUDYING ACCESS TO INFORMATION Intentionally or otherwise, organizations regulate individual access to information by two mechanisms. The f i r s t is the organization's policy of access to information: this determines which participants are author-ized to have access to what types of information. This is an organiza-t ion's direct regulation of access to information. The second, indirect form of regulation, is through mechanisms or f a c i l i t i e s the organization employs to enhance or hinder the retrieval and use of a particular type of information. Employees can be authorized to have access to a certain type of information but i t may be impractical to retrieve and use i t . For example, they may be allowed to see their personal evaluation reports by the organization but because they are kept in another factory they f ind i t impossible in practice to retrieve them. These two types of regu-la t ion, direct and indirect , form the framework with which i t is possible to examine the l i terature concerning access to information in organiza-tions. Some organizations wil l emphasize direct regulation while others wil l emphasize indirect regulation of access. The differences between organizations need not concern us at this point. Instead we wil l discuss access to information for a general organization type, returning later to examine the differences across organizations. - 7 -I. Direct Regulation of Access to Information The organization's policy concerning access to information may be formalized in a document, describing which people are authorized to have access to which type of information, (e.g. Cary, 1976) or i t may be trans-mitted informally through a social ization process or through the tac i t understanding that on joining the organization the employee generally accepts the authority structure in the organization. Buried in this struc-ture wil l be the policy concerning access: "In joining the organization he (the employee) accepts an authority re lat ion, i . e . , he agrees that within some limits (defined both expl ic i t l y and impl ic i t ly by the terms of the employment contract) he wil l accept as the premise of his behavior, orders and in -structions supplied to him by the organization. Associated with this acceptance are commonly under-stood procedures for "legitimating" communications and clothing them with authority for employees." (March and Simon, 1958:90) In the same way, the organization equips the participants with authority to access certain types of information whilst denying them access to other types of information. It is possible to depict an organization using an information pro-cessing model (Galbraith, 1973:8ff). Clearly, information and its process-ing are essential ingredients to the effectiveness of any organization, without which some organizational tasks would be impossible. "The absence of a suitable technique of communication would eliminate the possibi l i ty of adopting some pur-poses as a basis for an organization." (Barnard, 1938:90) One of the obvious areas of access to information that must be established by organizations is supplying employees with information (or the resources to get information) that is relevant to their job performance, although - 8 -even here some important restrict ions may apply: t "Most persons in most organizations feel they do not have access to al l the information they need. Some-times they lack the information speci f ica l ly needed to accomplish their duties." (Forrester, 1965) The degree to which organizations ensure that employees have access to job-related information is a key issue in what Thompson (1967) cal ls "technical rat ional i ty" . If a suff ic ient number of employees has authority of access to job-related information denied to them i t is unlikely that the organization can function effect ively. However, not a l l job-related data supplied to employees is of equal importance to the organization. A simple dichotomy is therefore proposed. Job-related data can be either confidential or non-confidential where confidential information is infor-mation which i f released to other organizations (or interested parties) would be (or is assumed to be) damaging to the performance of the company of or ig in. An obvious example is the leaking of customer l i s t s to a competitor resulting in a loss of business for the aggrieved organization. The reason for distinguishing between these two types of job-related information is that organizations are l ikely to adopt different policies in handling these types. A further categorization is made to distinguish between information (confidential or non-confidential) that is needed for the employee's job and that information which is not needed. Once again, i t is expected that some organizations' policies on access make clear distinctions along these lines while others do not. The remainder of the information held in an organization is further divided into personal and non-personal categories, the latter category being a catch-all for general information kept by the organization. - 9 -Personal information is subdivided according to whether employees can see their own personal f i l e s , see their subordinates' and see their colleagues' (peers). Once again access policy is expected to d i f fer according to the category. A summary of this detailed structure is given in Figure 1. All Organizational Information ordinates Figure 1 Information Categories - 10 -1.. Organizational-task information An organization typical ly has one or more principal tasks that i t performs. The data required for these tasks have been labelled organiza-tional-task information to distinguish i t from other, internally held information used for ancil lary purposes (e.g. personnel information needed for manpower functions). In a manufacturing organization, task informa-tion would typical ly consist of inventory reports, production scheduling reports, sales s ta t i s t i c s , and so forth. As a contrast, in an educational inst i tute, whose principal tasks may be identif ied as education and re-search, the organizational-task data would consist of such items as class schedules, curriculum notes, academic journals, books, etc. (i) Information needed for the individual 's job At the level of the individual the organization needs to furnish access to information related to the job. One of the functions of an execu-tive is to ensure effective communication, and hence access to information within organizations (Barnard, 1938:217). Nord makes a similar point when he refers to the tasks of managers and the flow of job-related in -formation. "One of the most important jobs of a manager is to aid in the establishment of communication networks that fac i l i ta te task performance by the channels of information exchange and expertise and transmitting both operational and technical knowledge." (Nord, 1972:368) The whole thrust of Galbraith's model is the acquisition and processing of task-related information in order to make the organization effective in the face of increasing uncertainty (Galbraith, 1973). - 11 -In order to enable the individual to perform effectively the organ-ization must recognize the person's need for formal authority to secure access to essential information (Barnard, 1938:175). This does not, how-ever, prevent other employees fa i l ing to acknowledge another's authority of access. A study of access to government information found that the perceived status of the inquirer was a strong determinant of the success of the inquiry•(Divorski et a l . , 1973). This and other barriers to access wil l be discussed in the next section, as the main purpose here is to examine an individual 's formal access authority. Additionally, the organ-ization may only authorize access to a particular form of the information. Much reporting in organizations involves the distribution of f i l t e r ed , interpreted information, or what Sorter cal ls "value" information to d is -tinguish i t from the original "events" information (Sorter, 1969). One department may supply another with reports summarizing i ts act iv i t ies while the raw data are not communicated by the department that "owns" i t . Again, the dysfunctions associated with interpreted information are covered in the next section, the point to be made here is that for the same individual, access may be authorized to one form of information (e.g. interpreted information) but not authorized for another form of the same information (e.g. the original information). Access to interpreted rather than the original information is an example of the situation that exists between managers at the same level in different departments. In general, i t is possible to identify three different levels for which formal access to information may be granted by the organization: at the same leve l , a lower leve l , and a higher leve l , a l l with respect to a particular individual (Bacharach and Aiken, 1977). - 12 -For the majority of the organizations where task specialization and hier-archical organization are normal, i t would be expected that the policy on access to job-related information would vary somewhat according to these three directions. At levels beneath an individual one would expect that access to job-related information would be potentially unlimited, subject to the information f a l l i n g within the individual's sphere of authority. Thus i t i s expected that departmental managers would be authorized to have access to any job-related information within t h e i r departments i f only because "the departmental organization defines reasonably well the groups within which sharing of information i s needed" (Cyert and March, 1963:109). At the same l e v e l , however, many managers would not have access to raw data held by other departments. For example, a production manager would expect to see summary sales forecasts but the manager would not normally be authorized to have access to the original figures used to construct the forecasts. Indeed, such r e s t r i c t i o n s on access may be beneficial to the organization. Ackoff (.1967). and others have commented on the dysfunctions possible t f unlimited interdepartmental access i s allowed. Formal access to data held at a higher level i s largely a function of the executives who control t t . An important task of an executive is to establish and maintain effective access authority to job-related informa-tion for others in the organization (Barnard, 1938). The access to information at higher levels i s enhanced by the increased attention that such information seems to be given (Sussman, 1974). When an individual accepts a position in an organization, that person is responsible for understanding the rules and regulations that are relevant to the position and those of the employees under the individual - 13 -and, as already noted, this includes the company policy with regard to access to information (.see for example: Mechanic, 1962). At the very minimum, the person should ensure that information is made available for the essential functions performed by the department. Additionally, the individual is responsible for the rights of access that subordinates may have to their personnel records. Typical ly, new managers face a period of "indoctrination" by which they learn such procedures through reading manuals or by on-the-job observations. Because every new position has some start-up "costs" of learning these new procedures (Arrow, 1974) this suggests that a relationship should exist between the perceived abi l i ty of access to job-related information and the length of time a per-son has spent in that position (Mechanic, 1962). ( i i ) Information not needed for individual 's job There would be very l i t t l e dispute among organizations about the need for authorizing access to information (confidential or non-confiden-t ia l ) that individuals need for their jobs. However, where the informa-tion is not needed for the individuals' jobs one would expect some variety in policy. The choice is normally between regulating access by direct means or by indirect means. In the case of confidential information i t is expected that, because of the general need of the organization to protect such information from fa l l ing into the hands of competitors or other interested parties, organ-izations wil l normally confine access to only those who need i t for their job (Cyert and March, 1963:109). This would be accomplished in some organ-izations by rules and regulations (direct means) while in others, pro-hibit ion would be achieved indirectly by making i t too "costly" for indi-- 14 -viduals who do not need the information to retrieve and use the data. In the accounting department we would expect to find only certain members with access to detailed information for cost accounting purposes. Pricing -formulae are typical ly confined to a few members of an organization. Similarly, information on important industrial processes and formulae wi l l sometimes be known only by one or two members of the organization with al l other members excluded from access. In competitive industries, such secrecy is v ital to the performance of the firm. If important infor-mation is leaked the consequences could be disastrous f inancia l ly , and hence one would expect the firm to use whatever resources necessary to prevent unauthorized access and disclosure. Thus, i t is not surprising to find companies pursuing ex-employees who attempt to exploit for per-sonal gain the data they were once privy to (Aiken, 1974:40). In the case of non-confidential information, the regulation can also be by direct means (rules prohibiting or allowing access) or by i n -direct means ("costs" making i t d i f f i cu l t or easy to retrieve and use the data) or by both means. Any prohibition of access of employees to non-confidential information is not to protect valuable information from outside leakage, as with confidential information; instead i t might be used to ref lect the atmosphere of regulation in an organization where everything is regulated. Hence, i f you do not need information for your job then in these organizations there is a rule preventing your access to that information. If the information is held by other departments then the structure of the departments normally prohibits access without the necessity for formal rules. One of the problems of large scale management information systems (MIS) is precisely that, unless they are carefully designed, they - 15 -do not recognize departmental boundaries and a f f i l i a t ions with respect to information. This point is clearly la id out in the following extract: "When organizational units have inappropriate measures of performance which puts them in conf l ict with each other, as is often the case, communication between them may hurt organizational performance, not help i t . Organizational structure and performance measures must be taken into account before opening the flood-gates and permitting the free flow of information between parts of the organization." (Ackoff, 1967) Both Argyris (1971) and Bariff and Galbraith (1978) discuss similar effects of MISs upon organizations. Within departments there are two potential advantages to not pro-hibit ing access to information even when i t is not needed for an individual 's job. F i r s t l y , the ava i labi l i ty of this information allows individuals some freedom to determine their own patterns of access. Downs, comment-ing on the extent of this behaviour, said, "The vast majority of a l l communication in large organizations [is informal]" (Downs, 1967:269). Barnard (1938:224) also recognized the importance of this type of access although, contrary to his be l ie f , i t may not be under the control of the top executives (Rogers and Agarwala-Rogers, 1976:101; Mechanic, 1962). The second reason for the importance of such information is that certain jobs require individuals to acquire and use information beyond that which can be specified formally for their jobs. This requires access to information for what Mintzberg (1973) cal ls the detection of opportunities. It is assumed that individuals have authority of access to information for their regular jobs but their performance can be improved i f they can find useful problems to solve or make their current problem-solving more comprehensive (Pounds, 1969). For example, a portfol io manager may gain - 16 -considerable advantage over peers by gaining access to the right kinds of financial journals even i f they are not considered s t r i c t l y necessary for a manager's job. Arrow i l lustrates the same point: "General news sources about business conditions may be read simply because of their intr ins ic interest and hence at v i r tual ly no cost; but these may constitute a certain amount of monitoring. F inal ly , simply social associa-tions with business connections may constitute a source of information, the stronger because much evidence shows that personal influences are regarded as more re l iab le , which means that they convey more information, subjectively measured, at a given cost." (Arrow, 1974:51) Without some incentive from the organization to make access to such infor-mation less costly, the manager's choice of access is l ikely to depend heavily on personal attributes. "He may find i t cheaper to open certain information channels in ways connected with [his] ab i l i t i es and knowledge . . . It is cheaper to proceed to the chemical analysis of compounds similar to those already studied" (Arrow, 1974:41). The discussion of the nature of the relationship between organiza-tions and the types of mechanisms they choose to regulate access wil l be postponed until the total structure of the access model has been presented. 2. Other information Apart from internal organizational-task information the organization maintains and provides varying degrees of formal access to personal and non-personal information. (i) Personal information In addition to information that individuals need to perform their jobs, there will be personal information that individuals may or may not have access to. F i r s t l y , there wil l be information about themselves that - 17 -the individuals may have some rights of access granted to them by the organization, the state, or by contractual agreement through the union. Secondly, there will be personal information about their subordinates to which they will have some degree of formal access, normally in order to carry out the personnel functions of managers (evaluations, writing letters of recommendations, e tc . ) , but not necessarily confined to these functions. F inal ly , individuals may have some access rights to personal information concerning other individuals in the organization who are not their subordinates, for example, other managers (peers). As al l of these areas touch on the rights of access to personal information the sub-ject wil l be discussed in general before drawing conclusions about the policy of organizations in the three areas mentioned above. There are two types of personal information that the individual wi l l be concerned with. The f i r s t type is the information collected and main-tained by organizations concerning the individual and as we shall see from the l i terature, the amount of such information gathering varies widely between organizations and between countries. The second is information relating to on-the-job safety or health hazards. For the f i r s t type of information the issues of privacy and confidential ity take precedence. The employees have an interest in the organization restr ict ing others' access to their personal information. In the second area, the employees are concerned to see unencumbered access to such information relating to hazards which personally affect them. There would of course be further information which would interest employees in general. For example, unions negotiating contracts for their members would value better access to detailed management accounting - 18 -information in order to improve their bargaining posit ion. As i t is they are normally restricted to summarized financial information such as annual reports or quarterly statements of earnings. Although the union could claim that i t is entit led to have access to such information, i t normally fa l l s in the category "access is prohibited". An exception to this policy is when organizations wish to demonstrate that they are bargain-ing in good faith by opening the detailed records to union scrutiny. When employees enter an organization or request government ser-vices they admit the organization's right to col lect and retain personal information about them. This admission is part of their decision to par t i -cipate (March and Simon, 1958). The extent of their right to determine for themselves when, how and to what extent this information is to be given to others is a popular definit ion of the privacy of information (Martin, 1976:271). One of the prime ways to recognize a person's privacy right is to restr ic t the volume and type of information collected by the employer to "...what is necessary to enable the employer to assess-his su i tabi l i ty for work on which he might be or is employed" (Great Br i ta in, Parliament, 1972:95). The extent of information gathering is l ikely to vary for cultural reasons as well as with different tasks and organizations, but most people would agree that some information gathering is necessary: "These included questions about height and weight at ages 18, 20, 25, 30, 35, 40 and 45; about reading matter; and about family background and marital history." (Great Br i ta in, Parliament, 1972:94) In the U.K., to overcome capricious information gathering, the main trades union body has called for a standardization of questions, "which could be recommended as appropriate for use in normal circumstances" - 19 -(Great Br i ta in, Parliament, 1972:94). Other questions such as surrep-t i t ious gathering of personal information and the use of questionable techniques (e.g. polygraphs) are also relevant here. Employees are not only interested in controll ing what personal information gets collected by the organization; they would also like to know that the information is being used as or iginal ly intended. This is the issue of confidential ity of information (Auerbach Publishers, 1976:5). Confidentiality is broken when unauthorized persons gain access to infor -mation in order to carry out functions not or iginal ly sanctioned by the individuals or the organization. If the information collected is being used by the organization for legitimate purposes, then the individuals wil l want to know that the infor -mation that others are using to make decisions concerning them is a true representation of their current situation. They may, of course, prefer the information to be biased in their favour but as that is an unlikely situation to be maintained al l the time they would l ike their personal information to be as complete, accurate, relevant, and timely as is reasonably possible (Auerbach Publishers, 1976). This is clear from the draft code of the UK Code of Industrial Relations Practice: "[Planning of manpower] . . . needs to be . . . based on adequate and up-to-date per T sonnel records" (Great Br i ta in, Parliament, 1972:97). A mechanism to ensure that information does satisfy these requirements is to allow ind i -viduals access to their personal records, with the right to challenge their content. One of the main means of providing privacy and confidential ity of information is through security bui l t into the organization's systems of - 20 -information handling. These include the control of physical access to information, technical safeguards such as passwords and encryption of data, and administrative safeguards, the latter being exemplified by the follow-ing suggestion: " . . . the principle of "separation of duties" may be applied: that i s , access to some data may be allowed only with the knowledge and consent of someone in a different part of the organization." (Great Br i ta in, Parliament, 1975:7) On the one hand, the employee would l ike to restr ict the organiza-t ion's information collection act iv i t ies with respect to personal informa-tion and to prevent disclosure (access) to unauthorized personnel. On the other hand, the employee would l ike the right of access to informa-tion which affects the employee personally, and the examples of health and safety related information have already been mentioned. This second area of access is what one author cal ls the person's right to know, and freedom of information rights in this area would go a long way to remove isecrecy and create a climate of trust (Riley, 1977:22). In Canada not only are there no laws requiring organizations to provide access to hazard information for their employees but, in addition, many of the government reports of studies into industrial hazards are not publicly available. Hence, the government may have evidence of the harmful effects of toxic material but i t is not obliged to provide access to that information for the employees, union, or organizations (Riley, 1977:22). Clearly, even i f the organizations themselves possessed such information there may be strong incentives against allowing individual or union access to take place (Aykroyd, 1980b:46). Similarly, the individual 's right to access his own record and be assured of reasonable privacy and confidential ity - 21 -al l involve a substantial cost to the organization (Goldstein, 1975). Until now the issues of privacy, confidential ity and freedom of information for an employee have been discussed as " ideals" . Frequently, what occurs in the organization is largely at the discretion of those who control access to the information. Although there have been some moves to ensure public access to government records the same has not occurred in the private area. In the U.S., the f i r s t major piece of legislat ion was introduced into this area in 1967 when the freedom of information act (FOIA) was implemented to allow the public to obtain information from federal agencies describing what information the agency is permitted to collect and maintain. Unfortunately the act was widely misinterpreted by agencies and Relyea (1977:327-8) reports that excessive fees were some-times charged, requests were delayed and refusals seemed arbitrary. These are al l substantial barriers to access. Such problems caused the act to be widely amended in 1974. In the same year, the Privacy Act was signed which subsequently allowed individuals access to their own records held by federal agencies, and this has been followed by some states enacting similar legis lat ion. Also, 1974 saw the introduction of the Buckley Amendment allowing individuals to access their higher educa-tion records. Some believe that because of this amendment the quality of letters of recommendation has changed to a "more noncommital format" (Auerbach Publishers, 1976:3). To overcome th is , i t is possible for applicants to sign away their right to access letters of recommendation and allow the schools to adopt more candid information gathering (Shaffer et a l . , 1976). Several European countries have adopted legisla-- 22 -tion governing government information, although the scope and enforcement mechanisms vary greatly (see Sieghart, 1976:195ff, for a summary of recent legis lat ion) . To add to the open government laws already in place, the U.S. in 1977 introduced the "Sunshine Act" requiring most federal agencies to open their meetings to the public. In addition, 39 states have enacted similar laws (Hirschhorn, 1977). Although there have been some efforts in Europe to introduce similar rules of access into industry (e.g. the industrial democracy movement), attempts to open up executive decision making to employees have not met with any success this side of the Atlant ic. In the private sector only limited provision has been made for employees to gain access to personal information. The Swedish Data Act of 1973 covers both public and private sectors, but does not apply to manually stored records. In the U.S. the proposed HR 1984 was a b i l l that would extend privacy laws to cover privately held data but this has not yet become law. Westin (1979) gives a summary of the current situation con-cerning employee rights of access. For example, the state of Michigan has had, since 1978, an "employee right to know" law. Generally, however, the way to deal with employee rights of access to information has been lef t largely to the discretion of the private industry in the U.S. (Benson, 1978). One organization that has taken a leading position on employee privacy is I.B.M. (U.S.). They have laid out stringent procedures to be adopted by their staff governing: what l ine managers can and cannot see; outside requests for information; information required by law; and the rights of the individual to access personal information (Cary, 1976). For instance, l ine managers have access to the following information on any of their subordinates; job-related information, performance appraisals, - 23 -performance plans, letters of recommendations, record of awards, sales records, production assignments, etc. The same managers have no access to their subordinates' medical history, personal finances, payroll deduc-tions and 1 ife insurance. All employees are authorized to access anything in their own records but are prohibited access to another's record unless the latter is being considered for a position in their department. There is some indication that U.S. industry might be wi l l ing to adopt similar procedures. A recent survey by Harvard Business Review showed that of those business subscribers who responded, 87% were in favour of employees being able to access their personal records, i f some sensitive informa-tion were excluded (Ewing, 1977). In Canada there is evidence that although the employee does not have legal rights to access personal information, most can gain some access. The Canadian Task Force on Privacy and Computers (1972) found: "Most, though not a l l employees have the right to see and rebut their records, although many employers made i t clear that they would want to know why an employee wanted to examine his f i l e before lett ing him do so. In the case of unionized employees, the terms of access and the nature of the information in the f i l es are often governed by clauses in the collective agreement." (1972:55) The right of access is granted by the organization to the employee either by tradition or by contractual arrangements. Clearly, to make an employee just i fy his reasons for wanting access to his own record could act as a strong barrier to access. In order to implement rights of access for the employee i t is pos-sible to identify three methods of regulation: the self-help solution, the Ombudsman solution, and the licensing solution (Sieghart, 1976:123). The three methods probably have differential impact upon the ease of - 24 -access. The self-help solution is the one that North Americans are more familiar with. It is the regulation adopted by the 1974 Privacy Act and implemented within I.B.M. (U.S.). It provides the access mechanism but i t rel ies on the employees' in i t ia t ive to ensure that their personal records are complete and accurate while giving them an opportunity to challenge the content. Sieghart associates this solution with the American tradition (1976:124). The Ombudsman solution is a European idea (also imported into Canada in some provinces for other purposes) and has found advocates in two of the German states (Hessen and Rhineland-Palatinate), although the experience there has not been an overwhelming success. The ombudsman or commissioner has had d i f f i cu l t y in securing information on what f i l es are maintained by the government. Finally the Swedish Data Act (1973) uses licensing as a mechanism for regulating access.. Al l p r i -vate operators of automated personal. f i les must register their f i les with the Data Inspection Board, which is given wide-ranging sanctions to ensure compliance. The results of the licensing method seem particularly en-couraging. " . . . i t has every appearance of being cheap, effective and not l ikely to obstruct the normal needs of data pro-cessing within an advanced industrial society, while s t i l l providing substantial protection for the privacy of personal information." (Sieghart, 1976:128) It is clear from the evidence that the ideals of employee rights of access vary widely from country to country and in places, l ike Canada, where there are no legal rights to know, access is l ike ly to vary at the discretion of the organization, subject to other factors such as union contracts or industry norms. Without legal sanctions the organizations wil l tend to provide access according to economic considerations. As - 25 -their benefits are largely intangible and d i f f i cu l t to measure, employee rights of access will tend to be low on the priority l i s t for organiza-tional resources. Returning br ief ly to the three categories of access for personal information i t is now possible to suggest some general rules for organiza-tions in Canada. F i r s t l y , one would expect managers to have high authority of access to information about subordinates for most types of personal information. It is expected that in most commercial organizations the need to get the job done overrides any movement to restr ict access to personal data. Secondly, because the right of access of individuals to their own records is not formalized in Canada, i t is largely determined by industry norms or union agreements. Hence such authority of access is expected to vary across industry. In the f inal category of access to other, non-subordinate records, i t is expected that authority to access is very low. The organization has nothing to gain from providing access to this type of information and some organizations clearly guard the i n -formation that concerns status (salary, grade, e tc . ) , often promoting a "climate" of ignorance in these areas particularly for middle management (Forrester, 1965). An exception to this is found in public bodies such as universities where both grade and salary of the members of faculty are theoretically public information (Forrester, 1965). In the particular case of the University of Brit ish Columbia, a report including salary details is published annually by order of a statute but the salary figures are presented in ways that often conceal the true facts (Financial State-ments, University of Brit ish Columbia, 1979). This is an example where members of the university and the public have-;"the authority to access this information but because of the way the information" is published the use-- 26 -fulness of access is severely limited in practice ( i .e . regulation of access is by indirect means). ( i i ) Non-personal information Non-personal information is a general category for information collected and used by the organization that does not f i t into any of the previous c lass i f icat ions. It includes the sort of information that the organization or union may use to promote access in general, such as news-letters and bul let ins. Because this wil l be covered by questions con-cerning f a c i l i t i e s to promote access, i t wil l not be considered further in this study. Our emphasis will be on the information concerning the main tasks of the organization and the personal information collected and maintained by the company. - 27 -11. Indirect Regulation of Access The last section discussed the direct regulation of access, the organization's formal policy on access. However, achieving authority of access to information is not the only barrier to information that ind i -viduals face in an organization. They must be capable of both retrieving the information and using i t (Dhalakia and Sternthal, 1977). The f a c i l i -t ies to enhance or hinder retrieval and use of information are the means by which the organization regulates access indirect ly . Individuals may have authority to access an information source but they may be unable to get the information because of technological barriers (perhaps they do not know how to use the retrieval system). Additionally, even i f they could overcome the technical barr iers, the individuals may not be able to comprehend the information they receive. For example, they may be able to obtain a three hundred page report on inventory levels, but because of its volume they find themselves unable to effectively process i t ; they may have access to information from foreign branch of f ices , but they cannot read the language. From an individual 's viewpoint there wil l be barriers associated with each of the components of the feas ib i l i t y of access, retrieval and use, and each of these barriers wil l have a subjectively estimated cost associated with i t (Arrow, 1974:51). What follows is a fu l l e r discussion of these barriers as they apply to any type of information used by the organization. 1. Barriers to Retrieval of Information There are several factors to retrieval which mitigate against the individual getting the information he is seeking. Some of these barriers wil l be more important for particular individuals and organizations but - 28 -the extent to which the individuals perceive these barriers to be "costly" in retrieving the information compared with the "value" of the informa-tion to them, determines the feas ib i l i t y of retr ieval . For some job-related needs, the importance of getting the information wil l be so great that they wil l bear almost any cost in overcoming the barriers to retr ieval . In other non job-related areas, the barriers may effect ively seal off access to a particular type of information. (i) Ignorance of information An obvious barrier to retrieval is the individuals' ignorance of a type of information that they actually have authority to access. They may need i t for the jobs they are performing, or they may be legally entit led to access i t , but because of their lack of knowledge of its existence they cannot retrieve i t . In the public sector, many of the statements of principles governing personal record-keeping argue against the secrecy of the existence of f i l e s . For example, the H.E.W. report in the U.S. states: "There must be no personal data record-keeping systems whose very existence is secret" (Great Br i ta in , Parliament, 1975:47). In Sweden, where the Data Act covers both public and private sectors, "a personal register may not be started or kept without permission of the Data Inspection Board" (Sieghart, 1976:165). In their U.S. study, Westin and Baker found that there were few cases where institutions believed that members of the public were kept in total ignorance of their records. A more major problem was the prevalent lack of access authority the public has to a l l or some of their records (Westin and Baker, 1972:431-2). In Canada there is no law of disclosure in either the public or the private areas (Riley, 1977). Where regulations do exist , the - 29 -rules cover personal record-keeping. No organization is obliged to d is -close the existence of non-personal information to employees. As previously mentioned, the organization seeks to remove the barrier to ignorance for new employees as they undertake a period of t ra in -ing or indoctrination into company pol ic ies . This effort wi l l normally apply to those areas that are valuable to the company, i .e. job-related areas. However, i f the firm is concerned to maintain a climate of trust and openness (Zand, 1972) then one indication of this is i ts efforts to remove the barrier of ignorance concerning information in general. It is possible, however, that many firms use ignorance as a means to promote secrecy and to discourage the informal sharing of data. If an organiza-tion wishes to protect information from access, maintaining ignorance of its existence wil l normally be a much less costly mechanism than security. ( i i ) Retrieval procedure Ignorance is obviously a total barrier to access. However, i f we assume that the individual does know about the existence of the informa-tion and is authorized to have access there remain further barriers to retrieving the data. One of these is the procedure that the individual must employ to retrieve data. This is what Arrow cal ls part of the "capital costs" of data (Arrow, 1974:39). In order to retrieve data one must f i r s t expend time and effort in learning the retrieval procedures. However, in a study of engineers i t was found that while ease of use and channel accessibi l i ty (used as a largely physical concept in their study) were both signif icant factors in choice among sources of data, accessi-b i l i t y dominated ease of use in the measure of cost to the engineers - 30 -(Gerstberger and Al len, 1968). In the retrieval of computerized data there have for several years been advocates for both the special procedural languages of retrieval (Zloof and de Jong, 1977) and the use of "natural" language ( i .e . English) to retrieve data (Codd, 1974). The special languages involve a large in i t i a l investment in learning but allow high rapidity in the on-going formulation of retrieval requests. For natural language retrieval the trade-off is reversed; almost no effort in learning is paid for by a lengthy dialogue with the machine in order to agree on the request. However, there is some evidence to show that managers do not perceive themselves to be computer terminal operators, preferring to use others where such retrieval is necessary (Keen and Scott Morton, 1978:152). If this is so, the argument for and against different procedural languages may be mis-placed. ( i i i ) Geographic barriers In the Gerstberger and Allen study (1968), accessibi l i ty and f re -quency of use were strongly related. It is not clear from the work exactly how accessibi l i ty was defined but i t is clear that the geography of access was strongly implied in their use of the term. The lower the perceived geographic barrier to access, the more frequent use of data was found. Geography is one of the factors in the difference in costs for different directions (Arrow, 1974:41). The same point is vividly brought out in the following extract: "So long as data could be obtained in the same factory, there was l i t t l e or no problem of access; so long as the data were located in the same city the problems were s l ight ; but where data were located in a different city from the person who needed access to them, the problem was frequently regarded as serious." (Simon et a l . , 1954:62) - 31 -One of the conclusions of the same study supported this : "The most important consequences of centralization or decentralization of the record function have to do with the accessibi l i ty of the documents . . . IThis] points in the direction of relat ively great geo-graphic decentralization." (1954:7) In a study of a l ibrary system at a major university, the distance of faculty off ices from the l ibrary was an important variable in the l ibrary use (Greene, 1973). The weight of these studies support the claim that the perceived geographic barrier is an important "cost" in determining the feas ib i l i t y of retr ieval . (iv) Timing With respect to the retrieval of information there are two major components of timing. The f i r s t is timeliness, which is measured as the time between the events occurring and the data being accessible (Burch and Strater, 1974:34). This is what Gregory and Van Horn cal l "reporting delay" (Gregory and Van Horn, 1960:352). Clearly, i f a manager beiieves a report is going to arrive after i t is needed for a decision i t is of only general value to him and this wil l act as a large barrier to retr ieva l . On the other hand, timely access to important data sources could be highly valuable to certain decisions (for example, the release of expected or current prof i t figures during wage negotiations). Mintzberg (1975) notes that much formal information in organizations arrives too late. The second component of timing is the elapsed time required to re-trieve data once i t is accessible. If individuals believe the retrieval time is too great they may not attempt to get the data. Experiments with persons using computer terminals indicate that is is not just the delay - 32 -that is important; the var iabi l i ty or uncertainty in time to retrieve can also lead to frustration (Shneiderman, 1978:429). Clearly, elapsed time to retrieve data should be strongly related to the d i f f i cu l t y of the retrieval procedure. (v) Confidence in source It is expected that lower confidence in a type of information results in a greater barrier to retrieval and use of that information. If the value of such information is perceived as being suff ic ient ly low, the individual may discount the information by either not retrieving i t or i f i t has been retrieved, by not using i t . In some circumstances, employees may have alternate sources which can be used to verify one another and hence reduce their dependence on one source (Thompson, 1967:32). In other cases where there are no alternatives or, for various reasons (e.g. economic), the employees choose not to use them, they then have the problems of bias, distort ion, and omission to deal with. This wil l be discussed in the section "Barriers to Use of Information", although, of course, these problems are closely tied to confidence in the source. An individual may have confidence in either an impersonal source of information or in other persons as a source and for each situation the degree of confidence will vary with the individual 's previous experiences. Koester and Luthans (1979) and Luthans and Koester (1976) found that managers lacking computer experience place s ignif icant ly higher confidence in a computerized report than computer-experienced managers. In the Pearl Habour attack i t was the people in the local intell igence unit that lacked c red ib i l i t y : "Members of mil itary intelligence fai led to respond to a communication of the impending attack sent by lower ranking - 33 -officers in a unit which had low repute with the Intelligence Off ice" (Nord, 1972:367). Wilensky (1967:48,59) ascribes this kind of problem to the structure of the organization. He also gives several examples of similar "pathologies". Managers w i l l , in general, evaluate messages from their superiors as being more valuable than those from their subor-dinates. In organizational hierarchies, there should be l i t t l e i f any reason for conscious distortion of downward communication (Sussman, 1974; Lawler, Porter and Tennenbaum, 1967). Even a source which is perceived as low in cred ib i l i ty may be valu-able to an individual as a verif icat ion mechanism: "For executives removed from actual operations, the greatest significance of attention-directing account-ing data l ies in the information they transmit inde-pendently of operating supervisors." (Simon et a l . , 1954:27) In the same study this also was found to be true of factory foremen except that their important sources of information were the face-to-face contacts they had access to on the factory f loor . "He regards accounting information as only a supplement . . . to the other sources" (Simon et a l . , 1954:23). The executive use of staff in conducting technical analyses is another source of ver i f icat ion. Where the use of alternatives is too costly to employees, they may select one person they respect as an expert in one or more f ie lds and use this person as a source of information. This expert is sometimes called the opinion-leader who f i l t e r s information through a "two-step" process (Lazarsfeld et a l . , 1968:151; Ladendorf, 1970; Porter, 1974). Apart from any considerations of low data ver i f i cat ion, the use of opinion-leaders may entail a high "cost" of exposure. The individual is - 34 -openly dependent on another's opinion (Dewhirst, 1971). (vi) Other barriers to retrieval There will normally be a monetary cost associated with the access of formally reported information. Although in some firms the cost to an individual may be zero, in many the cost of formal reporting is a budgeted item, and, depending on how the organization's budgetary control system is set up, this may be charged in real dollars or internal money (Turney, 1977). As the charging of these items is intended to motivate individuals and make them cost conscious, one would expect perceived cost to be a potential barrier to retr ieval . Other barriers to retrieval are those related to the individual. Some individuals may perceive that the retrieval of a certain type of information makes them fee ! awkward or embarrassed which acts as a deterrent or barrier (Canadian Task Force, 1972:55). A particular example has already been given where employees have to just i fy their requests for access to their superiors. It is l ikely that the importance of this ex-posure as a barrier is greater where the climate is non-trusting and where the information is held by other people (Zand, 1972). In such cases, the use of computerized retrieval of data as an alternative to personal contacts may appear more attractive because of the neutrality of such technology. The person is not so exposed to others. However, the manager's fear or reluctance to use such technology may outweigh this advantage (Carter, 1976). This wil l be discussed further under the section "Technology of Access". - 35 -2. Barriers to use of information Individual access to information, as has been shown, is dependent upon the rules of access of the organization, the barriers to retrieving the information, and f ina l l y , how readily the information can be used. Clearly, the decision to retrieve the information wil l depend not only upon the barriers to ret r ieva l , but also the ease of using the information. The barriers to use are discussed below. (i) Mode and format Once the information has been obtained the f i r s t thing of concern is the "mode of presentation" and this is a potential barrier to use of information. Mason and Mitroff (1973) claim that although computer usage has become an "art ic le of fa i th" for most MIS work, "Stories, drama, role playsi art , graphics, one-to-one contact and group discussions may be more effective in some information contexts" (Mason and Mitroff , 1973: 484). They then discuss why this is important: different psychological types wil l want different types of presentation in order to be individually effect ive, " . . . Feeling and Intuition types may react extremely negatively to the idea of computer generated information, especially when i t is numerical rather than verbal" (Mason and Mitroff , 1973:484; also cf. Keen and Scott Morton, 1978:152). Not only is the mode important, so also is the format (Dhalakia and Sternthal, 1977). The format is the way the information is arranged within a particular mode. For example, numerical information may be pre-sented in graphical or tabular formats, with or. without explanation text. Zmud (1978) found a performance improvement through the use of graphical techniques for presenting numerical information and this supports the - 36 -work of Simon and Newell (1971) concerning human limitations on the pro-cessing of information. Churchman and Shainblatt (1965) found that rather than reduce processing, individuals will actually ignore information that is presented to them in a format incompatible with what they expect. If computer displays are taken as one example, the range of potential factors affecting performance of processing is staggering: "The size of the display screen, brightness of the dis -play, glare, f l i cker , contrast, typefont s ize, type-font design, graphics or color capabil i ty, and physical placement may a l l affect users." (Shneiderman, 1978:429) (i i) Comprehension Once the barriers of format or mode of presentation have been over-come, the individual is faced with the task of turning the data into information - the problem of comprehending the data - and there are barriers associated with this (Arrow, 1974:40). First of a l l , the barrier of language must be faced; i f employees do not understand the coding mechanism they cannot comprehend the information or, worse s t i l l , they may misinterpret the information. Secondly, employees must be able to cope with the volume of information presented before comprehension can take place. F inal ly , the content of the message must be assessed for bias, inaccuracies, and omissions during the interpretation of the infor -mation. (i i i) Language Before comprehension can proceed, the individual must understand the language of the message and this involves an in i t i a l cost to the person: - 37 -"Learning a foreign language is an obvious example . . . the subsequent ab i l i ty to receive signals in French requires this in i t i a l investment. . . . the technical vocabulary of any science is [another] case in point. The issue here is that others have found i t economical to use one of a large number of possible coding methods, and for any individual i t is necessary to make an in i t i a l investment to acquire i t . " (Arrow, 1974:39-40). Attempts have been made to measure the general level of education required to read messages and this has found some application in the accounting l iterature (Smith and Smith, 1971). (iv) Volume of information Another barrier to the processing of information is its volume. This will be more apparent for those managers working under pressure of time. One of Ackoff's complaints is not that MIS provides too l i t t l e relevant data but that managers are given an "over abundance of i r r e l e -vant information . . . I have seen a daily stock status report that con-sists of approximately six hundred pages of computer print-out. This re-port is circulated daily across managers' desks" (Ackoff, 1967). Managers under time constraint cannot process such large volumes of information (cf. Simon and Newell, 1971). One way to reduce the volume of data is to take the raw data ("events") and summarize or condense them ("value" data) (Sorter, 1969). There is some evidence that the use of summary information is better for technical decision making under relative certainty, but that performance has to be traded-off against decision time and decision confidence (Chervany and Dickson, 1974). Mitzberg claims that one of the problems with formal information systems is that they tend to aggregate information too much. This makes reports too general to be of use to the manager - 38 -(Mintzberg, 1975). Others have tr ied to explore the individual differences between people as measured by their cognitive "style" and hence match people with volume of output (Benbasat and Dexter, 1979; Benbasat and Taylor, 1978). Very l i t t l e experimental work, though, has been done on the effect of time pressure on the choice of information volume and per-formance (Wright, 1974). However, for many organizations the manager may have l i t t l e choice of mode, format, and volume of output for formal report-ing, especially i f others are receiving the same or similar reports. (v) Interpretation The information may be perceived as biased, inaccurate, or incomplete. These three barriers to interpreting the information are considered to-gether. The question of incompatibility of information sources will be addressed later. When a manager in an organization considers a source of information, consciously or otherwise, i t wil l be assessed for biases, inaccuracies, and omissions. The manager may rely heavily on one person as a source of information needed for task performance. This saves the managers time and effort in gathering and processing information from a variety of sources. The other person acts as an interpreter for the manager, f i l t e r -ing out "unwanted" information and condensing and processing i t to more manageable proportions. An example of this would be a manager's request to the accounting department for special reports or opinions. As with the retrieval of information, the manager has a degree of confidence in the source which is expressed in the bel ief that bias, inaccuracy, and relevant omissions are low. Of course, the benefits of using other people as interpreters of data may have to be "paid" for by such items as low veri-- 39 -f icat ion, dependence on others, and exposure to others. In hierarchical-type organizations the distortion of upward communi-cations has long been recognized (Athanassiades, 1973; Downs, 1967:119). At each level inferences are drawn from the raw data and these are communi-cated, not the raw data. This is important to the manager because: "his interpretation must be based primarily on his confidence in the source and his knowledge of thebiasesto which the source is subject, rather than on a direct examination of the evidence" (March and Simon, 1958:165). This process the authors cal l the "absorption of uncertainty" and i t has dist inct implications for the effectiveness of the organization: "If i t is true that typical ly only a very small portion of the total available information is ever recorded by the organization, the processes by which the in i t i a l screening takes place has extraordinary importance in determin-ing the f inal decision" (Cyert and March, 1963:20). Again: " . . . infor-mation is a resource that symbolizes status, enhances authority and shapes careers. In reporting at every leve l , hierarchy is conducive to concealment and misrepresentation" (Wilensky, 1967:43). Others have noted the reluctance to communicate undesirable information or the "mum" effect (Rosen and Tesser, 1970; Berry and Otley, 1975:180). If the upward communication from subordinates is distorted, what can be done about i t? Cyert and March (1963:110) claim that counter-biasing occurs but this may only apply to tasks where the decisions are made fa i r l y frequently (Pettigrew, 1972:202). In fact , the greater the uncertainty surrounding the task, the wider the range of values a variable may assume and the wider the range of latitude o f f i c i a l s have in empha-sizing one part of i t without being proved wrong (Downs, 1967). For more - 40 -routine tasks, constant bias can be allowed for without even removing i t (Feltham, 1972:123). Inaccuracies and omissions require other remedies. The primary way of reducing these is to use alternative sources for the same information (Berry and Otley, 1975:180; Simon et a l . , 1954:23,27; Downs, 1967:119). (vi) Incompatibility of information F inal ly , the individual , faced with integrating information from several sources, has a potential barrier to interpretation. For example, one report may be summarized by month, another by product, and there may be no opportunity for disaggregation for comparison purposes. This d i f f i -culty is in addition to the problem of geographic dispersion of informa-tion types (Simon et a l . , 1954:79). Summary This chapter has laid out the structure by which we can examine internal access to information. Organizations regulate access to infor-mation by direct and indirect means. The direct means by which access is regulated is the organization's policy on access, the rules that specify, for individuals, what types of information they are authorized to access and what types they are prohibited from. These rules may be formalized in a document or, more l ikely be informally established by an internal social ization process that occurs after an individual joins an organization. Indirectly, the organization regulates access.to information by the way i t promotes or hinders the retrieval and use of information. The extent of this regulation varies across both individuals and types of information. Thus, for a particular employee, access to one type of infor-- 41 -mation may be made highly convenient, while another type of information may have s u f f i c i e n t barriers associated with i t to e f f e c t i v e l y prohibit access. It has been demonstrated that access is an important area for study. In the next chapter we w i l l attempt to explain some of .the variations in direct and ind i r e c t regulation of access to information in organizations. - 42 -CHAPTER TWO DETERMINANTS OF ACCESS In the last chapter, access was treated largely as an isolated sub-ject with only cursory reference to the context within which the direct and indirect regulation of access operates. The purpose of the current chapter is to identify some of the more important features of the organ-izat ion's context which directly affect access to information or which interact to make an impact upon access to information. This wil l lead in the next chapter to a series of hypotheses linking access to its deter-minants and these hypotheses wil l form the basis for the subsequent empirical work. The f i r s t factors considered are those external variables that can have an impact upon access to information. These are shown to act largely as constraints upon thedirect regulation of access. Secondly, the chapter examines some of the important structural variables as they have an impact upon access. The variables considered are s ize , decentralization of authority, level of authority, shape of the organization, and routineness of technology. The third factor is the subjects' perceptions of the general attitude towards data sharing. Fourthly, technology of access is examined for i ts effects upon access. Lastly, two variables (depart-mental a f f i l i a t ion and the subjects' experience of their positions) are discussed in relationship to their influence on access to information. - 43 -I. External Regulation of Information The organization's enterprise governs the market within which i t must operate. Although the organization may seek to control or influence i t , the market is assumed to be a fixed part of the organization's environ-ment, especially in the short term (Pfeffer, 1978). In an informational sense the organization's survival is dependent upon securing access to and processing of information suff ic ient to match the organization's task(s) (Galbraith, 1973). This in turn governs the organization's policy with regard to individual access to information for job-related purposes and generally organizations wil l want to ensure access to infor -mation in these cases. But another condition of survival , the degree of which wil l vary from industry to industry, is the need to protect certain types of information from unauthorized access. For a manufacturer, these information types might include research and development reports, pro-duction figures, customer l i s t s and so on, a l l of which may have a sub-stantial value to competitors. Thus direct and indirect regulation of access to work-related data depends somewhat on the extent and value of confidential information held by the organization. This is at least partly determined by the competitive nature of the organization's environ-ment. For example, in the competitive chemical manufacturing industry, where the survival of a company is dependent on its ab i l i t y to protect proprietary formulae, i t is expected that this confidential information would be s t r i c t l y confined to those who need i t for their jobs. Addi-t ional ly , an emphasis is l ikely to be placed on security mechanisms to prevent unauthorized access, internally or externally. In research laboratories employees are typical ly required to sign agreements pro-- 44 -hi bi t i ng the release of technical information to those outside the organ-izat ion. In many cases, sanctions are used to ensure that individuals comply (Aiken, 1974). In non-competitive organizations or regulated mon-opolies, such as u t i l i t i e s , where the extent of confidential information is much lower or non-existent, the leakage of certain information types may no longer affect the commercial advantage of the organization. In-stead, leakage of such information needs to be prevented because of the embarrassment i t could cause the organization which may in turn lead to further external interference in the running of the organization. For example, a non-profit educational institute might want to protect the details of the wages of i ts c ler ical s taf f , not because of the commercial effect of leaking this information, but because of the embarrassment caused to the institute by wages that are generally lower compared to those available outside. The second item in the informational environment that could have an impact upon access to information is the general and specif ic regula-tion of organizational information by governments. These requirements may be in direct conf l ict with the organization's aim of protecting key i n -formation sources. For instance, the government may require a manufacturer to disclose data concerning toxic substances to an appropriate agency. The information may have enormous commercial value to the firm and i t wil l therefore want to ensure that the agency applies stringent security to such information before disclosure takes place (Dueltgen, 1979). The law may also require that information concerning the health and safety of employees is made readily available. In Canada, only Newfoundland has legislat ion requiring that employers disclose health- and safety-related - 45 -data to employees although Quebec is proposing similar regulation (Aykroyd, 1980a:40). In Cal ifornia the National Labor Relations Board ruled in one case that unions did have the right to access information, "on raw mat-er ia ls and chemicals stored, handled and processed in the company's plant . . . this was rather a hollow success, however, as unions succeeded in obtaining only the trade, and not the generic names, of the chemicals i n -volved" (Aykroyd, 1980b:46). Clearly, one way for the organization to pro-tect sensitive information is to control the form of the information re-leased. In the Californian example, the organization was able to bypass the intention of the ruling by releasing non-sensitive information instead of the original data. The law may also require that personal information is regulated. Whereas the confidential data for an organization wil l be suitably pro-tected from unauthorized access, without legislat ion there is l i t t l e i n -centive to protect personal information from such access, especially as the implementation of security wil l normally be costly (Goldstein, 1975). It is expected, therefore, that in such cases as in Canada where there is no regulation, the organization will apply l i t t l e or no resources in this area, the exceptions being the cases where a particular industry sets a standard concerning personal data (Cary, 1976) or where unions have negotiated an agreement that includes rights of access (Canadian Task Force, 1972). The informational environment is highly dependent upon the country in which the organization operates. In the area of personal information, this can be easily demonstrated by comparing legislat ion across several countries (Sieghart, 1976). In the regulation of toxic substances there are already differences between Canadian and American practices (Aykroyd, - 46 -1980a, 1980b). It is clear that many less developed countries have l i t t l e or no regulation of information. Consequently some industries wil l find strong incentives to establish subsidiaries or research laboratories in these countries, using such countries as data havens rather l ike some companies seek out and use tax havens. As the external regulation of information acts largely as a con-straint upon access to information and because the study will be conducted in one country, this factor wil l be largely controlled. Any industry-specif ic exceptions to this wil l be commented upon during the presentation of the results. II. Structure The manner in which relations are patterned and differentiated within organizations is called structure (Thompson, 1967:51). The organ-ization needs to balance the needs of differentiating between various functions while providing suff ic ient coordination of those functions in order to achieve overall effectiveness (Lawrence and Lorsch, 1967). The typical hierarchical structure adopted by most organizations is very effec-tive in restr ict ing access to information and therefore provides some ordering to the patterns of access. If everybody had access to everybody, the organization would collapse under an overload of information (Galbraith, 1973). However, this; restr ict ion also brings with i t certain costs to the organization, costs in the sense of d istort ion, biasing, and omission of data - information pathologies (Wilensky, 1967; Rogers and Agarwala-Rogers, 1976:106). Although there are many ways of dimensioning struc-- 47 -tures (Steers, 1977:57-69) the following variables have been selected because of their acceptance in the l iterature and their relevance to the access question. (i) Size The organization's size can be thought of as a contextual variable (Pugh et a l . , 1969) or as a structural factor. From the point of view of this study, the interest l ies in the effect the size of the organization and the size of the work unit have upon access to information. Bacharach and Aiken (1977) correlated organizational size with communication behaviour at two levels of authority. Using a logarithmic measure of the number of employees, they found that size and communication act iv i ty were positively related, as hypothesized, for supervisors but not for department heads. Size was also strongly correlated to their measures of shape (width and height). This confirms the work of Pugh et a l . (1969:93) who found size to be highly correlated with the structuring of ac t i v i t ies . Steers (1977:67) reports that size appears to be positively related to increased efficiency while being negatively related to employee attach-ment to an organization. The larger the organization, the lower the identif icat ion an employee feels towards a firm. Also, the work unit size has been found to be positively associated with job dissatisfactions. In the study of access, the size of the organization and work-unit may be important in two dimensions. F i r s t l y , some of the physical barriers wil l appear to be greater as size increases. The individual is l ike ly to have further to travel in order to retrieve the information needed. This will be heightened by the increased need to use sources of information (Bacharach and Aiken, 1977). Secondly, as well as increasing physical barr iers, the l iterature suggests that in larger organizations the at t i-- 48 -tudinal barriers wi l l , also appear larger. As size increases, the ind i -vidual generally feels less attachment to the firm or work group. It is expected that both organizational size and work-unit size wil l be related to access to data (Comstock and Scott, 1977:179). However, because of the "washout" effect of treating a collection of subsystems (e.g. departments) as a united system, i t is expected that the size of the work-unit wil l be of greater significance than the size of the whole organ-ization. A similar argument on the treatment of technology at an organ-izational and a work-unit level is presented in Steers (1977:78). (i i) Decentralization The extent to which lower participants in a hierarchy have authority to make decisions concerning their tasks is called decentralization of authority, or just decentralization (Steers, 1977:60). Tushman (1979) used the ratio between vertical and horizontal communication behaviour as a measure of central izat ion; the higher the ratio the more centralized decision making would be. Hage et a l . (1971) measured decentralization by asking individuals questions about their authority to make decisions in several areas. They found a strong positive relationship between decen-tra l izat ion and communication behaviour. As decentralization is related to participative decision-making i t was argued that more decentralization led to more participation and hence greater communication (Steers, 1977:60). Bacharach and Aiken (1977), using the Hage et a l . (1971) instrument, found consistently strong relations between communication behaviour and decentralization for both levels studied in the organizations. Read (1962) found that "open communications" were strongly related to decentra-l i za t ion . This wil l be captured by measuring the attitude of individuals to data-sharing. - 49 -A seemingly confl ict ing notion has been raised by Galbraith's information processing model (Galbraith, 1973) where decentralization is proposed as one method of reducing communication. The conf l i c t , however, is vacuous. The communication studies al l refer to act iv i ty within the work-unit, not between the work unit and others in the hierarchy. This distinction is useful nonetheless when we look at the effects of decentra-l izat ion upon access to information. Namely, i t is expected that the demand for access would be greater for information held within the decentralized work-unit while less demand should be found to information stored outside the unit (Cyert and March, 1963:109). Additionally, because the decentra-l ized unit should contain a l l of the information required to perform the tasks, the physical barriers to access will appear to be correspondingly lower. This is contingent upon the recordkeeping function also being decentralized and therefore under the authority of the unit: "The most important consequences of centralization or decentralization of the records functions have to do with the accessibi l i ty of documents and the re l i ab i l i t y of the source records. Both of these c r i te r ia point in the direction of relat ively great geographic decen-t ra l iza t ion . " (Simon et a l . , 1954:7) ( i i i ) Organizational level Bacharach and Aiken (1977) studied the communication behaviour of supervisors and department heads within bureaucracies. One of their most s ignif icant findings is the differences between the two levels: "Regarding department heads, the major finding is the lack of effect of these structural dimensions upon the frequency of department head communication." (Bacharach and Aiken, 1977:373) With regard to access, i t is expected that individuals at differ ing levels - 50 -will perceive differences in barriers to access of information. For example, managers have more resources available to access information. As well as being able to afford retrieval equipment and on-going costs, they also have available manpower to retrieve and process information on their be-half. Although they may find access personally d i f f i cu l t because of the complexity of technology, for instance, they may have several alternative paths available to them (Keen and Scott Morton, 1978:152). At the highest levels, executives are l ike ly to believe they can get any information. Clearly, this is not true of the workers on the shop floor who, i f they need to access an information source, have no alternative but to get i t themselves unless they can find alternative sources of information (e.g. such as the "opinion leader", Rogers and Agarwala-Rogers (1976)). It is expected that the higher the individual is in the hierarchy, the less barriers wil l be perceived to the access of information. A potential confounding effect is the expected positive relationship between the level of the individual in the hierarchy and the informational con-tent associated with the person's job. Some higher level jobs are almost total ly data-oriented while others lower down in the hierarchy are re la -t ively data-poor (Poole, 1978). Hence, the workers on the shop floor may find large barriers to accessing information but not find this a great problem because they rarely need to retrieve information to perform their jobs. On the other hand, the manager in a "data-rich" position will develop a greater expertise in handling documents than the shop-floor worker (Weber, 1947; Mechanic, 1962). Put another way, access to informa-tion is part of a manager's normal work. - 51 -(iv) Shape of organization For this variable two dimensions are ident i f ied: horizontal and vertical di f ferent iat ion. Horizontal differentiation refers to the degree to which the organization organizes i t se l f into departments, each special -izing in a particular function (e.g. marketing, production, etc . ) . Vertical differentiation is a measure of the number of dist inct authority levels that exist from the employees to the executives. Both dimensions are believed to have an effect upon access to information. Horizontal differentiation may be thought of as an attitudinal variable, a structural variable, or both (Lawrence and Lorsch, 1967). Horizontal differentiation or departmentalization is a way of clustering together act iv i t ies which can then be performed relatively independently of each other. This has a clear effect upon the needs for information access: "The departmental organization defines reasonably well the groups within which sharing of information is needed" (Cyert and March, 1963:109). March and Simon (1958:169) refer to this as loosely-coupled programs of action. However, there wil l be important occasions when interdepartmental access of data is necessary. In those cases the individual will normally find access to another department's detailed data to be d i f f i cu l t (Ackoff, 1967). The tendency wil l be for individuals to l imit access of informa-tion to that which can be easily obtained within the department: "Generally, persons tended to communicate with others who were within easy reach, and also with others who were closely related [structurally] in the organization" (Aguilar, 1967:112). Thus, although departmental differentiat ion may be important to the organization's effectiveness, i t also increases the barriers to access between departments. The provision - 52 -of extensive data processing technology to force data sharing between departments may only serve to make the situation worse (Argyris, 1971; Bariff and Galbraith, 1978). Bacharach and Aiken (1977) defined a department as any unit with at least two persons and two levels. The head of the department also had to report to the chief executive, in their case the mayors or other local government o f f i c i a l s . The number of such departments was the way they measured the width of the organization. They found a weak inverse re la -tionship between the number of departments and communication behaviour, but only for subordinates; no relationship could be found for department heads. Lawrence and Lorsch used a composite measure which used the d i f -ferences in orientation as well as differences in the formality of the structure (Lawrence and Lorsch, 1967:10). Bacharach and Aiken used a count of the number of levels of authority as their measure of vertical d i f ferent iat ion. They found a strong posi-tive relationship to communication behaviour but, once again, only for sub-ordinates. In relation to access to information, the number of levels of. authority is l ikely to provide a measure of the potential distortions and omissions, that vert ica l ly transmitted data can suffer. The more people that screen the information on its way through the hierarchy, the more "concealment and misrepresentation" is l ikely to take place (Wilensky, 1967) and the more uncertainty gets absorbed at each level (March and Simon, 1958:165). Others have noted the tendency of individuals to trans-mit only favourable information to their supervisors (Rosen and Tesser, 1970; Berry and Otley, 1975:180). It is to be expected, therefore, that a l l else being equal, the number of levels in the organization will determine the inaccuracy, bias, and incompleteness of data generated in-- 53 -ternally. In organizations with large vertical d i f ferent iat ion, i t is expected that the individual wil l face larger barriers to access of infor -mation from sources down the hierarchy and wi l l therefore seek alternative access paths or sources when counterbiasing is not feasible (Berry and Otley, 1975; Pettigrew, 1972; Cyert and March, 1963). As was mentioned, the higher the level of the individual in the organization, the more re-sources the person can muster to find these alternative access paths or sources, somewhat mitigating the effect. (v) Routineness of technology The technology of an organization can be thought of as processes (Woodward, 1965), as relationships between units or individuals (Thompson, 1967:15-18), or as existing at the level of individual tasks (Hage and Aiken, 1969; Comstock and Scott, 1977:181). Because the focus in this study is on the routineness of technology, i t is this latter measure of technology that will be used here (Bacharach and Aiken, 1977:370-1 ). This level of focus seems to be the one best suited to the empirical work in which the subjects are individual managers. One dimension of tasks performed in organizations is their degree of routineness or predictabi l i ty . Thompson (1967:134) suggested a four-way c lass i f icat ion of tasks according to the degree of uncertainty both in the goals of the task and in the methods used to achieve those goals. A task.is predictable or routine to the extent that i ts goals are well defined and its means/ends relationships can be specif ied. Simon (1960) cal ls this kind of task "programmed" where a procedure can be constructed and applied every time. This has also been called a "structured" task (Mason and Mitroff , 1973). Simon (1960) contrasts a programmed task with - 54 -a nonprogrammed or unstructured task where the problem faced is unique and wil l not y ie ld to an established procedure. Although he dichotomizes tasks in this extreme way, Simon does acknowledge a continuum between these two extremes. Keen and Scott Morton (1978) introduce a th i rd , intermediate category of semi structured tasks, where the problem is only partly pro-grammable. Several writers have suggested that communication behaviour will be affected by the degree of uncertainty within the task. Predictable tasks, "will have low information-processing requirements which can be f u l f i l l e d by fixed programs, rules, and standard operating procedures. Furthermore, since relevant information is l ike ly to be higher in the hierarchy, routine tasks can be accomplished with more supervisory decision making and less extensive peer communication" (Tushman, 1979:84). Comstock and Scott (1977:177) make a similar point; routine or predictable tasks can be handled with more standardized procedures and less internal communication. The opposite is true of nonroutine tasks: "It is clear that when the purpose is not simple - that is when the requirements are complex and not obvious, or the conditions require precision of coordinated move-ments, or the nature of the individual action is d i f f i -cult to grasp by the actors . . . - much more communica-tion is necessary than under the contrary conditions." (Barnard, 1938:107) This is elaborated on by Tushman (1979): "If a subunit's task is nonroutine, i t must attend to substantial information processing requirements, since complex tasks require generating and evaluating a l ter -native approaches to solutions. Furthermore, the more complex the task, the less l ikely that the requisite task information wil l be available from any one ind i -vidual, even the supervisor. To deal with this com-plexity requires peer decision making and extensive peer contact." (Tushman, 1979:84) - 55 -Tushman used four three-point scales to identify the routineness of tech-nology. Because he was studying a research and development laboratory, his four scales were: Basic Research, Applied Research, Development, and Technical Services, representing the range from nonroutine tasks to highly routine ones. Tushman also used Pelz and Andrews (1966) technique to weight the scores according to the percentage of tasks in each category (Tushman, 1979:87). Bacharach and Aiken (1977:370) used a simple six question measurement to capture routineness of technology. Tushman found significance between task routineness and communication in the direction hypothesized. Bacharach and Aiken found signif icant findings for depart-mental heads but not for subordinates. What are the implications of routineness of technology for access to information? For routine, predictable tasks, the organization can specify in advance the types of information that individuals need to have access to in order to perform their jobs effect ively. That i s , the patterns of access can be predetermined by the organization. However, when the task is less routine, i t is not possible to total ly plan for the information requirements of the job and hence the organization can only partly specify the information individuals are required to have access to. The individual , in order to perform nonroutine tasks, requires more access to information than for predictable tasks. But the organization can only support the task by authorizing access to pre-determined information (Keen and Scott Morton, 1978:1). In order to perform their nonroutine tasks effect ively , therefore, individuals must exercise discretion and in i t ia t ive in gaining access to additional information. To be effect ive, they need more access to information that the organization might normally - 56 -classify as not needed for their jobs. Without support from the organiza-t ion, the individuals' choice of access will depend on the barriers they perceive to access. These types of nonroutine jobs are l ike ly to be assigned to workers "having suff ic ient general training and experience to be able to act effectively in uncertain situations" (Comstock and Scott, 1977:181). But the degree to which the organization perceives this type of job to be important to its effectiveness will determine the extent to which the barriers of access to non-task information are minimized for individuals performing this type of work. If, as.seems to be widely held, managerial tasks consist largely of unstructured or "wicked" problems (Mason and Mitroff, 1973), this issue wil l be pervasive in organizations. One clear method to support this type of task is for the organization to improve the attitude to data sharing, thereby reducing many of the barriers to access. III. Attitude to Data Sharing In the communication l i terature the individual 's attitude to data sharing has been consistently found to be strongly related to communication behaviour (Dewhirst, 1971; O'Reil ly and Roberts, 1976; Athanassiades, 1973; Zand, 1972). Dewhirst measured the influence of norms of data sharing upon infor -mation channel ut i l izat ion (Dewhirst, 1971). Using one perceptual question to measure the individual 's perception of sharing norms, he confirmed his hypothesis that the use of interpersonal channels is directly related to the strength of the perceived norms. O'Reil ly and Roberts (1976) found a - 57 -relationship between source credib i l i ty and a series of independent var i -ables, two of which were perceived information accuracy and perceived open-ness of communication. Athanassiades (1973) proposed that upward distortions in communica-tion behaviour would depend upon the organization's climate. In fact, he did not measure climate but he selected two normative extremes of organizations for his study. A police department was chosen as a "hetrono-mous" climate, typif ied by elaborate rules and regulations. A university was used to represent an "autonomous" climate where emphasis is placed on individual freedom. As hypothesized, the " f reer" , more open climate was associated with lower distortion of communication. Zand (1972), using a concept developed by Gibb (1964), claimed that an attitude to trust would influence information behaviour: "One who does not trust others wil l conceal or distort relevant information, and avoid stating or wil l d is -guise facts, ideas, conclusions, and feelings that he believes wil l increase his exposure to others, so that the information he provides will be low in accuracy, comprehensiveness, and timeliness; and wil l therefore have low congruence with rea l i ty . One who does not trust wil l try to minimize his dependence on others." (Zand, 1972:230) Using a controlled laboratory experiment, Zand found that trust was s ign i -f icantly related to a l l the variables he examined. Of particular impor-tance to access to information, trust was related to openness about f ee l -ings and search behaviour. The greater the perceived trust, the greater the perceptions of openness and search behaviour. Similarly, Roberts and O'Reil ly (1974) and Burke and Wilcox (1969) found that a subordinate's trust in a supervisor acted as a fac i l i t a tor of open communication ex-change. Also, i f subordinates perceive that their supervisor is supportive, - 58 -then s i g n i f i c a n t increases in t h e i r reliance on the supervisor as an i n f o r -mation source have been found (O'Reilly, 1977). Piecing a l l these elements together i t i s possible to see the following s i m p l i f i e d relationships with respect to access: Norms of Data Shari ng Trust and Openness Sharing of Data Source Credibi1 • i t y Percei ved barriers to access Sharing of information w i l l be pos i t i v e l y associated with organizations where the perceived norm is to share. However, sharing norms may not be related to increased source c r e d i b i l i t y . An individual may share informa-tion because everyone expects i t but i t i s not clear that this sharing w i l l a l t e r the information pathologies discussed e a r l i e r (e.g., Wilensky, 1967). On the other hand, trust (and openness, which i s strongly related to trust) w i l l influence both sharing behaviour and confidence in the source of information (source c r e d i b i l i t y ) . Trust w i l l also influence the norms of data sharing ( i . e . , data sharing i s a necessary but not s u f f i c i e n t condi-tion of trust and openness). Both sharing of information and increased source c r e d i b i l i t y w i l l reduce the percetved barriers to access of i n f o r -mation. As these attitudes may be loc a l i z e d in an organization, i t would be necessary to distinguish between these variations. One way to accomplish this i s to measure trust and "norms" of data sharing at the - 59 -work unit leve l , rather than for the whole organization. Attitudes to sharing data may be influenced by structural effects (Steers, 1977:106; Athanassiades, 1973). Both size of organizations and level in the hierarchy have been found to influence an organization's climate. It may also be that a similar relationship wil l be found for attitudes to data sharing which would indicate that data sharing is one possible dimension of an organization's climate (Muchinsky, 1977). This finding would also support Forrester's claim that, "Much of the character and atmosphere of an organization can be deduced from the way i t internally extends and withholds information" (Forrester, 1965). F inal ly , the attitudes found towards data sharing can also be a demonstration of the desire of individuals to create data monopolies and accrue power for themselves: "To possess information is to possess power. A mono-poly of information can give a form of security. There are, in al l organizations at a l l levels, a selective withholding and extending of information. Sole possession of information can make others dependent on oneself . . . Control of information channels can isolate certain per-sons from the remainder of the organization and keep them within one's own sphere of influence." (Forrester, 1965) In one recorded case, maintenance engineers were able to gain a power ad-vantage over the organization by retaining control over access to manuals required to service the machinery (Crozier, 1964). The expertise of the maintenance engineers is a good example of one of the factors that con-tributes to the power of lower participants in the organization (Mechanic, 1962). As the same author notes, "lower participants do not usually achieve control by using the role structure of the organization but rather by circumventing, sabotaging, and manipulating i t " . The engineers were - 60 -successful in maintaining their powerful position by manipulating the access to important information allowed to others. At the same time, organ-izations are continually seeking ways to neutralize their dependence on experts by making their tasks more routine by using, for example, standard operating procedures (March and Simon, 1963; Hickson et a l . , 1971; Crozier, 1964). The "gatekeeper" is another good example of an individual who effectively controls the access to information for others (Pettigrew, 1973; Rogers and Agarwala-Rogers, 1976). Power gained through an individual 's control of access to information is not treated expl ic i t l y in this study. The major thrust of the work is to treat access to information as the potential to retrieve and use infor-mation. The study does not deal with individual choice behaviour and ind i -vidual manipulation of access for others even though these are important subjects. The study is interested, however, in how the organization as a whole is seen by individuals to withhold or extend access to information. Something of the emphasis taken in this study is apparent in the following statement: "Just as an individual hoards information, so does the organization as a whole. Competitive position . is often believed to rest on secrecy to a far greater extent than is the fact. Information is withheld from individuals inside the organization on the ex-cuse that this keeps information from outsiders." (Forrester, 1965) IV. Technology of Access In this section, the technology used to access information is examined for i ts effects upon the direct and indirect regulation of access (Mason and Mitroff, 1973). Generally, i t is proposed that technology wil l in-- 61 -fluence access to information in two connected ways. F i r s t l y , technology has a direct impact upon an individual 's perception of access to informa-tion. Access to a particular information type may appear to be faster./ slower, easy/dif f icul t , etc. according to the type of technology employed Secondly, the technology can influence the relationships between informa-tion sources. Some technologies can provide communication links between sources and hence change the feas ib i l i t y of access over that available without such technology. The following is a simple representation of the relationships: I Individual access rmation ships be-C C I I auurces of information Because technology is important to both organizational-task information and other information in different ways, these will be considered separately after a discussion of the effect of technology upon the direct regulation of access. Effect of Technology upon Direct Regulation of Access The question being raised here i s : Can a company's policy on access to information be influenced by the type of access technology employed? The only indirect evidence we have for this in the l i terature is when non-computer technology is replaced by computer technology. As one writer has noted, there are certain bui l t-in ineff ic iencies in operating administrative, non-computerized retrieval systems (Canadian Task Force, 1972). Different - 62 -types of information in such systems tend to be stored in different places and often have different formats. Additionally, several people may be needed to obtain the information that is wanted (e.g. secretaries and clerks). The introduction of computer technology and especially telecommunication links brings with i t the potential for removing some or a l l of the protec-tive ineff ic iencies that were previously there. The new system also means that the access procedures must be formally stated i f only because the soft-ware demands such formalization. Systems designers refer to this process as "designing the environment" of the computer system in recognition of the importance of formalizing the procedures that the computer system wil l affect when i t is brought into operation. Where i t was previously possible for an organization to "muddle through" with an informal policy on access, the introduction of computer systems wil l often act as an organizational trigger to formalize the direct regulation of access. This process is l ike ly to apply for al l types of information that are affected by computer technology. Effect of Technology upon Indirect Regulation Indirect regulation refers to the barriers to retrieval and use of data, and the effect of technology wil l be treated in that order for both organizational-task information and other information. (i) Organizational-task Information The f i r s t requirement is to find a c lass i f icat ion of technology of access. Clearly, i t must be a c lass i f icat ion that individuals can readily assess as i t wil l be used in the data collection phase. Burch and Strater (1974:28) use a four-category c lass i f icat ion of data processing methods; - 63 -manual method, electromechanical method, punch card equipment method, and the electronic computer method. Although the various classes do include, or at least imply the access technology, the c lass i f icat ion scheme suffers from at least two problems. F i r s t l y , the manual method includes access to both documents and people. Several writers have indicated the importance of personal contact rather than impersonal (e.g. Arrow, 1974). Secondly, punch card equipment has largely been eliminated from organizations with the advent of small, cheap computers. Aguilar c lass i f ied management sources of information according to whether they were personal or impersonal: "Examples of personal sources include direct te le-phone conversations, le t ters , personal memoranda, and so forth. Examples of impersonal sources in -clude publications, conventions, and scheduled meetings." (Aguilar, 1967:65-66) If we take these classif icat ions i t is possible to exploit the advantages of both. The following is a proposed c lass i f icat ion for the technology required to access (retrieve and process) information: 1. No Technology. e.g. face to face personal contacts, meetings, use of others. 2. Manual Technology. e.g. Internal Mailing system - regular reports, memos, letters. Use of mechanical f i l i n g system - clerks, secretaries. Requesting special reports, use of reference l ibrary. 3. Electromechanical Technology, e.g. Telephones Conference ca l l s , computer conferencing Mi crofi1m/microfi che readers. 4. Electronic Computer Technology. e.g. Requesting regular or special printed reports -batch system Use of online computer terminals. - 64 -Using this c lass i f icat ion i t is now possible to examine the l ike ly effect of the different kinds of technology upon the indirect regulation of infor-mation. Retrieval The f i r s t barrier previously discussed was ignorance. It is reason-able to expect that a properly designed electronic system can maintain ignorance of a source of information more rel iably than a manual system using f a l l i b l e people. For example, the human "grapevine" can be very ef f ic ient in redistributing information around the organization (Rogers and Agarwala-Rogers, 1976:100-1 ), whereas computer-stored information can be s t r i c t l y controlled so that only those who need i t can gain access to i t . The retrieval procedure is clearly affected by the technology of access. If the access is to be face-to-face then there wil l be certain procedures to be adhered to before access can take place (etiquette) and the degree of cooperation wil l somewhat vary with the relationship (e.g. peer to peer, subordinate to supervisor). Building-up a network of personal relationships is what Arrow (.1974:39) would call part of the "capital costs" of acquiring information. With manual access to information, the procedure, once established, might be very "cheap" to use. For example, the receipt of regular reports, memos, and other notices involves the manager in no effort of re t r ieva l , whereas establishing those channels of information may have been very costly. An individual could also find i t costly to personally retrieve manual records. But this would be accomplished at no cost i f intermediaries are used to access the f i l es for the manager (e.g." clerks or secretaries). Generally, the higher the technology that is used to retrieve information, the greater the in i t i a l effort to learn the re-- 65 -trieval procedures. This can, however, be mitigated somewhat by the use of intermediaries. In computer retrieval of data the use of difference access languages involves a tradeoff between a large investment in learning the language and the on-going ease of use of those languages (Zloof and De Jong, 1977; Codd, 1974). For the issue of geographic barriers to access, there is only one clear effect on access. Using electronic or electromechanical technology i t is possible to "compress" the distances between people and between people and information. The individual need only be as far away as the telephone or computer terminal. This would appear to have a clear advan-tage over face-to-face access when the geographic barrier is great and at least one study has shown that this technology can increase interpersonal communication (Turoff and Starr, 1978). This effect was also apparent in Aguilar's study; "Generally persons communicated with others who were with-in easy reach, and also with others who were closely related in the organ-ization" (Aguilar, 1967:1121. On the other hand, just because the te le -phone brings people near does not mean that an individual prefers to use i t . Mintzberg (1973) found a clear managerial preference for face-to-face contact with others. Technology is l ikely to influence both aspects of timing: the report-ing delay, and the retrieval delay. Electronic computer technology has a clear speed advantage over manual processing of numerical information and hence the reporting delay, once a l l of the data transactions are available, should be drast ical ly reduced using computers. Computer reporting can be made available for retrieval more readily than manual reporting (Stewart, 1971:255). Similarly, on-line information retrieval is potentially much - 66 -faster than manual retrieval (Burch and Strater, 1974:31). If the computer f i l es are available, an on-line retrieval is clearly faster than waiting for a batched-processed report to be sent by the internal mail system. This advantage may have to be paid for , however, by frustration caused by the uncertainty in retrieval times exhibited by some computer systems (Shneiderman, 1978). The only results of studying the effect of computer technology upon the confidence in the source of information have been stated elsewhere and they show that non computer-experienced individuals place more confidence in computerized output than non-computerized output (Luthans and Koester, 1976; Koester and Luthans, 1979). Other studies have indicated the power-ful influence of personal information sources over impersonal (Arrow, 1974; Lazarsfeld et a l . , 1968). The direct cost of access to information may be important or not to the individual depending on how the organization makes individuals account-able for their access to information. If the employee is direct ly b i l led for a l l equipment needed (terminals, microfilm readers, etc.) and computer time used, this would be a clear barrier to using more technology. If not, then other considerations might govern the individual 's choice of technology, such as the d i f f i cu l t y of learning the retrieval procedure. Generally, however, the in i t i a l set-up costs for electronic computer systems are going to be high compared with, other technologies (Burch and Strater, 1974:31). The last barriers to retrieval to be considered are the ones asso-ciated with the individual 's attitude to access. There is sometimes a certain psychological "cost" or exposure associated with access of informa-- 67 -tion (Zand, 1972; Dewhirst, 1971). Dewhirst mentions this in connection with the use of informal, interpersonal channels. The use of others as sources of information can indicate your dependence upon them. In such cases i t may be thought that the use of a "neutral" technology such as computerized data retrieval may be preferred. There i s , however, evidence to indicate that managers do not see themselves as terminal operators (Keen and Scott Morton, 1978: 152). Use of Information Both the mode of output and the format of the output are potential barriers to processing the information (Mason and Mitroff , 1973). For managers, more technology may be resisted and a preference for more personal sources of information shown. If the technological mode of output can be chosen, the format can be designed to suit the individual. In practice i t may not be. Once individuals have obtained the information they must be able to comprehend i t and technology can have an impact on comprehension. The out-put wil l be presented in a certain language or code. It is not clear that the technology of access wil l d i f ferent ia l ly affect this barrier to access. A computer-generated report may be identical in language (coding) to a manually-generated one. In the area of volume of information, however, i t is possible to spot a difference in the technologies. In manual systems, the production of reports is often expensive as is the marginal cost of each extra requirement. For example, a secretary has to type each addi-tional page requested. In computer systems, the marginal cost of producing extra pages is much lower once the programs have been written and this can result in a superabundance of reported information, not a l l of i t relevant - 68 -(Ackoff, 1977). Because cost may not be a great factor in this kind of reporting, care must be taken to ensure that the report is designed so that i t is at least capable of being processed by the individual and does not swamp the person with an overload of information (Chervany and Dickson, 1974). Once the information is captured by a computer system i t can generally be transmitted, processed and outputted without any loss of accuracy (Burch and Strater, 1974:31). It may s t i l l suffer from the bias and incomplete-ness that existed when i t entered the system, but even here the computer systems can do l imited, automatic tests to check for these. Any other technology suffers from inaccuracies, biases, and omissions, particularly person-to-person contacts, although individuals develop methodologies to deal with these, such as counterbiasing and the use of alternative sources or paths of access (Cyert and March, 1963; Berry and Otley, 1975; Downs, 1967). The technology of access can also affect the relationships between sources of information. For example, the centralization of information in in one or more computer f i l es can permit access to diverse information sources where once i t was not feasible. The keyword retrieval systems pro-viding access to abstracts is a good example of this gathering process. An individual can use the computer's technology to search quickly through a l l the required indices. However, in organizations, the bringing together of several previously unlinked sources of data can be the cause of many dysfunc-tions (Bariff and Galbraith, 1978). Managers, who previously had control over access to their information, can find themselves forced to share i t with other departments (Argyris, 1971; Ackoff, 1967). Whereas the physical - 69 -barriers to access can be made much lower by the use of computer technology, the psychological affects on the manager may outweigh any other considera-tions (Argyris, 1971). In the section covering structure, the relationship between routine-ness of technology (task structure) and access was examined. Is there also a link between the manager's job and the technology of access? Simon (1960) used the terms programmable and non-programmable to indicate the extreme types of jobs possible in an organization. This indicates that some jobs can be so structured and defined that computer programs can be written to replace a manual series of instructions, whereas other jobs require human judgement or compromise and therefore cannot be programmed (Thompson, 1967). The influence of computer technology has largely been seen in organizations in routine, structured jobs (b i l l i ng , payrol l , etc.) whereas the more i11-structured tasks have scarcely been affected (Brady, 1967), and where the type of technology used to access information is non-computerized (see also Mason, 1969). Computer technology is being intro-duced into this less structured area of problem solving but as a support to the individual, not as a replacement (Keen and Scott Morton, 1978:1). ( i i ) Other information For other types of information the issues of access feas ib i l i t y are different from those of organizational-task information. For information that the individual needs to access (job-related information) the organiza-tion wil l try to ensure that access is feasible. For his own personal records, the individual , and, rf legislat ion is tn effect , the organization, is concerned with keeping these from unauthorized access while ensuring that health and safety information is easily available where this is impor-- 70 -tant. On the other hand, for information considered prohibited to an ind i -vidual, i t is the organization that wants to make sure i t is secure from internal and external access or leaks. F inal ly , non-personal information wil l normally have no requirements attached to i t . With this framework, i t is now possible to look at the impact of technology upon privacy, con-f ident ia l i t y , and security. To simplify the discussion, the f i r s t three categories wil l be considered together as "manual" technology, the other category being electronic computer technology. These ref lect the two major dimensions identif ied in the l iterature on privacy, confidential i ty, and security. One of the early concerns of writers on privacy was that computer technology would mean an inevitable increase in data col lect ion: " . . . once an organization purchases a giant computer i t inevitably begins to col lect more information about i ts employees, c l ients , members, taxpayers, or other persons in the interest of the organization." CWestin, 1967:161) The second concern was for the misuse of information once i t was col lected, that i s , the loss of confidential i ty: " . . . the three [fears] which seem to be uppermost in the public 's mind are [the computer's] f a c i l i t y to compile "personal p ro f i l es " , i ts capacity to corre-late information and its provision of new opportuni-ties for unauthorized access to personal information." (Great Br i ta in, Parliament, 1972:180) Before computers, personal f i l es were generally scattered throughout the organization. With the advent of fast telecommunication links and high-speed processors the linking of f i l es became feasible whereas because of poor access to information i t was previously infeasible: "Computers, as a consequence of their own eff ic iency, break down many of the protective - 71 -barriers of ineff ic iency, which in the past have helped to shelter privacy" (Canadian Task Force, 1972:111). This does not mean that data sharing did not previously occur: "the mails, teletype, telephone, phototransmission and radio sets were used for the data transmission, and many organizations moved extremely high volumes of information about people through these media", and Westin found evidence of manual building of individual "pro-f i l e s " (Westin and Baker, 1972:252). However, computer technology does have the potential for making these links much more easi ly. Once the links are established i t can handle almost any volume of information. Both of these concerns have not been confirmed in practice. Westin and Baker comment "the organizations that we vis i ted have not extended the scope of their information collection about individuals as a direct result of computerization" (Westin and Baker, 1972:249, their emphasis). Simi-l a r l y , they found that for data-sharing, " . . . nothing in computerization i t se l f has produced a sharing of identif ied Information to a broader class of users within multibureau organizations or among organizations before computers" (Westin and Baker, 1972:255). This was confirmed in the U.K. study (Great Br i ta in , Parliament, 1972:179). Nonetheless, the threat of computer technology to privacy and confidential ity remains a real one (.Great Br i ta in, Parliament, 1972:181 ). and the attempt to ban the use of single identifying numbers (e.g. S.I.N.) as a universal record ident i f ier is evidence that some individuals fear the implications of using such in -tegrating devices (Canadian Task Force, 1972:85-90). Individuals are not only concerned that personal information is kept from unauthorized access; they also need to know that the information being used is accurate: "inaccuracy in a personal record may result in - 72 -dangers to privacy; a conviction for dishonesty might, for instance, be attributed to the wrong John Smith; a taxpayer might be l isted as a de-faulter when his assessment was in fact under appeal" (Great Br i ta in, Parliament, 1975:5). Although accuracy is somewhat a function of computer hardware and software, the general consensus is that computer technology results in an increase in accuracy compared with manual system equivalents (Westin and Baker, 1972:298; Great Br i ta in, Parliament, 1975:5). Among the reasons for this are the errors that are found and removed in convert-ing manual f i l es to computerized ones and the (potential) use of software to check for accuracy, completeness and reasonableness (Westin and Baker, 1972:299-300). Security is the means to ensure the privacy and confidential ity of information (personal or otherwise). Again computer technology can be a means to breach the security systems in organizations. Using a computer terminal and having access to the organization's computer system, i t is possible to access information without even physically entering the organ-izat ion. However, the practice does not match the potential , leading Westin and Baker to comment: "We found no instances of complete-outsider intrusion, solely by technological means, into computer f i l es to obtain information content, . . . We found far more examples of information breaches from manual f i l e s , reflecting their presently greater value in confidential information." (Westin and Baker, 1972:314) The later UK study confirmed this finding (Great Br i ta in , Parliament, 1975:7). This leads to the conclusion that security is probably at least as good for computer f i l es as for their manual equivalents and there are good reasons for believing that security of computer-held information can - 73 -be far superior compared with manual systems. For example; terminal entry to computer systems can be restricted physically and by passwords, usage can be recorded, communication links can be made free of intrusion by electronic encryption of information, computer f i l es can be centralized, and they can be easily duplicated to reduce the chance of total loss (Sieghart, 1976:90-94; Martin, 1976; Canadian Task Force, 1972:101-110). But whether organizations make their computer systems secure is a function of outside pressure: "Although no system operating in the active world of government, commercial, and private l i f e can be made permanently and completely safe, there are available techniques for providing far more security for infor-mation in computerized f i l es than are presently being used . . . [This] wil l depend primarily on outside pres-sures, especially the attitudes of regulatory agencies and law makers on how important i t is to ensure the confidential ity of information in various sectors of record-keeping." (Westin and Baker, 1972:315) Maturity of the Information System Finally in this section on technology of access we wil l examine the effect of the maturity of the computer information system upon the ind i -vidual 's view of access to information. Gibson and Nolan (1974) were the f i r s t to propose that information systems are subject to the learning curve within organizations. This re-sulted in a model consisting of four stages in the growth patterns: i n i t i a t i on ; expansion; formalization; and maturity. Each stage is asso-ciated with different types of applications that are computerized, a growth in specialized personnel, and a need for different management-techniques. The model has been modified subsequently to include a f i f th stage (Nolan, 1975), and even further stages were later added with the - 74 -model adapted to distinguish between growth patterns in different functional units (Nolan, 1979). Empirically the stage hypothesis was- tested by Lucas and Sutton (1977). Instead of using the absolute budgeted amount spent in the D.P. department as or iginal ly proposed by Gibson and Nolan (1974), they normal-ized budgets in terms of the G.N.P. to account for inf la t ion. Using 12 years of data, the researchers tested three models of growth: l inear, exponential, and the learning curve. The best f i t was a linear relat ion-ship with the learning curve providing the worst explanatory power. They conclude that D.P. budgets do not provide a good predictor of the stage of maturity. As a different approach the number of years since major problems were involved wil l be used as a simple measure of maturity of the computer system. This, coupled with the measure of personal use of computer ret r ieva l , wil l provide a composite measure that wil l be tested against the findings on access. Where a computer system is mature (most of the problems having been solved) and the individual use is high then i t is suggested that the barriers to job-related information wil l be assessed as lower than when there are s t i l l major problems with the computer system but the usage is s t i l l high (Bjorn-Anderson and Hedberg, 1977). V. Other Determinants As well as the factors treated so far there are two further variables that suggest themselves as being associated with access to information. The f i r s t is the departmental a f f i l i a t ion of the individuals and the second is the number of years the individuals have been employed in their - 75 -current positions. (i) Departmental a f f i l i a t i on As the study wil l concern i t s e l f with subjects from various departments in various organizations, i t would be valuable to test i f access to infor-mation is dependent on the subjects' departmental a f f i l i a t i on . Are certain departments in a better position to gain access to a given type of informa-tion? As mentioned before, departments largely contain the information that they need to function within their boundaries (Cyert and March, 1963: 109). This means generally that the individuals in a department would normally have better access (direct and indirect), to information peculiar to their department. For example, the production department manager is l ikely to have good access to production information and the marketing manager would be privy to market detai ls . The personnel department should have good access to personal information but may have poor access to many other types of information. The one department that has a potential to handle a great deal of the information used by most of the other departments is data processing. This, of course, depends upon the pervasiveness of computing in the organ-ization and what stage of maturity of applications the system has reached (Gibson and Nolan, 1974). In some organizations computer applications may only have touched job-related information. In other companies i t may have affected both job-related information and personal information. Clearly, because of i ts expertise the data processing department wil l have access to the information from most applications that are computerized and i t is therefore expected that D.P. managers wil l believe they have - 76 -better access to most types of information than managers in comparable posi-tions in other departments. If this is so i t would somewhat confirm the work of Bari f f and Galbraith (1978) who suggest that power wil l accrue to the data processing department because they control several strategic con-tingencies (Hinnings et a l . , 1974). ( i i ) Experience of position In addition to their departmental a f f i l i a t i o n , the number of years individuals spend in their jobs is expected to improve their view of access to information. Arrow (1974) has commented"upon the in i t i a l investment in learning new procedures and gaining contacts before signals can be obtained and interpreted. Weber noted that the time spent by a bureaucrat in becom-ing familiar with an organization's rules and regulations etc. gave a con-siderable advantage to the bureaucrat over the new (pol i t ica l ) incumbent (Weber, 1952). Mechanic (1962). has developed the following formal hypo-thesis: "Other factors remaining constant, as a participant's length of time in an organization increases, he has increased access to persons, information, and instrumentalities". Although Mechanic deals primarily with lower participants in organizations, his hypothesis is clearly ex-tendable to other ranks of employees and can be (partial ly) tested in this study. Summary Chapter Two has identif ied several organizational and other variables which are l ike ly to influence the direct and indirect regulation of access to information. For each variable the l i terature was presented and dis -cussed and l ikely relationships to access were ident i f ied. - 77 -It was shown that the external regulation of access through mechanisms such as government legislat ion acts as a constraint upon the direct and indirect regulation of access to specif ic information types. By using a Canadian sample of middle managers these factors are largely controlled in this study. Next, a series of structural variables were considered. It was argued that increasing size is l ike ly to have a negative impact on both the convenience of access and on the manager's authority to access non job-related information. Decentralization of authority ts expected to be positively related both to a manager's authority to access information and to the manager's ab i l i t y to be able to retrieve and use information. The l iterature suggests that the level of the manager in the company's hier-archy wil l be positively related to the ab i l i t y to access information. The manager may require more information access than other employees lower in the hierarchy, but the organization can supply the manager with more resources to effect such access. Increases in the "height" or the "width" of an organization are l ikely to increase a manager's barriers to access. In the case of the "width" i t was- argued that this relationship should only hold for non-routine tasks. Higher routtneness of the manager's job is expected to reduce the barriers to information required for the manager's job. However, because these jobs are predictable in their in -formation requirements, an organization can determine more clearly what information the managers are not authorized to access. Both high "norms" of data sharing and a positive attitude of data sharing are expected to reduce many of the barriers to information access. It was also argued that the effect should be more pronounced for the latter variable. The technology used to access information is expected to - 78 -influence the direct and indirect regulation of access. For routine tasks, i t was argued that the high deployment of computer technology would probably increase the convenience of access to required information, while reducing the authority and convenience of access for information not re-quired for the job. A discussion of the l i terature on access to personal records using various forms of technology was also presented. For managers in companies with "immature" computerized information systems i t was suggested that they would have impaired access i f their work demanded the high use of computers. Some l iterature was presented showing that data processing managers are l ike ly to have better access over information than other managers. F inal ly , the manager's greater experience in a job was shown to improve the ab i l i ty to access information. The presentation of the relationships as formal hypotheses is given in Chapter Three. - 79 -CHAPTER THREE THE STUDY The objective of the empirical study was to test hypotheses linking access to the independent variables described in Chapter Two. This chapter begins with a formal statement of the hypotheses that were tested. The chapter also describes the development of the questionnaire, the interview technique employed, and the population sampled. Although the study is concerned with how organizations regulate access, the empirical work focussed largely on managers' perceptions of these regulations. In this study we assumed that the managers' perceptions of their organization's direct and indirect regulation of access to informa-tion formed the "enacted environment" to which they reacted (Weick, 1969). Of course, individual managers may not understand correctly their duties or their rights of access. However, i t is assumed that their perceptions of the regulation of access formed part of the basis for the subsequent actions they took concerning the retrieval and use of information. - 80 -I. The Hypotheses For each independent variable one or more formal hypotheses are stated.''' After each hypothesis a summary of the rationale is presented. 1. Access and Size The "costs" or barriers to access of information in larger companies/departments are higher than those experienced in smaller companies/departments. Rationale: As the size of the company/department increases i t is expected that attitudinal barriers to access wil l also increase as a result of employees experiencing greater alienation from their company (Steers, 1977). Additionally, the size of the unit affects the geographic dispersion of information and the need for greater communication (Bacharach and Aiken, 1977). In each of these cases the direction of the relat ion-ship suggests that managers should experience greater barriers to access in larger companies/departments. Because of the "washout" effect of treating a collection of different de-partments as a whole company, the relationship between the barriers to access and size should be more pronounced for departments than for companies as a whole. W^'- A manager's authority to access non job-related information is lower in larger companies/departments than in smaller units. Rationale: For larger companies/departments i t is expected that more formal control of access has been established. This results in greater restrictions being placed on informa-The hypotheses are al l written as the alternatives to corresponding null hypotheses. - 81 -t ion, particularly, that which the manager does not require to perform his job. 2. Access and Decentralization of Authority H :^ A manager's authority to access information is higher in those companies with greater decentralization of authority. Rationale: A manager able to exercise greater authority over decision-making is also expected to be able to exercise greater discretion over work-related and personal informa-t ion, i .e . , a manager's local authority includes authority over information. H :^ A manager's "costs" or barriers to access of information are lower where authority is decentralized compared with those cases where authority is not decentralized. Rationale: Part of the reason for decentralizing decision-making is to reduce the need for inter-department communication (Galbraith, 1973). The department should contain most of the information a manager needs to perform his job. Hence, access to information within the department is mainly re-quired. Consequently, the barriers to information are expected to be lower. 3. Access and Organizational Level In this study, organizational level was largely controlled by choosing a population of middle managers. For this reason the effect of the subjects' level upon access to information was not tested. 4. Access and the Shape of the Organization The shape of a company was measured by the number of levels of authority in a manager's department (an indication of the hierarchical depth of the company), and the number of dist inct departments (the width of the company) - 82 -(Bacharach and Aiken, 1977). Hg: A manager's "costs" (barriers) to using information are greater in companies with greater vertical differentiat ion (levels of authority). Rationale: The number of levels of authority governs the number of different people that process the information before the manager receives i t . The more intermediate processing that takes place the greater the likelihood for distortion and omissions to occur (Wilensky, 1967; March and Simon, 1958). Hg: For managers performing non-routine jobs in companies with high horizontal differentiat ion (more departments) the "costs of access to job-related information are higher. Rationale: For routine, predictable jobs the proliferation of departments may improve the managers' access to information by bringing closer together those whose jobs are similar (Cyert and March, 1963; Aguilar, 1967). For non-routine jobs, however, a greater number of departments is l ikely to result in greater "costs" of access to job-related informa-tion because the local department wil l not contain al l of the information the managers need for their jobs. In these cases the managers wil l normally find access to information held in other departments to be more d i f f i cu l t and incon-venient (Ackoff, 1967). 5. Access and Routineness of Technology H ? : A manager who performs more routine tasks faces lower "costs" (barriers) to access of information required for such tasks and higher "costs" (barriers) to the access of information that is not required for the tasks. Rationale: For routine, predictable jobs the organization can spec in advance the types of information the manager needs and the - 83 -patterns of access necessary to retrieve the information. The same reasoning can be used to argue that the manager who performs routine jobs can be barred from the access of infor-mation not required for the job (Comstock and Scott, 1977). Access and Attitudes to Data Sharing Hg: Managers in departments where the norm of data-sharing is high face lower "costs" (barriers) to the retrieval of job-related information and have higher authority to access the same information. HQ: Managers in departments where the attitude to data sharing is characterized by trust and openness, face lower "costs" (barriers) to the retrieval and use of a l l types of infor-mation and have higher authority to access this information. Rationale: The norm of data sharing in a department results in a greater sharing of information required for the job. Trust and openness, however, result in both an increase in informa-tion sharing and increased cred ib i l i t y for the information source. This leads to a lowering of the "costs" of both retrieval and use of a l l types of information (Zand, 1972; Dewhirst, 1971; O'Reilly and Roberts, 1976). For each type of information for which the manager has increased access there will be a corresponding increase in authority. Access and the Technology of Access H l f ) Managers who perform routine jobs and who use computers frequently face lower "costs" (barriers) of access to infor-mation. Rationale: The use of computer technology can affect access d i rect ly , by increasing the speed of ret r ieva l , for example, or indirectly by altering the patterns of access to previously unconnected sources of information. It is to be expected, therefore, that for routine jobs where the requirements for information - 84 -are known and the patterns of access are relat ively stable, the retrieval of such information wil l appear to be more convenient to the manager. Additionally, there is some evidence to suggest that managers have increased confidence in using computer-generated information (Koester and Luthans, 1979; Luthans and Koester, 1976). H,,: Managers who use computers frequently for their jobs have lower authority to access information not required for their jobs and face higher "costs" (barriers) for the retrieval of such information. Rationale: Computer software can enforce more effectively the access authority structure than manual administrative systems. There-fore, managers who use computers frequently should have lower authority to access information not required for their jobs and face increased barriers to the retrieval of this informa-t ion. 8. Access and the Maturity of the Information System H-.2: Managers in companies where the computerized information system is immature but where their work demands frequent computer use face higher "costs" (barriers) of access to job-related information. Rationale: The managers' access to information is made harder when both the information system s t i l l has major problems and the managers' work requires its use. 9. Access and Departmental A f f i l i a t ion H,.,: Managers from data processing departments face lower "costs" (barriers) of retrieval to a l l types of information compared with managers from other departments. Rationale: It is expected that data processing managers wil l have more convenient access to most or a l l types of information - 85 -that are administered through their department. Although they may be capable of retrieving al l of the information stored on the computer f i l e s , i t is not expected that d.p. managers are capable of or interested in using a large pro-portion of the data that passes through their department. 10. Access and Experience General information processing sk i l l is assumed to be positively related to the manager's age. Specif ic , job-related information processing sk i l l is measured by the number of years the manager has held the current position in the company. H^ : Managers with greater experience (general and job-specific) face lower "costs" (barriers) of access to al l types of information. Rationale: All new jobs require substantial start-up "costs" before managers are familiar with the rules and practices of access. With experience (general and job-specif ic) , managers develop personal networks of contacts to enchance their access ab i l i t y (Arrow, 1974; Mechanic, 1962). 11. Relationships Between Authority, Retrieval, and Use In addition to testing the formal hypotheses presented above, the following relationships were considered to be of special interest and, therefore, they were the object of further analysis. One type of analysis was performed to demonstrate the relationship between the authority ratings for different types of information. It was hypothesized that access to confidential information would be less authorized than access to non-confi-dential information and that authority to access work-related information would be greater than authority to access non work-related information. It - 86 -was also postulated that authority to access subordinates' records would be higher than the authority to access either the subjects' own records or those of other managers. The analysis produced a complete ordering of these authority measures. A similar analysis was carried out for the measures of indirect regulation for each type of information. A parallel ordering to the authority profi les was envisaged for the general question on access ab i l i t y . A second type of analysis was performed to investigate the links between authority and access ab i l i t y . For this purpose, the component variables that a priori measure retrieval and use were aggregated to form two composite variables for retrieval and use for each type of information^ Each variable was tested for intertest r e l i ab i l i t y . Correlational analyses were then con-ducted between each rating of authority to access and the corresponding composite variables for retrieval and use, as well as between authority and each individual variable. In the final analysis, the relationship between general access ab i l i ty and the composite variables for retrieval and use were sought. The ques-tions addressed here were, what are the relative weights given to retrieval and use in determining access, and how do these weights change given the different information types? - 87 -II. Data Collection Before testing the hypotheses presented in the previous section a combination of questionnaires and interviews was employed to gather the data. They are both described below. 1. Questionnaire Development A copy of the questionnaire used in this study is presented as appendix I. The four sections in the questionnaire are as follows: (i) Rules of access to information ( i i ) The practice of access ( i i i ) General questions about access (iv) Questions about the company, the work, and the manager (i) Rules of access to information In this section, the objective is to measure the managers' perceptions of their authority to access different types of information. The types of information are those discussed in the f i r s t part of Chapter One. For job-related information the categories used are confidential/non-confidential, and needed for the job/not needed for the job. It was found that in the course of the pi lot study that i t was necessary to be more precise over some of these terms. For example, in order to help managers understand what was being referred to as confidential information, the term was de-fined as "information which i f released to other companies would prove harmful to the performance of your company". This was done to stress the value of this type of information i f leaked to competitors in the belief that companies would handle this information dif ferent ly . Subsequently, this was modified to cover those companies which do not have competitors, - 88 -such as regulated monopolies. A further suggestion made was incorporated into section I which enabled the managers to focus on specif ic examples of confidential and non-confi-dential information. A range of examples was developed and modified in the pi lot study to include those found frequently in commercial companies. An open-ended category was included for further examples peculiar to the individual subjects. Personal data is divided according to whether managers are authorized to see-their own f i l e s , their subordinates' f i l e s , and other managers' personal information. The personal items are subdivided into four broad categories, the most obvious one being job status (salary, grade, job history) as this has had special mention in the l iterature (Forrester, 1965). A seven-point Likert-like rating scale was used throughout this section. ( i i ) The practice of access In this section of the instrument the objective was to obtain the manager's assessment of various factors that may enhance or hinder access to data. Again, seven-point Likert-like rating scales were employed throughout this section. The factors were those developed in detail in the second part of Chapter One; each statement forms a summary of the discussion presented there. One additional statement was added after the pi lot study, "others fa i l to recognize the manager's legitimate authority to get this information", otherwise al l of the statements were derived from the l i terature. The statements were divided into two, approximately equal groups and one group was reversed. That i s , there are some statements that describe the company - 89 -as enhancing access to information while others refer to barriers to access faced by managers. The rationale for this was to prevent subjects provid-ing patterned responses to the statements. Several subjects confirmed in the interviews the effectiveness of this device in making them carefully consider their responses to the statements, while others commented on the comprehensiveness of the statements on access. F inal ly , the statements were randomized to prevent order effects, and a summary statement on re-trieval and use of information was appended. The order of the information types is the same as in section I. In cases 1-2 and 3-4, the subject is asked to bring forward the examples used in section I. This has had the disadvantage of making the subject turn back pages i f the examples previously used cannot be recalled. However, i t does ensure that the subject is focussing on the same information types used in the previous responses and i t is thought that this outweighs the inconvenience caused by the exercise. Where i t is appropriate, the manager is given- the option of leaving a particular column blank i f the data type cannot be retrieved. This is logical ly necessary because i f the data type cannot even be obtained, i t is pointless to ask the manager to respond to statements about i ts retrieval and use. ( i i i ) General questions about access These two questions were used to supplement the information gathered from the f i r s t two sections. The f i r s t question was designed to assess the manager's view of how much data the company c lass i f ies as confidential. The subject has already indicated in section I the items that are conf i -dential and the greater number of items ticked would be a rough indication of how pervasive confidential information is in the company. However, - 90 -there may be items not covered by the examples given that are nonetheless confidential . The answers to this statement give a more comprehensive assessment. The second question asked the managers to indicate which features exist in their company to promote access to information and how effective these features are. The responses to this question were used to supplement the data gathered in section II of the questionnaire. (iv) Questions about the company, the work, and the manager The questions in this section follow closely the discussion of the independent variables in Chapter Three. The general objective in this sec-tion was not to develop new instruments to measure these independent var i -ables but to use, where possible, established measures. Size of the Company Two measures of size were used in this study. The f i r s t is the total number of employees in the local organization (Bacharach and Aiken, 1977: 369). As i t is unreasonable to expect managers to know the exact number of employees, five equal interval categories were used with an additional open-ended category for large companies (Bouchard, 1972). The second measure, which is believed to be the more important one, is the number of employees in the manager's department and each subject was asked for an exact rating. Shape of the Company The horizontal differentiation is measured by the number of depart-ments and this was done by asking the subjects to c i rc le a number up to 10. If the number of departments is greater than 10 (the normal case in large companies) the subjects were asked to specify the number. Vertical differ-- 91 -entiation was obtained from the number of levels of authority from the department head to the lowest category of worker. Original ly , the question was put without the word "authority" and several subjects in the pi lot test interpreted this to mean job grades. F inal ly , the same part asked the sub-ject to indicate which level he/she is at, using "1" to represent the de-partment head or higher. Routineness.of Technology The six items to measure routineness of technology at the individual level came direct ly from Bacharach and Aiken (1977) who follow the t rad i -tion of Hage and Aiken (1969) in measuring technology at an individual leve l . Their instrument was adapted to include seven-point Likert-like rating scales to provide consistency with the previous sections. Three of the statements were reversed. Decentralization of Authority To measure decentralization of authority, an instrument was used that was f i r s t developed by Hage et a l . (1971). They asked managers to respond to whether they could make decisions in various areas (18 in a l l ) . The instrument was modified In two ways during the pi lot study. F i r s t l y , sub-jects found that for some areas they could make decisions but only after they had referred the matter to their superior. Secondly, several subjects had trouble with the word "sanctioning". The f i r s t problem was eliminated by asking them to respond YES/NO to whether they could make decisions in these areas without reference to a superior. The second ambiguity was cleared up by changing the word "sanctioning" to "d isc ip l in ing" . These were both considered minor changes that should not materially affect the instrument. - 92 -Retrieval Technology The categories for retrieval technology were developed from those suggested by Burch and Strater (1974) and Aguilar (1967). An additional question was appended to ask the managers to estimate, to the nearest ten percent, the frequency of use of the four technologies for work-related information. Some concern was expressed prior to the pi lot study about the feas ib i l i t y of using this question but several subjects in the pi lot test commented on i t s usefulness in interpreting the other questions on retrieval technology. Attitudes Towards Data-Sharing Attitudes to data-sharing were measured in two ways. F i r s t l y , the "norms" of data-sharing were measured with a single question from a study of communication behaviour (Dewhirst, 1971). Attitudes of trust and open-ness of communication were measured by four questions adapted from another communication study (O'Reilly and Roberts, 1976). Once again, a seven-point Likert-like rating scale was used for consistency with the rest of the instruments. No particular type of data was focussed upon in these questions. The measures for routineness, decentralization, and attitudes to data-sharing were subjected to an intertest r e l i ab i l i t y measure (Siegel, 1956:229). Maturity of the Information System Two measures were used to measure maturity of the computer information system. The f i r s t measure of maturity was the number of years computers have been used for technical, or administrative functions. The second ques-tion requested an estimate of the number of years since most of the original - 93 -computer problems were overcome and the computer was accepted into the normal procedures of the company. Subjects were asked to tick a box with a given range associated with i t . For the second question, an additional box was given for subjects to indicate i f major problems s t i l l exist . Biographical Details The f inal part of the questionnaire requested biographical details from the subjects. The items were; age, sex, highest level of education, job t i t l e , number of years at present leve l , department, and company type. These are largely used for sample reporting purposes. "Age", "number of years in present posi t ion" , and "department" were used in testing specif ic hypotheses. 2. Pi lot Study Eleven subjects were used as a pi lot test to reveal any weaknesses in the instrument. This resulted in the following discoveries and changes in addition to the ones mentioned above in the text: 1. The instrument was too long. At nineteen pages i t was fe l t that sub-jects would resist answering the questionnaire. The length of the ques-tionnaire was cut by two methods. F i r s t l y , the sections on general infor -mation (non-personal) were dropped. Although the subject of access to general information was of some interest, i ts elimination was made possible by the questions in section III on features to promote access to data. Additionally, several subjects mentioned their d i f f i cu l ty in understanding exactly what general information was. Secondly, the whole questionnaire was retyped on large-sized paper and then photo-reduced by 25 percent. To-gether, these methods produced a questionnaire that was nine pages in - 94 -length without any loss of readabil ity. Although the number of questions had only changed s l ight ly , i t was fe l t that the new questionnaire was of greater acceptability than the older version. Timing results indicated that most managers could complete the questionnaire using twenty to forty minutes. This was important because the data collection involves a mail-ing sample as well as a number of interviews. F inal ly , a covering letter was added to the instrument. This explained the purpose of the study and assured the subjects of the confidential ity of their responses. The letter also offered them a final report comparing access across several companies. 3. The Research Interview Because i t was not possible to gather a l l of the data required on access using a questionnaire only, one part of the sample was interviewed. For both the p i lot test and the main study, these subjects were contacted by let ter , telephone or both. If they agreed to participate in the study, a questionnaire was sent to them by mail ( i f they had not already been sent one) which they were asked to complete prior to an interview. The interviews were arranged over the telephone and they were normally scheduled during work hours. The interview method is what Bouchard (1972) cal ls a type II, where specif ic questions were asked by the researchers but the subjects could respond in an open-ended manner. This leaves considerable f l ex ib i l i t y for the researcher and the subject to explore areas that require extensive elaboration. The responses from the questionnaire represent the more structured approach (Type I). Additionally, a method of questioning called "funneling" was adopted where possible. This method is to sequence - 95 -questions from the most general type f i r s t (e.g. does the company have a formal policy on access to information?) leading to more specif ic questions later (e.g. can you give specif ic examples of the policy formation process?). Tandem interviews are reputed to be extremely effective especially at the exploratory stage (Bouchard, 1972; Kincaid and Bright, 1957). This method was adopted for some of the pi lot study interviews, otherwise solo interviews were employed throughout. The length of the interviews gen-eral ly f e l l within the range of 40-60 minutes although some subjects were so enthusiastic that up to ninety minutes would be used in the course of the interview. The objectives of the interviews were as follows: 1. They enable the subjects to clear up any d i f f i cu l t i es they may have had with the questionnaire and therefore increase the val idity of their responses. Few subjects found any d i f f i cu l t i es and where they did these were only minor points. The majority of subjects had l i t t l e d i f f i cu l ty in both reading and answering the questionnaire. 2. The interviews allow the researcher to supplement the questions in the questionnaire. For example, several of the respondents had been in the organization for a great deal longer than they had been in their pre-sent position. It was apparent that general knowledge about the company and the contracts made over a number of years in several departments was important to the subject's view of access to information. 3. They lead the interviewer to new areas of importance to access not covered in the questionnaire. For instance, industry competition seemed to be an important factor in influencing the policy and practice of access. Although this was not covered direct ly in the questionnaire, i t did become apparent during the course of the interviews. - 96 -4. The interviews reveal many anecdotes about access to information. For example, several subjects mentioned examples of information leaks and how these led to policy or practice changes. These examples are invaluable to the understanding of access and point to further areas for study. At the completion of each interview, the questionnaire was collected i f i t had been completed or i f i t had not, a return stamped-addressed envelope was lef t with the subject. 4. Population of Subjects Subjects were chosen from a population of approximately 1,150 middle managers in various companies throughout the province of Brit ish Columbia, Canada. For the mailing sample, subjects were systematically selected from two mailing l i s t s . For the interview sample (questionnaire and inter-view) the subjects were chosen judgementally from middle managers in the lower mainland of Brit ish Columbia. The cr i ter ia for selection in the interview sample was to use middle managers from a wide variety of companies and departments. The in i t i a l contact in most of these companies was the data processing manager and further subjects were sol ic i ted from these contacts. In order to discover any differences, the samples were subject to an inter-sample difference test for each variable. The total number of questionnaire subjects selected from the popula-tion was under 500. This number was chosen on the assumption that a mini-mum response rate of twenty percent would provide a suff ic ient number of useful questionnaires to enable a meaningful interpretation of the results and to gain increased significance from the measures of the relationships. - 97 -CHAPTER FOUR THE FINDINGS I. The Sample 1. Response rate For the total sample, 466 questionnaires were distributed to managers. One hundred and seventy usable scripts were returned, giving an overall response rate of 36.5%. In the case of the interview sub-sample, a total of 55 questionnaires were sent to managers and 53 interviews were conducted. From this sample, 51 valid questionnaires were obtained for a response rate of 92.7%. Because the response rates varied somewhat between the sub-samples (one interview, two mailings) and because of the different selection tech-niques employed, a K-sample median test was conducted for each dependent variable. Significant differences in seven percent of the variables were obtained (10 out of 135). This inter-sample difference was judged to be suff ic ient ly small to enable the sample to be treated as a whole in sub-sequent analyses. 2. Total sample The distributions by industry and departments of the responses received from the total sample are given below in tables I and II. - 98 -Table I Composition of Total Responses by Interview Forestry 11 Mining 6 Manufacturing Industries 17 Construction Industries , 6 transportation, Communication, and other U t i l i t i es 29 Trade (wholesale and retai l ) 17 Finance, Insurance, and Real Estate 19 Community, Business and Personal service Industries 30 Public Administration and Defence 35 Total 170 The companies ranged in size from less than ten employees to twenty-three thousand. The mean size was 2,537 while the median was 503 employees. Table II Composition of Total Responses by Department Marketing 8 Medical 2 Sales 5 Data Processing 14 Finance 24 Transportation 2 Accounting 42 Maintenance 1 Purchasing 3 Public Relations 3 Personnel 9 Warehouse 1 Customer Service 5 Education 1 Engineering 7 Training 1 Labour Relations 2 Credit '1 General Administration 25 Planning 7 Building 2 Production 1 Real Estate 4 Total 170 Departments varied in size from one to over 1,000 employees. mean was 63 while the median was 18 employees. - 99 -3. Interview sample The distributions by industry and departments of the responses received from the interview sample are given below in tables III and IV. Table III Composition of Interview Responses by Industry Manufacturing Industries 2 Transportation, Communication, and other U t i l i t i es 16 Trade (wholesale and retai l ) 8 Finance, Insurance and Real Estate 5 Community, Business and Personal service Industries 10 Public Administration and Defence 10 Total 51 Table IV Composition of Interview Responses by Department Marketing 2 General Administration 8 Sales 1 Medical 2 Finance 4 Data Processing 13 Accounting 4 Maintenance 1 Purchasing 2 Public Relations 3 Personnel 3 Warehouse 1 Customer Service 4 Education 1 Engineering 1 Planning 1 Total - 100 -11. Profiles of Access and Authority In this section we report on the differences in authority and barriers to access of information perceived by managers. The types of information that are considered are given below: Case.1: Confidential information required for the manager's job Case 2: Confidential information not required for the manager's job Case 3: Non confidential information required for the manager's job Case 4: Non confidential information not required for the manager's job Case 5 Case 6 Case 7 Own personal details Subordinates' personal details Other managers' (peers') personal details, The indices used to measure authority for cases 5, 6, and 7 were simple linear combinations of four measures. The four measures were job status (salary), job performance, general comments, and biographical detai ls . These combinations were submitted to a r e l i ab i l i t y test (Siegel, 1956:229) and their components were found to be s ignif icant ly associated with one another (p < . 0 1 ) . 1 . Types of information chosen by the managers In section I of the questionnaire (Appendix I), the managers were asked to tick those types of information that were considered confidential to their organization. The following types of information were subsequently chosen by managers to represent important examples in their companies of confidential and non-confidential information: (i) Confidential Information n % Minutes of Board Meetings 31 18.2 Detailed Sales Reports 16 9.4 Customer Lists 14 8.2 Pricing Formulae 14 8.2 "Other" category 45 26.5 - 101 -( i i ) Non Confidential Information n % Production Figures 40 23.5 Inventory levels 14 8.2 Machine Service Reports 14 8.2 Supplier DetaiIs 13 7.6 "Other" category 27 15.9 Discussion: Almost one in five subjects chose minutes of board meetings as the most important example of confidential information in their company, suggesting the 'important and strategic nature of such details to many companies, even though these details are often terse or highly summarized. Similarly, detailed sales figures were prominent in the choice of several managers. Both of these items of information featured highly in the inter-views. An analysis of the "other" category for confidential information revealed that approximately 50% of subjects chose types of information that were peculiar to their companies. 2. Distributions of authority and access measures (i) Authority Using a Likert-like seven-point rating for authority to access the seven types of information, the following median values were recorded with 1 and 7 representing the lowest and highest possible authority, respectively. Percentage of subjects with: Authority Lowest Highest Rating Authority (1) Authority (7) Case 1: Confidential informa-tion required 6.68 7.1 63.3 Case 2: Confidential information not required 3.32 28.9 24.7 Case 3: Non confidential informa-tion required 6.89 4.3 74.4 Case 4: Non confidential infor-mation not required 6.54 4.8 51.8 - 102 -Authority Rating Percentage of subjects with: Lowest Highest Authority (1) Authority (7) Case 5: Own personal details Case 6: Subordinates' personal 6.91 0.0 58.8 details Case 7: Other managers' (peers') details 6.95 1.2 69.6 1.48 45.8 4.8 (i i ) Access For each category of information, a general question on the ease of access was answered by the subjects. The question was worded, "Generally, i t is easy to get and use this type of information", with ratings of 1 and 7 representing extremely poor and extremely good access, respectively. For each of the cases 2, 4, 5, 6, and 7 a signif icant number of subjects''' indicated that information could not even be retrieved and consequently these subjects provided no response for the question on access. In the d i s t r i bu t i on of access given below, the adjusted values for lowest and highest access were recalculated on the assumption that i f managers cannot obtain a certain type of information, then their access is the lowest (1). The recalculated percentages are given in parentheses. Otherwise the figures are for subjects who had at least some access. The numbers of subjects who indicated that a particular kind of informa-tion could not be obtained at a l l were 66, 25, 5, 9, and 101, respectively. - 103 -Access Percentage of subjects with: Rati ng Lowest Highest (Median) access (1) access (7) Case 1: Confidential information, requi red 5.59 14.1 32.1 Case 2: Confidential information, not required 4.47 9.7(45.3) 24.3(14.7) Case 3: Non confidential , requi red 6.35 7.1 47.6 Case 4: Non confidential , not required 5.72 8.4(23.0) 41.3(34.7) Case 5: Own personal details 5.71 10.3(12.9) 32.7(31.8) Case 6: Subordinates' personal details 5.87 • 9.3(14.1) 36.6(34.7) Case 7: Other managers' (peers') 37.7(74.7) 10.1(4.1) personal details 2.57 The two groups of managers who could not obtain work-related information when i t is not required for their jobs (cases 2 and 4) wi l l be used in the analysis later to provide examples of those managers who are barred ex-p l i c i t l y from obtaining information. The median values for each variable used to measure access ( i .e . the barriers to access) are presented for the seven cases of information as Appendix II. In considering the barr iers, the lowest value 1 represents the highest access whereas 7 is the poorest access. ( i i i ) Techniques used by companies to prevent access In the interview sample, managers related details of several tech-niques.their companies used to regulate access directly and indirect ly . The details of these mechanisms are presented in Appendix VI and are summarized below in table V using the categories developed for the question-naire. - 104 -Table V Summary of Techniques to Prevent Access Direct Regulation Directives from management concerning access policy Channeling of external statements (use of public relations or a designated spokesperson) Sanctions or threats to employees Special report markings Indirect Regulation Retrieval 1. D i f f i cu l t or lengthy access procedures Physical security (guards, badges, designated areas) Use of computer technology to deter unauthorized access 2. Embarrassment Retrieval d i f f i cu l t i es (need to just i fy requests to othe 3. Permission of superior Retrieval d i f f i cu l t i es 4. Location Retrieval, problems (inconvenient distance) 5. Existence not known Promotion of ignorance of information existence Non-recorded information Visual access to information 6. Failure of authority Retrieval problems 7. Timing 8. Cost Use of computer to promote information cost awareness Use 1. D i f f i cu l t language, symbols, or jargon Coding of information 2. Irrelevant details Information overload 3. Missing details Use of summarized data - 105 -3. Profiles of authority and access The values presented in previous tables give a qualitative picture of the relative order of authority and access. Additionally the individual subject's ratings were also subjected to an inter-case Wilcoxon matched-pairs ranked-signs test (Siegel, 1956:75). If case 1, for example, has authority ratings consistently higher than case 2 at a .05 level of significance, this is denoted as 1 > 2 in the presentation below. (i) Ordering of authority The analysis produced the following order for authority In addition, case 6 was found to be signif icantly greater than case 5. The ordering shows that non-confidential information required for the managers' job has approximately the same ratings as personal information about themselves and about their subordinates with the one exception men-tioned above. These three types of information are authorized s i g n i f i -cantly more than confidential , required information. Confidential infor -mation which is required for the job,as expected5ihas higher authorization than non confidential information that managers do not require for their jobs. Confidential information not required has lower authority ratings than non confidential information which is not required. F inal ly , infor-mation concerning other managers is found to be s ignif icant ly lower in authorization than any other type of information.' ( i i ) Ordering of access The second analysis produced the following order for access. > 1 > 4 >,2 > 7 - 106 -3 > | 5 j > l > 2 > 7 Additionally, although cases 4, 5 and 6 were indistinguishable from one another, cases 5 and 1 were also not s ignif icantly different. The profi le shows that the best access was obtained by managers to non conf i -dential information that they require for their jobs. Access to this i n -formation was s ignif icant ly better than access to non-confidential infor -mation not required for the job, and to the f i r s t two types of personal information. These three categories were generally more accessible than confidential information required for the managers' jobs (see above for an exception to th is , though). However, access to confidential information was s ignif icant ly lower in the situation where i t is not required than when i t is required. Once again, access to other managers' personal details provided the lowest ratings of a l l . Discussion: The authority and access ratings provide good support for the hypothesized relationships. Authority to access confidential i n -formation is lower than the ratings for non-confidential information. Authorization to access work-related information is much higher than man-agers' authority to access information not required for their jobs. Also, authority of access to personal details of subordinates is higher than the authority to access one.'s own details (marginally higher) and those of other managers (substantially higher). The pattern for the general questions on access parallels the ratings for authorization. The relationships that were hypothesized for authority and access across the information types formed only a partial ordering of the ratings. Using the Wilcoxon method i t was possible to obtain a complete ordering of - 107 -the ratings. This raises few surprises but there are some interesting differences between the order for authority and the order for access. F i r s t l y , for authorization there is a clear pr ior i ty for work-related in -formation. Cases 3, 6, and 1 are a l l relevant to the work of managers and feature highly in the order. . The exception is case 5, information about themselves, which is rated approximately level with the authority to access non-confidential information and authority to access subordinates' detai ls. Both types of information (cases 4 and 2) which managers do not require for their jobs are generally low in authority. The lowest authority of a l l was found for access to other managers' detai ls . This means that managers perceive that they have s ignif icant ly higher authority to access confidential information not required for their jobs than they have to each others' personal detai ls . Secondly, a similar pattern appears in the ab i l i ty to access information. However, non-confidential information has obtained the dominant rating in terms of ease of access. Case 3, non-confidential information that is required by the managers, is easier to access than any other type. When the non-confidential information is not required for the manager's job (case 4) the ease of access is only s l ight ly lower. Towards the lower end of the scale of ease of access we find con-f idential information (cases 1 and 2) and associated with this is the ease of access to other managers' details (case 7). The ease of access ratings do not include those subjects who could not obtain the specified information. The following conclusions are suggested by this evidence. F i r s t l y , managers in the sample were authorized to have access to information needed for their jobs and generally had lower authority or no. authority to access information not needed for their work. Secondly, non confidential - 108 -information was easier to obtain and use than confidential information. For example, managers had greater access to non confidential information not required for the job than to -confidential information required for the job. This probably reflects the greater security and secrecy that is applied to confidential information and implies that the regulation of confidential information is generally indirect. Thirdly, personal information about themselves or their subordinates is perceived by the managers in a similar way to non-confidential information. Generally, access is authorized and i t is reasonably easy to obtain. Fourthly, access to personal details about other managers is almost universally unauthorized and extremely d i f f i -cult to obtain in practice even when i t is authorized. For this type of information the closest parallel is confidential information where i t is not required for the managers' jobs. However, access to other managers' details would seem to be treated with even greater care than confidential 2 information. Lastly, the managers generally believe that their authority and abi l i ty to access information concerning themselves is relat ively high, providing some additional support for a similar finding from another Canadian study (Canadian Task Force, 1972). Further evidence was obtained from the interviews. These results generally support the above conclusions and provide some of the details that were not possible to col lect in a structured questionnaire. For work-related information (cases 1-4) over f i f t y percent of managers claimed that access in their company is supplied on a "need to know" basis. Sometimes this rule was applied to confidential information only. In other cases i t would be applied to both confidential and non confidential This supports the assertion made by Forrester (1965). - 109 -information alike-. Apart from denying authority to access information that is not needed, the companies in the interview sample used a variety of indirect regulation mechanisms in order to seal off access to certain types of information (ignorance of existence, non-recording of information, coding, physical security, e tc . ) . Each of these is fu l ly discussed in Appendix VI. The point to be made here is the clear separation made in many companies between information needed and not needed for a manager's job. A few examples wil l i l lustrate the point. In a department store, the manager received detailed monthly operating figures for her department and for similar departments in other store locations. Similar reports for other departments ( i .e . not needed for the manager's job) were not available. In several companies, the minutes of manager's meetings were only available to managers in those departments which were materially affected by items discussed at the meeting. Although the pervasive rule was access to information on a "need to know" basis, there were exceptions. Two managers from the same company noted that most work-related information was available to them whether they needed i t or not. Their concern was that they were sent too much un-necessary information (Ackoff, 1967; Mintzberg, 1975). Usually this i n -formation would not be confidential . However, in one other company, man-agers openly received confidential prof i t plan information regardless of whether i t was needed for their jobs. In contrast to policies concerning access to work-related informa-t ion, personal (personnel and payroll) information on employees at a l l levels was often the subject of elaborate rules of access. In ten of the companies, the rules of access were written down and made available to man-agement and occasionally to other employees. In al l of the companies - 110 -visited the managers understood that rules of access to this kind of infor -mation did exist even i f they were not always personally familiar with the rules. All of the managers interviewed were provided with access to per-sonal information at least on a "need to know" basis. At least some infor -mation on subordinates and other employees who are seeking employment in the manager's department was always available (cf. Cary, 1976). The rules on the managers' access to their own records varied a great deal from company to company. In nine companies managers had the right to access information kept on them. In four companies the managers were s t r i c t l y for -bidden from such access. Managers in three companies commented that they had never tr ied to access their own record and did not know what their rights of access were. Although the questionnaire results make i t appear as i f access to their own records is good for managers, the interview findings indicate that many of these managers may be assuming that their authority and abi l i ty to access this information is adequate. In practice they may have never attempted to retrieve this information. Almost a l l managers agreed that the one area they were prevented from accessing was information, par t i -cularly on sa lar ies , about other managers both at their own level and at a higher leve l . The exception to this was in those organizations where they are required by law to publish salary and expense payments. This strongly supports the questionnaire findings. In three of the companies the rules of access to personal informa-tion were very elaborate. This situation was found where the rules of access were bui l t into the collective agreement with the union. For example, in an industrial board the rules of access to personal information covered - I l l -grievance procedures. Detrimental comments on employees have to be purged after a set time period. The agreement covers the extent of personal data collection and excludes management's keeping of informal records. Only the formal records can be used in grievance procedures. Personnel and payroll information were often treated differently in the companies. In a l l but three companies, personnel information was kept in a central department in a documentary form. Several managers commented that most of this information could not be computerized. In a l l of the medium or large companies, payroll information was computerized and was often administered by a payroll department. The amount of per-sonnel information collected varied a great deal between the companies. In one privately owned company very l i t t l e personal information was used. Neither sick leave was recorded nor were medical records kept. In con-trast , the human resources department of a large corporation kept exten-sive records in an effort to be seen to be fa i r to the employees, includ-ing managers. In one company, the managers were required to take a psycho-logical test. The results were not shown to the individual concerned. The same company also kept extensive records on discipl inary actions. This was apparently in response to a decision in the courts to tighten labour regulations concerning grievance processes. The important factor that determined the access policy on payroll information was salary detai ls . The exceptions to this were found in those organizations covered by the need to report the salaries of man-agers publicly or for salaries of employees covered by col lective agree-ments. In a l l other cases, salaries of middle or higher managers were treated as very sensitive information. Again, this was reflected in the questionnaire findings. In two cases, talking to other employees about - 112 -salaries was a cause for discipl inary action. In four of the organizations there were at least two payroll systems corresponding to the levels of the individuals in the organization. For example, one company had three levels of payrol l . The f i r s t level was a general payroll and was used largely for unionized workers where the pay rates are published and reason-ably well known internal ly. The next level was a "semi-private" payroll for middle management which was administered by the control ler 's secretary. Some status was attached to being on this payrol l . Managers were s t r i c t l y prohibited from accessing each other's detai ls . Indeed, when one manager wanted the general salary levels, not of specif ic individuals, but of junior managers in another department for comparative purposes he was told that the information was not available to him. The last and highest level of payroll was for executives only and the president's secretary handled this through an outside bank. An interesting situation arose in some companies where the unionized workers' salaries began to exceed those paid to some managers. The access to this information was often only one-way; the managers knew what the workers were paid because i t was published in an agreement that many managers had, but the reverse did not apply. In one company this had led to some f r i c t ion . In another company, the problem was solved by tying the managers' salaries in the unionized department to union increases. Because the policy was not well known in the company, there were no complaints from managers in other departments. On the whole, most managers were content with the policy of secrecy that surrounded their salaries. It is clearly beneficial to the companies as many anomalies concerning salaries are bound to arise over years of employment and the effect of making salaries public would tend to push - 113 -levels up as managers aim for compatibility with each other. But i t was surprising that the general comment voiced by the managers was that i t would only make them discontent i f they knew what other managers earned. The personnel department or i ts equivalent was more often than not centralized in the organization. In only three of the companies inter-viewed were the personnel records decentralized. In one of these three companies this was in accord with the organization's general decentraliza-tion policy. This company issued only guidelines to departments on the ways in which personnel and payroll information might be handled, although in practice most departments used the guidelines as directives. In the other two companies, the managers kept their own o f f i c i a l personnel records on employees. These were copies of those kept in a central system. In a l l other cases the personnel department kept the records. Depending on the location of the department the managers had varying degrees of ease of access. In one situation the department was moved to a separate loca-tion several miles from the main building; in an another example some of the important back records on leave and attendance were maintained in Ottawa. The d i f f i cu l ty of obtaining this information, particularly on subordinates, was a key factor in the manager's decision to maintain his own informal records on his s ta f f , a widespread practice in the companies which wil l be discussed later (Section V). For those using decentralized systems the incentive to develop an informal system was much less and in the three cases recorded, no informal record-keeping was reported. Some of the indirect means of controll ing access to personal infor -mation are parallel to the techniques employed for work-related informa-tion (see Appendix VI). Where the record-keeping is decentralized, some of these barriers to access did not apply. In general, the lowest barriers - 114 -to access were associated with retrieving subordinates' details and this is in accordance with the emphasis found in companies on access to information that a manager "needs to know". Obtaining his/her own record was sometimes more d i f f i cu l t and getting information on other managers was normally impossible. Summary The results presented in section II can be summarized as follows: 1. Organizations use a variety of techniques to prevent access to information (see Table V). 2. Access to work-related information is regulated more in -direct ly than access to personal information. The major exception to this is the "need to know" rule which is used widely to regulate access to al l types of informa-tion. 3. Authority and access ab i l i t y vary s ignif icant ly according to the type of information. Again, the "need to know" rule applies generally to both authority and access. 4. Access to confidential information is s ignif icant ly more d i f f i cu l t than access to non confidential informa-t ion. Access to work-related information is s i g n i f i -cantly better than access to information that is not work-related. 5. ' Managers' access to their own f i l es or those of their subordinates is generally regulated in organizations in the same way as access to non confidential information. 6. Access to details about other managers, particularly salary deta i ls , is largely unauthorized and extremely d i f f i cu l t in practice. - 115 -III. Constituents of Access In this section, the findings were analysed to provide evidence for two types of relationships. F i r s t l y , the connection between direct and indirect regulation of access was sought. That i s , is a manager's authority to access information matched by the ab i l i t y to retrieve and use this information? Secondly, which components of indirect regulation of access are more important in determining overall access to a particula type of information and how do the relative weights of these components vary across the different types of information (cases 1-7)? 1. Direct and indirect regulation of access The following two variables for retrieval and use were constructed from a simple linear combination of the eight variables for retrieval and the eight variables for use that were defined in Chapter 1. This was repeated for each type of information. a. Retrieval b. Use 1. D i f f i cu l t procedures 1. Errors 2. Embarrassment 2. D i f f i cu l t to compare 3. Permission of superior 3. Missing details 4. Location 4. Biased 5. Existence not known 5. Badly presented 6. Authority fa i lure 6. Bad layout 7. Untimely 7. Jargon 8. Cost 8. Irrelevant details - 116 -A high value for retrieval or use (maximum = 7) represents the highest measure of a barrier or the poorest access, whereas the value 1 represents the lowest measure of a barrier or the best access. A table of intercorrelations of barriers is presented in Appendix 111 The type of information chosen for this purpose was non confidential , needed for the manager's job. An analysis of the other types of informa-tion showed similar patterns of intercorrelations. Several features of this table are worth noting. F i rs t l y , the majority of correlations are less than x = .250; the highest being T = .550. Secondly, the higher inter-correlations tend to be among either retrieval or use variables. (i) Distributions of retrieval and use As the components of retrieval and use were aggregated to form new variables, an intertest r e l i ab i l i t y test was performed (Kendall) and the results, given in Appendix IV, show high significance for a l l of the new variables (P < .001). Additionally, the variables were subjected to a Kolmogorov-Smirnov one-sample test (Siegel, 1956:47) to investigate i f the new, composite variables for retrieval and use could be considered normally distr ibuted. The findings supported the normality assumption for a l l of the variables except for the one that measures the use of con-f idential information required for the job. These findings were used as the basis for the multivariate analyses that were subsequently performed linking access to i ts components. Retrieval Mean Use Mean Case 1 3.09 Case 1 2.94 Case 2 3.57 Case 2 2.94 Case 3 2.63 Case 3 2.86 Case 4 2.93 Case 4 2.90 Case 5 2.95 Case 5 2.79 Case 6 2.73 Case 6 2.80 Case 7 4.23 Case 7 2.99 - 117 -Discussion: The mean values for Retrieval follow the same pattern for the general question on access presented in section II. However, the same is not found for the mean values for Use which show a remarkable consis-tency across the different information types (cases 1-7). This suggests that the managers perceived variations in access (indirect regulation) across information types as differences in retrieval d i f f i cu l ty but not as variations in the d i f f i cu l ty of using information. ( i i ) Relationship between authority, ret r ieva l , and use Kendall rank correlations (x) were performed between authority and the general question on access and between authority and the composite variables for retrieval and use. Case 1: Confidential information required x sig 1. General access .2156 .001 2. Retrieval -.2728 .001 3. Use -.2257 .001 Case 2: Confidential information, not required 1. General access .3271 .001 2. Retrieval -.3429 .001 3. Use n.s. Case.3: Non confidential information, required 1. General access .2007 .003 2. Retrieval n.s. 3. Use n.s. Case 4: Non confidential information, not required 1. General access .1395 .045 2. Retrieval -.2359 .001 3. Use -.1627 .012 Case 5: Own personal details 1. General access .2657 .001 2. Retrieval -.3057 .001 3. Use -.1319 .027 Case 6: Subordinates' personal details 1. General access .3682 .001 2. Retrieval -.3351 .001 3. Use -.1272 .001 - 118 -T s i g Case 7: Other managers' personal details 1. General access 2. Retrieval 3. Use .2820 -.4159 n.s. .002 .002 Discussion: Several comments can be made based on these findings. F i rs t l y , a l l s ignif icant relationships (p < .05) are in the anticipated directions. That i s , authority and the general access questions are al l positively related while authority and the barriers (retrieval and use) are negatively related. This means that higher authority is generally associated with better access and lower barriers to access. Secondly, the relationships are only of moderate strength (x) and in some cases (e.g. Case 3) are not even signif icant. In section II i t was noted that the profi les of authority and access differed and here is further evidence for this difference. Thirdly, there is a systematic difference between the relationships for retrieval and use. For every case where s i g n i f i -cant results were found, authority and retrieval are more strongly related than authority and use, providing support for the conclusion presented above, namely, that retrieval plays a larger part than use in determining access. 2. Constituents of access (i) Access and individual barriers Although a multiple regression could not be run linking access to 3 i ts constituents, Kendall correlations were run and the strongest re la -tionships (all s ignif icant at p < .05) are presented in Appendix V. For The constituents had in many cases a J type of distribution which cannot be transformed - see Rummel, 1970:286. - 119 -each case, the letters R and U denote whether a variables measures retrieval or use, respectively. Discussion: F i r s t l y , a l l of the relationships are in the anticipated direction. Because the general question to measure access is measured positively while the individual barriers are measured inversely, the re la -tionships were expected to be inverse, and this was found to be the case. Secondly, there is a dominance displayed by certain of the retrieval var i -ables for a l l types of information that is not paralleled by the variables to measure use. Analysis of the data suggests that d i f f i cu l t procedures come up most consistently across al l types of information as providing the largest barrier to access taken individually. For work-related infor -mation, location was almost as important a barrier to access as d i f f i cu l t procedures, confirming the work of some ear l ier studies (Simon et a l . , 1954; Greene, 1973). For access to personal data, bad timing, embarrass-ment, and authority fai lure were individually important in determining the inconvenience of access to this information. The interviews were helpful in interpreting some of these results. For example, a common barrier to the access of personal information was the manager's need to just i fy a request for information to a member of the personnel department. This vetting of requests was particular prevalent when managers made requests to view their own f i l es (cf. Canadian Task Force, 1972). In one company, such requests to see their own details have to be reviewed by the personnel manager. In another company, managers have to get the approval of their vice president before the personnel department wil l release the information. Clearly, such a method of regu-lating access would explain why bad access to personal records was asso-ciated with d i f f i cu l t procedures, embarrassment, and authority fa i lure . - 120 -Although timing was not mentioned in the interviews with respect to the access of personal records, the questionnaire results indicated that i t too is strongly related to the inconvenience of access. Its absence from the interview data is thought to be explained by the frequently mentioned problem of the location of personal records. Badly located records are one cause of the bad timing of information. This is especially true for subordinates' records which are normally required by managers on a daily basis. In a l l but three of the large companies v is i ted , the personnel department contained the only copies of personal records. This resulted in many managers being a considerable distance from information that was needed frequently. In one company, the personnel department was located a considerable distance (several miles) from the main building, and although a courier service was provided, the problems of access were judged by managers to be high. ( i i ) Access and retrieval and use By exploiting the normality of the distributions of the composite measures of retrieval and use, seven multiple regressions were run using the basic regression equation: Access = g,*retrieval + e0*use The following regressions were obtained. Case 1: Confidential information, required. N = 167 Access = -.479*retrieval - .096*use = 1.609 (n.s.) R = 27.6% r = .4818 - 121 -Case 2: Confidential information, not required N _ i n ? Access = -.553*retrieval - . l l l *use F l j l 0 1 = 42.787 F 1 J 0 0 = 1.729 (n.s.) R2 = 34.22% r = .2917 Case 3: Non confidential information, required N _ l f i 7 Access = -.462*retrieval - .146*use F, i a a = 32.085 F, n c c = 3.226 .(n.s.) 1 ,166 1 ,165 R2 = 30.44% r = .6122 Case 4: Non confidential information, not required N _ -,n9 Access = -.537*retrieval - . l l l *use F1 J 4 1 = 46.795 F l 5 l 4 0 ' = 1 - 9 9 4 (n-s-) R2 = 35.10% r = .5025 Case 5: Own personal details N _ , f i o Access = -.535*.retrieval - .047*use n i 0 6 F, 1 C 0 = 47.885 F, , c l = .037 (n.s) 1,162 1,161 R2 = 26.23% r = .4797 Case 6: Subordinates' personal details N = -.eq Access = -.502*retrieval - .019*use n F7 , , q = 43.000 F, 1 K 7 = .064 (n.s.) 1,158 1 ,157 R2 = 24.43% r = .4185 Case 7: Other managers' personal details co Access = -.599*retrieval - .228*use I N D a F1 6 7 = 30,645 F1 6 6 = 4.430 (sig. p < .05) R2 = 32.09% r = .3316 For a l l but case seven, the Betas for use were not signif icant at the .05 level . For a l l cases, the Betas for retrieval were highly s i g n i f i -cant (p < .001). The Pearson product-moment intercorrelations between - 122 -retrieval and use are given as ' r 1 . For a l l but case 7, the R is pre-sented for the retrieval variable and does not include the contribution from use. Discussion: For six of the seven cases i t is possible to "explain" from 24% to over 35% of access using just the single composite variable for retr ieval . Once again, here is further evidence that access is per-ceived by managers as being governed largely by the problems of retrieving the information. Only for the seventh case did the variable use enter s ignif icant ly into the regression equation. Secondly, access is better explained by retrieval for those types of information that the managers have no need for or have no rights to (Cases 2, 4, and 7). For each of 2 these cases, R exceeds 30%. This suggests that the problems of retrieval are brought more sharply into focus for those types of information to which access is not needed. One can also suggest that for information re-quired for the job, managers have learned how to reduce the barriers ("costs") of retrieval and use. Summary The results presented in section III can be summarized as follows: 1. Access to information is inversely related to the problems of retrieval and is almost independent of the d i f f i cu l t i es in using information. 2. General access, ret r ieva l , and use are only moderately related to the authority to access each type of information. 3. General access is "explained" better by the measure for retrieval for those types of information managers do not need to know than for those types of information that are needed for the job. - 123 -IV. The Determinants of Access In the previous chapter, relationships were hypothesized connecting access (direct and ind i r e c t regulation) to i t s determinants. The f i n d -ings for each variable are presented below together with any appropriate material from the interviews. Unless otherwise stated, the correlations 4 are a l l Kendall's T and are s i g n i f i c a n t at the .05 l e v e l . Before the detailed findings are given, the results of the study are presented in summarized form in table VI. This shows each hypothesis and an assessment of the support i t received based on the findings. 4 For authority and access a positive sign represents an increasing r e l a -tionship. For individual barriers and retr i e v a l and use, a positive sign means increasing barriers or more inconvenient access. - 124 -Table VI Summary of Results Hypothesis Structural Variables Hy The "costs" of access to information in larger companies/departments are higher than those experienced in smaller companies/departments. H 2 : A manager's authority to access non job-related information is lower in larger companies/departments than in smaller units. H :^ A manager's authority to access i n -formation is higher in those com-panies with greater decentralization of authority. H :^ A manager's "costs" of access to information are lower where authority is decentralized compared with those cases where authority is not decen-tra l ized. H :^ A manager's "costs" of using infor -mation are greater in companies with greater vertical di f feren-t iat ion. H :^ For managers performing non-routine jobs in companies with high hor i -zontal differentiat ion the "costs" of access to job-related informa-tion are higher. H ? : A manager who performs more routine tasks faces lower "costs" of access to information required for the tasks and higher "costs" to access information not required for the tasks. Result Some support. There was some indication that access was s l ight ly worse in larger organ-izations. Some support. For confidential information authority to access information not required for the job was lower in larger companies. Some support. For information about other managers there was an increase in authority;for non confidential information not required for the job there was a decrease. Some support. Some procedural barriers to access were reduced. Cost of non confidential infor -mation was signif icant ly higher. Not supported. Barriers to using information were lower in organizations with greater vertical d i f ferent iat ion, Not supported. Substantial support. There is a lowering of "costs" of access to non confidential information but an increase in "costs" for con-fidential information. Infor-mation not required for the job has higher barriers to access. - 125 -T a b l e VI ( c o n f H y p o t h e s i s A t t i t u d e s t o d a t a s h a r i n g H • Managers i n d e p a r t m e n t s where t h e norm o f d a t a s h a r i n g i s h i g h f a c e l o w e r " c o s t s " t o t h e r e t r i e v a l o f j o b - r e l a t e d i n f o r m a t i o n and have h i g h e r a u t h o r i t y t o a c c e s s t h i s i n -f o r m a t i o n . H q : Managers i n d e p a r t m e n t s where t h e a t t i t u d e t o d a t a s h a r i n g i s c h a r -a c t e r i z e d by t r u s t and o p e n n e s s , f a c e l o w e r c o s t s t o t h e r e t r i e v a l and use o f a l l t y p e s o f i n f o r m a t i o n and have h i g h e r a u t h o r i t y t o a c c e s s t h i s i n f o r m a t i o n . d) R e s u l t S u p p o r t e d . T h e r e i s a d e c r e a s e i n t h e a u t h o r i t y to o t h e r man-a g e r s ' d e t a i l s ( c a s e 7), o t h e r -w i s e t h i s h y p o t h e s i s i s s t r o n g l y s u p p o r t e d . Many b a r r i e r s t o u s i n g i n f o r m a t i o n a r e a l s o l o w e r . S t r o n g l y s u p p o r t e d . O n l y i n t he c a s e o f a u t h o r i t y t o a c c e s s o t h e r manage r s ' i n f o r m a t i o n was t h e r e a r e s u l t t h a t d i d n o t s u p p o r t t h e h y p o t h e s i s . T e c h n o l o g y o f a c c e s s H , n : Managers who p e r f o r m r o u t i n e j o b s Not s u p p o r t e d , and have a h i g h p e r c e n t a g e o f compu te r use f a c e l o w e r " c o s t s " o f a c c e s s t o i n f o r m a t i o n . H- , - . : Managers who use compute r s f r e q u e n t l y f o r t h e i r j o b s have l o w e r a u t h o r i t y t o a c c e s s i n f o r m a t i o n n o t r e q u i r e d f o r t h e i r j o b s and f a c e h i g h e r " c o s t s " f o r t h e r e t r i e v a l o f such i n f o r m a t i o n . S u p p o r t e d . Howeve r , a u t h o r i t y t o a c c e s s i n f o r m a t i o n n o t r e -q u i r e d f o r t h e j o b was n o t a f f e c t e d by the use o f compute r t e c h n o l o g y . H-.p : Managers i n compan ies where t h e com-p u t e r i z e d i n f o r m a t i o n s y s t e m i s immature b u t where t h e i r work d e -mands f r e q u e n t compute r use f a c e h i g h e r " c o s t s " o f a c c e s s t o j o b -r e l a t e d i n f o r m a t i o n . Not s u p p o r t e d . S t r o n g s u p p o r t was f o u n d , h o w e v e r , f o r t he a s s e r t i o n t h a t more mature sys tems p r o v i d e b e t t e r m a n a g e r i a l a c c e s s t o i n f o r m a t i o n . O t h e r d e t e r m i n a n t s H-,o : Managers f rom d a t a p r o c e s s i n g d e - Not s u p p o r t e d , p a r t m e n t s f a c e l o w e r c o s t s o f r e t r i e v a l t o a l l t y p e s o f i n f o r -m a t i o n compared w i t h managers f rom o t h e r d e p a r t m e n t s . H-, . : Managers w i t h g r e a t e r e x p e r i e n c e S u b s t a n t i a l s u p p o r t , have l o w e r " c o s t s " o f a c c e s s t o a l l t y p e s o f i n f o r m a t i o n . - 126 -1. Size of Company/Department (i) Company size a. Access There were no signif icant relationships between size and the general question on access. b. Specific barriers Case 2: Confidential, not re-quired None Case 4: Non confidential , not required Jargon +.1405 Existence not known -.1303 Case 6: Subordinates' personal details Jargon +.1662 Existence not known -.1379 Cost +.1708 Case 1: Confidential Informa-t ion, required None Case 3: Non confidential , required Jargon +.1198 Existence not known -.1202 Case 5: Own personal details Jargon +.1304 Existence not known -.1275 Cost +.1406 Case 7: .Other managers' details-. Jargon - + .'2141 Cost +.2449 c. Authority Case 2: Confidential information not required. -.1218 ( i i ) Department size a. Access No signif icant findings. b. Specific barriers Case 1: Confidential informa-t ion, required None Case 3: Non confidential , required None Case 5: Own personal details None Case 7: Other managers' details Cost +.2089 Case 2: Confidential, not required Existence not known -.1410 Case 4: Non confidential , required Existence not known -.1610 Cost +.1288 Case 6: Subordinates' personal details None - 127 -c. Authority None. Discussion: Generally the relationships between the two size variables and direct and indirect access provide only modest support for the f i r s t and second hypotheses. The Bacharach and Aiken study of communication behaviour and size found a similar lack of results for managers (Bacharach and Aiken, 1977). It was also expected that the work-unit size would be more strongly related to access than the size for the whole company. The results indicate that this is not so and that the reverse may be the case. Speci f ica l ly , for the f i r s t hypothesis (H-j), the results indicate that for larger companies, access to non-confidential information may be s l ight ly worse than in smaller firms. For the second hypothesis (H^) there is some support showing that a manager's authority to access conf i -dential information not required is s ignif icantly lower in larger companies. A common element in the specif ic barriers that did relate to size seems to be formalization. Large companies may tend to develop formal costing systems, including the chargeback of information costs. Size of the company is also strongly related to the number of departments (x = .5052) which may explain the increase in jargon. Each department is l ike ly to develop its own special ist language. A decrease in the authority to access confidential information not required for the job is another indication that more formal rules of access exist in larger companies. It is interesting to note, however, that the decrease in authority was not paralleled by a decrease in access ab i l i t y for confidential informa-tion not required by the managers. Size of the organization was mentioned twice in the interviews as having a direct impact upon the policy of access. In both cases the - 128 -companies had experienced rapid growth during which many policies had not been established. The growth from small, informal firms to medium-sized ones was accompanied by the companies gradually developing formal policy statements including those relating to access to information. Another result of size was the increased reliance placed upon formal documents for providing information. It would appear, therefore, that the size of a company and formaliza-tion have some association with respect to access. However, some of the more physical barriers to access such as location problems were not s ignif icant ly related to s ize , contrary to expectations. 2. Decentralization of authority a. Access Case 4 , non confidential information not required, had higher barriers to Retrieval (x = + . 1 5 8 8 ) . b. Specific barriers Case 1: Confidential informa- Case 2 : Confidential, not required Failure of authority - . 1 3 1 5 Case 3 : Non confidential , required Case 4 : Non confidential , not Cost +.1703 required D i f f i cu l t procedures +.1497 t ion, required None Location +.1478 Jargon +.1343 Cost +.1908 Case 5: Own personal details Case 6 : Subordinates' personal details D i f f i cu l t procedures =.1311 Di f f i cu l t procedure - . 2 1 0 8 Cost +.1184 Case 7: Other managers' details None c. Authority Case 4 : Non confidential information not required - . 1 4 9 8 Case 7: Other managers' job performance +.1265 Case 7: Other managers' general correspondence +.1357 - 129 -d. Further results Those who could not get confidential information when i t was not required for their jobs were from organizations that had signif icant ly more centralized decision-making networks than those with good access to the same information. Discussion: For confidential information, decentralized decision making is associated with easier retrieval procedures. Also managers who find i t relat ively easy to access confidential information when i t is not required,frequently have authority over several decision areas. For non confidential information not required for the job, the situation is reversed. Access is generally less authorized and retrieval is more inconvenient. A common finding for non confidential information is the increased cost awareness (also for case 5) for those managers with higher decision-making authority. This result would be consistent with the use of cost centres or prof i t centres in decentralized organizations where the cost of information is charged to the centre (Turney, 1977). In the interviews only two of the companies were found to be exten-sively decentralized. For one of these companies, the decentralization of authority enabled the managers to exercise considerable autonomy over access to information of a l l types. A drawback to such an arrangement occurs when information needs to be gathered centrally. This led to pro-blems of incompatibility in one company because of the variations in information formats used by different branches. For personal information, there are modest increases in authority over other managers' information and some reductions in the procedural d i f f i cu l t i es in accessing the managers' own records and those of their subordinates. These findings are consistent with the two hypotheses. - 130 -In one decentralized firm v is i ted , the senior managers were also given the authority to organize the personal record-keeping system and this clearly produced better access to these records for managers. Generally, there is some support in these findings for hypotheses three and four. The one type of information that contradicts these results is non-confidential information where i t is not required for the job. One possible explanation for this anomaly is that in decentralized organizations decision terr i tor ies are separable and v is ib le . The autonomy of one manager over the terr itory means the exclusion of others. Confi-dential information is often managed centrally for security reasons and is less prone to balkanization. A result that was not expected but which is nonetheless plausible is the increasing awareness of information costs in decentralized organizations. 4. Shape of the organization (i) Vertical differentiat ion (Number of levels) a. Access No signif icant findings b. Specific barriers Case 1: Confidential, required Permission of supervisor -.1308 Case 3: Non confidential , required Comparison problems -.1338 Missing details -.1295 Existence not known -.1891 Case 5: Own personal details None Case 7: Other managers' details None Case 2: Confidential, not required Missing details -.1691 Bias -.1856 Case 4: Non confidential , required Comparison problems -.1544 Location -.1488 Missing details -.1549 Widely known -.1763 Case 6: Subordinates' personal details None - 131 -c. Authority Case 7: Other managers' details -.1374 ( i i ) Horizontal differentiat ion (Number of departments) a. Access No signif icant findings. b. Specific barriers None c. Authority No signif icant findings d. Further Results For non confidential information not required for the job, better access is associated with those companies with more departmentalization. After a suitable division of the distributions for high and low number of departments and high and low routineness of technology, an ANOVA was per-formed using authority, retrieval and use as dependent variables. No signif icant interactions were found. Discussion: It would appear from the results that greater vertical differentiation is positively related to lower barriers to access of work-related information. Furthermore, a substantial proportion of the s ign i -ficant barriers are those related to the use of information (comparison problems, missing deta i ls , and bias). Both of these results are contrary to expectations. Two opposing arguments can be identif ied concerning the effect of the number of levels of authority upon managerial access. The f i r s t states that more levels of authority lead to greater distortion and omission of information as i t passes up the hierarchy from level to level (Wilensky, 1967; March and Simon, 1958; Rosen and Tesser, 1970; Berry and Otley, 1975). This leads to higher barriers to access particularly - 132 -those relating to the use of the information. The second, opposing argu-ment, emphasizes the limitations of human information processing (Simon and Newell, 1971). This argument would claim that without the f i l t r a -tion of information by lower level employees the manager would be deluged by too much raw data (Ackoff, 1967; Chervany and Dickson, 1974; but cf. Sorter, 1969). The hypothesis tested assumed that the effects described by the f i r s t argument were more important than the effects discussed in the second. On balance, however, i t would appear that access is improved where more f i l t r a t ion of information is l ikely to occur, i .e. where there are more levels of authority. The problems of processing biased, sum-marized information s t i l l remain for managers in hierarchies with many levels of authority. This may further suggest that counterbiasing is a feasible option for many managers facing this situation (Cyert and March, 1963; Simon et a l . , 1954). 5. Routineness of technology a. Access Case 2: Confidential information, not required +.1552 Case 6: Subordinates' details -.1556 Case 3: Use of non confidential information, required -.1075 b. Specific barriers Case 1: Confidential, required Errors +.1757 Medium of presentation -.1195 Authority fai lure +.1181 Case 3: Non confidential , required Irrelevant details -.1357 Case 5" Own personal details Permission of superior +.1283 Errors +.1357 Authority fai lure +.1529 Case 2: Confidential, not required Authority fai lure +.2139 Case 4: Non confidential , required Authority fai lure +.1379 Case 6: Subordinates' personal details D i f f i cu l t procedures +.1496 Authority fai lure +.2086 - 133 -Case 7: Other managers' details None c. Authority Case 3: Non confidential information, required -.1808 Case 6: Subordinates' details -.1398 Case 7: Other managers' details +.1203 d. Further results For non confidential information that is not required by managers for their jobs, low access is s ignif icant ly associated with high routineness of technology. Discussion: Overall the evidence supports the seventh hypothesis. Access to non confidential information is s l ight ly improved where i t is needed for the job. There is a suggestion in the results that the problems of information overload have been resolved (Ackoff, 1967). Access is made more d i f f i c u l t , though, where the non confidential information is not needed. For confidential information the picture is not so clear. Where the information is not required for the job then there is some increase in authority fai lure but a possible increase in overall access. If the information is required then the managers performing more routine tasks seem to face higher barriers to retrieving and using i t . For those man-agers who perform routine tasks i t is conceivable that their authority to access confidential information might be questioned by others even i f the information is required. This also reflects the greater care that appears to be taken with handling confidential information. A similar pattern is found for personal information (cases 5 and 6) where even greater procedural barriers are found for those performing more routine tasks. - 134 -6. Attitudes to data-sharing (i) Norms a. Access Case 1: Confidential information, Case 2: requi red Access +.2018 Retrieval -.2074 Use -.1865 Case 3: Non confidential , required Case 4: Retrieval -.1492 Use -.1405 Case 5: Own personal details Case 6: Access +.1902 Retrieval -.1239 Use -.1302 Confidential information, not required Access +.2863 Retrieval -.1693 Use -.2011 Non confidential , not required None Subordinates1 detaiIs Access Retrieval Use personal +.2915 -.1602 -.1694 Case 7: Other managers' details None b. Specific barriers Case 1: Confidential information, requi red D i f f i cu l t procedures -.2256 Embarrassment -.1421 Permission of superior-.2465 ! Errors -.2555 Missing details -.1770 Layout -.1467 Jargon -.2265 Authority fa i lure -.2278 Timing -.1531 Case 2: Confidential, not required. D i f f i cu l t procedures Errors Missing details Jargon Authority fai lure Irrelevant details .1794 -.2404 -.2012 -.2482 -.1807 -.1983 not Case 3: Non confidential , required Case 4: Non confidential requi red Embarrassment -.1735 Permission of superior-.1626 Embarrassment -.1764 Permission of superior-.2153 Location -.1740 Layout -.1610 Jargon -.1961 Authority fai lure -.1509 Widely known -.1312 Case 5: Own personal details Missing details -.1749 Layout. -.1880 Authority fai lure -.1851 Timing -.2208 Case 6: Subordinates' personal details D i f f i cu l t procedures Permission of superior Errors Comparison Missing details Location Layout -.1975 -.1902 -.1961 -.1575 -.2343 -.1300 -.1964 - 135 -Case 7: Other managers' details Errors -.2660 Case 6: Subordinates personal details (cont'd) Authority fai lure -.2837 Timing -.2400 Irrelevant details -.1326 c. Authority Case 1 Case 4 Case 5 Case 6 Case 7 Confidential information, required +.1344 Non confidential information, not required +.1491 Own personal details +.1435 Subordinates' personal details +.2492 Others managers' personal details -.2556 d. Further results Those managers who had good access to confidential information not required for their jobs tended to be from departments where the norms of data sharing were high. ( i i ) Trust and openness a. Access Case 1: Confidential information, Case 2: Confidential, not required required Access +.1450(p=.05) Access +.2378 Use -.2789 Retrieval -.2410 Use -.3066 Non confidential , required Access +.2004 Retrieval -.2774 Use -.2406 Case 4: Non confidential , not requi red Access +.1833 Retrieval -.2612 Use -.2083 Case 5: Own personal details Access +.2840 Retrieval -.2821 Use -.2835 Case 6: Subordinates' details Access +.3127 Retrieval -.2988 Use -.3164 Case 7: Other managers' details Access n.s. Retrieval -.2966 Use -.2412 - 136 -b. Specific barriers Case 1: Confidential information, required D i f f i cu l t procedures -.1631 Embarrassment -.2348 Permission of superior-.1584 Errors -.3284 Comparison -.2747 Missing details -.1658 Layout -.3073 Jargon -.2839 Authority fai lure -.3191 Timing -.1598 , Irrelvant details -.2686 Cost -.1399 Case 3: Non confidential , required D i f f i cu l t procedures -.2236 Embarrassment -.2057 Errors -.2469 Comparison -.2563 Location -.2054 Widely known -.1322 Missing details -.2282 Layout -.1765 Jargon -.1801 Authority fai lure -.2588 Timing -.1785 Irrelevant details -.1905 Cost -.1423 Case 5: Own personal details D i f f i cu l t procedure -.2489 Embarrassment -.1574 Permission of superior -. 1452 Errors -.2965 Comparison -.3465 Location -.1236 Missing details -.2658 Layout -.2172 Jargon -.2762 Authority fai lure -.2580 Timing -.2282 Irrelevant details -.1682 Case 7: Other managers' details Errors -.3419 Comparison -.3442 Existence not known -.2277 Missing details -.2399 Layout -.1891 Jargon -.2568 Case 2: Confidential, not required Errors -.2738 Comparison -.2926 Layout -.2138 Jargon -.2541 Timing -.2225 Irrelevant details -.2865 Case 4: Non confidential , not requi red D i f f i cu l t procedures -.2048 Embarrassment -.1976 Permission of superior-.1407 Errors -.2100 Comparison -.2511 Widely known -.1770 Missing details -.2056 Jargon -.1587 Authority fai lure -.1666 Timing -.1699 Irrelevant details -.1627 Cost -.1720 Case 6: Subordinates' personal details D i f f i cu l t procedures -.3000 Embarrassment -.1415 Permission of superior-.1806 Errors -.2977 Comparison -.3656 Location -.1453 Missing details -.2652 Layout -.2355 Jargon -.2638 Authority fai lure -.3406 Timing -.2319 Irrelevant details -.1864 - 137 -Case 7: Other managers' details (cont'd) Authority fai lure -.4031 Timing -.2082 c. Authority Case 1: Confidential information required Case 5: Own personal details Case 6: Subordinates' personal details Case 7: Other managers' details + .1618 +.1739 +.1674 -.1630 d. Further results Managers who had good access to confidential information when i t was not required came from departments with s ignif icant ly higher trust and open-ness than those who could not access this information at a l l . Discussion: Generally, hypotheses eight and nine are substantially supported. Al l but two of the relationships are in the hypothesized direc-t ion. In the case of access, both the norms of data-sharing and an attitude of trust and openness are influential in improving the manager's view of access. There is also some support for the assertion that when norms of data-sharing are considered alone, the effect is confined to job-related (needed) information but even where the information is not needed for the job there is some lowering of barriers. For the second variable (trust and openness), the effect of increased access is even greater and is ex-tended to a l l types of information, even to the access of other managers' details which was hardly related to increased norms of data-sharing. For a l l but information about other managers the barriers affected cover both retrieval problems and d i f f i cu l t i es in using information, strongly supporting the work of other writers (Zand, 1972; O'Reil ly and Roberts, 1976; Dewhirst, 1971). Although i t is not possible to te l l from this - 138 -study the direction of causation, i t is expected that the attitude of trust and openness and good access are mutually supportive. There was also a considerable impact upon the authority measures for both norms and trust and openness. For a l l but information about other managers, the direction was as hypothesized. A better attitude to data sharing increased the managers' perceptions of how much authority they had over the majority of information types. In the case of other managers' information (Case 7) i t is interesting to note that authority is reduced when attitudes are better. This may result from the strengthen-ing of privacy rights which result from higher interpersonal trust. Man-agers who are more open and trustful are conscious of the need to protect the privacy of other managers' detai ls . However, the actual access to this information is somewhat improved in spite of the apparent reduction in authority. During the interviews the topic of attitudes to data sharing f re -quently arose. Eighteen of the managers indicated that their own pre-ferences and attitudes and those of top management had a signif icant i n f l u -ence on the policy and practice of access to information. These pre-ferences for certain procedures and practices formed part of management's "style" of information although management "style" in a general sense is a much wider topic than the subject of access to information (Nord, 1972: 538-544; Learned and Sproat, 1966:61ff). Eleven subjects mentioned the importance of the company executives in the determination of access patterns in their company. Five chief executive officers were perceived as encouraging an open attitude to infor -mation sharing at the level of middle management and above. This a t t i -tude manifested i t s e l f through the encouragement of development courses - 139 -for managers, open formal and informal meetings for managers and execu-t i v e s , and through giving more autonomy to managers, a l l of which were cited by subjects. This open style was used by one manager to explain why security in the company was not high. In f a c t , although a rule prohibited access to some confidential information, i t could be obtained informally with l i t t l e d i f f i c u l t y . Again, this practice seemed to result from the open style of top management. In four other companies the executive style was to dis-courage open access to information. In one of these companies the execu-tives were perceived by managers as promoting an atmosphere of secrecy, which resulted in t i g h t l y controlled access to information. This closed s t y l e of management was demonstrated i n the r e s t r i c t i o n s placed on unionized o f f i c e workers, who were kept s t r i c t l y to information required for the i r tasks even i f the information was not confidential. This stemmed partly from management's experience with the union. A copy of the month-end statements containing details of payments to non-union personnel found i t s way from the unionized typists to the union leaders <and was subsequently used in negotiations with management. The company, no longer able to trust union employees to type sensitive information with-out leaking i t , began using non-union s t a f f for such work. Other managers commented on the changes in access "s t y l e " which occurred when top management changed. In a l l these cases the old style of management was associated with secrecy and s t r i c t control of informa-t i o n . The new style of top management provided middle managers with greater authority and more opportunities to participate in decision-making and, consequently, more open access to information. - 140 -Five of the managers interviewed claimed that they also had a style of management that influenced access to information within their depart-ments. One manager asserted that he was very trustful of his subordinates. He would leave confidential information on his desk and would often inform his subordinates of details that, s t r i c t l y speaking, they did not need to know. Another manager encouraged a "family" type atmosphere in his department where information was shared freely with subordinates. A man-ager in another department claimed that the only reason she received some confidential information was her head of department's open style of manage-ment. Several other similar examples were given to the researchers. 7. Technology of access An analysis of the use of computers by management showed that over f i f t y percent of managers used computer technology either not at al l or only very occasionally. The median value for the percent frequency of use of computers was less than 10%. It would appear that a substantial pro-portion of managers are not affected direct ly by computers, a result which supports a much earl ier finding by Brady (1967). It also shows that the prediction given by early researchers of how computers were going to have an impact upon middle management's decision-making were over stated, certainly for this sample (Brady, 1967; Simon, 1965). This lack of com-puter use by management was also apparent in the interview sample. How-ever, the distribution of responses for computer use was suff ic ient ly broad to enable the tests to be made. The findings are presented below. Computer Use a. Access Case 2: Confidential information, not required - Retrieval +.1990 - 141 -b. Specific barriers Case 1: Confidential 1 information, Case 2: Confidential, not required required Permission of superior +.1900 Cost +.1274 Cost +.1531(p=.053) Existence not known +.1753 Case 3: Non confidential , required Case 4: Non confidential , not Cost +.1183 ,(p=.053) required Comparison +.1372 Embarrassment +.1550 Permission of superior +.1712 Case 5: Own personal details Case 6: Subordinates' personal Layout +.1614 details Layout +.1727 Case 7: Other managers' details Layout +.2388 c. Authority Case 7: Other managers' job status +.1389 Case 7: Other managers' biographical details +.1523 d. Further evidence An ANOVA was run for high and low frequency of computer use and high and low routineness of technology. There were no signif icant interaction effects. There was a signif icant increase in their authority to access non confidential information for those who used manual technology to access work-related information a great deal. The specif ic barriers that were affected by telephone use are presented below. Telephone Use Case 1: Confidential information, Case 2: Confidential, not required required None Errors -.1294 Embarrassment -.1568 Existence not known -.1542 Irrelevant details -.1547 Case 4: Non confidential , not required Case 3: Non confidential , required Permission of superior -.1783 Permission of superior-.1637 Authority fai lure -.1917 Authority fa i lure -.1583 Jargon -.1405 - 142 -Case 5: Own personal details None Case 7: Other managers' details D i f f i cu l t procedures +.3539 Embarrassment +.1974 Permission of superior +.2654 Jargon +.2018 Cost +.2212 Location +.2844 Retrieval +.3473 Final ly , the use of face-to-face contact (no technology) for retrieving information had almost no signif icant impact upon either authority or the barriers to access. Discussion: There was no support for the hypothesis that those managers who used computer technology in performing routine tasks faced lower barriers to access. In fact, higher barriers were experienced by managers who used computers frequently. This may just indicate the higher expectations of improved performance on the part of managers who use com-puters. As a contrast, in the interview sample there was clear evidence (13 cases) of improved performance available using computers, although this was generally for operational level data. In one f irm, the use of an online/realtime computer system to process bookings and inquiries had re-sulted in measurable improvements in service. In another company, on-l ine access to customer f i les meant that a branch off ice in a distant part of the province had as convenient access to the data f i l es as the head of f i ce . If access can be improved directly by the use of computer tech-nology, there were seven examples given of where the lack of an appro-priate technology hindered access. In a health care organization, where the access technology was largely manual, i t was impossible for managers to know at any one time the level of inventory. This could clearly have Case 6: Subordinates personal details Authority fai lure -.1391 - 143 -been improved by the use of a more appropriate technology to store and access this information. A very similar situation occurred in another organ-ization where branch offices would not te l l what was stored at headquarters and vice versa. This meant that large volumes of inquiries were sent by mail and resulted in delays and frustrations. For hypothesis eleven there was good support. The "costs" of access-ing information not required for the job (cases 2 and 4) were signif icant ly higher for those who used computers extensively. In particular, i t was the barriers to retrieval that were increased, as hypothesized. This was, however, no support for the assertion that managers' authority to these types of information would be reduced. The interview sample generally supports the finding of increased "costs" for computer users to informa-tion that is not required for the job. Seven of the companies visited used computer sytems that had bui lt-in f a c i l i t i e s for restr ict ing certain f i l es and programs from unauthorized use. These systems use special terminals, passwords, and levels of authority in various combinations to restr ic t access. The companies were exploiting the new capability by using the systems to keep employees to what information they "need to know". Indeed, for two of the firms, the introduction of the capability had precipitated the discussion among managers of what rules of access should be implemented. Each case demonstrated the influence of the new technology in being able to support and enforce predetermined access patterns. As one manager noted, computers can be very effective in seal -ing off access to information from those who do not need i t . The evidence from the interviews concerning the effect of computer systems on access to information for most workers was clearly to restr ict them to what is needed for the job. However, the evidence is ambiguous - 144 -for managers. F irst of a l l , many of the managers interviewed were not users of computer systems. Secondly, where they used computer systems some found improved access to al l information while others found reduced access. As examples, one manager indicated that computer access had reduced his ab i l i t y to retrieve information he could once get informally. In direct contrast, another manager found that his informal access to information had improved with computerization, largely as a result of the easy avai labi l i ty of passwords to him. The company did not enforce the access rules for people at his level whereas access patterns for administrative workers were s t r i c t l y adhered to. Some of the data also indicate findings beyond those hypothesized. Cost ($) is positively related to increased use of computer access for work-related data. This suggests that managers who use computers frequently become conscious that the information produced is costly. Indeed, i f the interview sample is taken as representative then many of the managers operate in environments where the cost of computerized information is charged back to the department. To i l lustrate th is , in one of the companies vis i ted the data processing department tr ied to make managers aware of the cost of printouts by printing the cost in dollars in a prominent position on the front page of each report. Those managers who frequently used the telephone to access work related information indicated that their barriers of access to work-related information were s ignif icant ly reduced. At the same time, their barriers of access (particularly retrieval) to other managers' personal details were much greater. Spec i f ica l ly , for non confidential information the results seem to signify that the use of the telephone is effective in bypassing some of the more formal channels to this information. - 145 -8 . a . M a t u r i t y o f i n f o r m a t i o n s y s t e m A c c e s s Case 3 : Non c o n f i d e n t i a l i n f o r m a t i o n , r e q u i r e d + .1646 Case 4 : Non c o n f i d e n t i a l i n f o r m a t i o n , n o t r e q u i r e d + .1663 Case 5: Own p e r s o n a l d e t a i l s + .1619 Case 6: S u b o r d i n a t e s ' p e r s o n a l d e t a i l s + .1797 Case 1: C o n f i d e n t i a l i n f o r m a t i o n , r e q u i r e d - Use _ .1406 Case 3 : Non c o n f i d e n t i a l i n f o r m a t i o n , r e q u i r e d - Use - . 1278 Case 5 : Own p e r s o n a l d e t a i l s - R e t r i e v a l .1436 Case 6 : S u b o r d i n a t e s ' p e r s o n a l d e t a i l s - R e t r i e v a l - .1625 b . S p e c i f i c b a r r i e r s Case 1: C o n f i d e n t i a l i n f o r m a t i o n r e q u i r e d C o m p a r i s o n - . 1 7 0 4 Embar rassment - . 1 4 1 9 P r e s e n t a t i o n p r o b l e m s - . 1 4 3 8 Case 3 : Non c o n f i d e n t i a l , r e q u i r e d E r r o r s - . 1 4 4 4 L o c a t i o n - . 1 3 0 0 B i a s - . 1 6 1 0 Case 5: Own p e r s o n a l d e t a i l s C o s t L o c a t i o n E x i s t e n c e no t known T i m i n g +.1240 (p=.054) C o s t Case 2 : C o n f i d e n t i a l , n o t r e q u i r e d C o m p a r i s o n - . 2 8 2 5 Case 4 : Non c o n f i d e n t i a l , n o t r e q u i r e d E r r o r s - . 1 5 9 6 B i a s - . 1 5 0 6 Case 6 : S u b o r d i n a t e s p e r s o n a l d e t a i l s - . 1 3 8 3 - . 1 5 9 3 - . 1 3 4 9 Embar ras smen t E x i s t e n c e n o t known B i a s Case 7 : O t h e r manage r s ' d e t a i l s C o s t +.2404 + .1269(p= .052) - . 1 4 7 7 - . 1 9 5 4 - . 1 3 8 6 c . A u t h o r i t y Case 7 Case 7 Case 6 O t h e r manage r s ' j o b s t a t u s ( s a l a r y ) O t h e r manage r s ' g e n e r a l i n f o r m a t i o n S u b o r d i n a t e s ' g e n e r a l i n f o r m a t i o n F u r t h e r r e s u l t s - . 1 7 2 0 - . 1 6 4 1 + .1413 An AN0VA was run u s i n g h i g h and low m a t u r i t y and h i g h and low com-p u t e r use b u t no s i g n i f i c a n t i n t e r a c t i o n e f f e c t s were f o u n d . The ma in e f f e c t f o r m a t u r i t y s u p p o r t e d one o f t he above r e s u l t s t h a t h i g h m a t u r i t y i s a s s o c i a t e d w i t h l o w e r r e t r i e v a l p rob lems f o r t h e m a n a g e r ' s own d e t a i l s - 146 -(case 5). Discussion There was no support for the interaction effect between the maturity of the computer system and the frequency of computer use of the managers. In general, however, the results do indicate good support for the asser-tion that more mature information systems do provide better managerial access to information. Access ab i l i ty was s ignif icant ly higher for four types of information (cases 3, 4, 5, and 6). Better retrieval was provided to personal information (cases 5 and 6) and better use of information was available where i t is needed for the job (cases 1 and 3). For authority there appears to be higher ratings over information about subordinates and lower ratings for information concerning other managers in companies with more mature information systems. This signif ies a growing formalization in companies with more mature information systems. Access to personal information about employees is authorized i f you need i t (case 6) but is not authorized i f i t is not required (case 7). An analysis of the specif ic barriers shows that for work-related information, mature systems are associated with lower barriers to access. Furthermore, most of the lower barriers are concerned with the ease of processing the information (comparison with other information, errors, bias and presentation problems) which is only to be expected in systems where most of the computer problems have been solved. This supports the current wisdom in the MIS implementation l iterature (Lucas, 1978; Bjorn-Anderson and Hedberg, 1977). F inal ly , there is some evidence of increased cost awareness with more mature information systems. The relationships are nearly s ignif icant in two cases. When similar analyses of the results are conducted for the - 147 -age of the information system, the cost barriers do become signif icant. In the case of non-confidential information, not required for the job, cost of information also begins to show some importance (x = + . 1 2 6 7 ; p = . 0 5 9 ) . It is to be expected that the age of the system is a better indicator of the chargeback of information costs rather than the maturity of the system. This would then be in accord with those who claim that chargeback systems are a feature of the later stages in computer systems development (Gibson and Nolan, 1 9 7 4 ; Nolan, 1 9 7 5 ; Nolan, 1 9 7 9 ) . 9 . Departmental a f f i l i a t ion The data processing managers were selected as a group (n = 14) and compared with the remainder of the sample using a Mann-Whitney U test (1-tai led) . The aggregate measures for Retreival and Use were used as i n -dependent variables in the test. The analysis revealed no signif icant findings in the hypothesized direction that managers would have lower costs of access to information. In fact , in using personal information about themselves or their sub-ordinates, they faced signif icant ly higher barriers. This may be an ind i -cation of the problems of using information over which the data processing managers have no control as i t is usually kept in a centralized personnel department. In contrast to the above f inding, a l l of the data processing man-agers interviewed generally had more extensive access to more information than most other managers at a comparable leve l . Although they did not normally have greater authority over a l l of the computer-stored informa-t ion, in practice their access and that of some other employees in the data processing department was unrestricted. Of course, much of the data - 148 -that passes through the d.p. department is of no interest to its employees. However, one would have to conclude from the interviews that d.p. managers, as information experts in the organizations, have more privileged access to information than comparable managers from other departments. 10. Experience (i) General experience (age) a. Access Case 5: Own personal details Case 5: Own personal details b. Specific barriers Case 1: Confidential information, required None Case 3: Non confidential , required Layout -.1310 Errors -.-175 Case 5: Own personal details D i f f i cu l t procedures -.1211 Permission of superior -.1525 Failure of authority -.1356 Embarrassment -.1760 Location -.1321 Timing -.1708 Case 7: Other managers' details None +.1393 Retrieval -.2071 Case 2: Confidential, not required None Case 4: Non confidential , not required None Case 6: Subordinates' personal details Embarrassment -.1420 Location -.1224 Authori ty Case 4 Case 5 Case 7 Non confidential , not required +.1420 Own personal details +.1530 Other managers' biographical details -.1365 ( i i ) Specific Experience a. Access Case 2: Confidential information, not required - Use -.1416 b. Specific barriers Case 1: Confidential information, Case 2: Confidential, not required required - 149 -Location -.1527 Irrelevant details -.2256 Timing -.1527 Case 3: Non confidential , required Case 4: Non confidential , not Location -.1322 required Jargon , -.1575 Jargon -.1705 Case 5: Own personal details Case 6: Subordinates' personal details Embarrassment -.1487 None Timing -.1387 Case 7: Other managers' details Missing details -.2057 c. Authority No signif icant findings. Discussion: In general, hypothesis fourteen is supported. Managers with greater experience (general and job specific) face lower "costs" of access to most types of information. For each result the direction of the relationship is as hypothesized. For increased general experience, the barriers to access personal information were substantially reduced. In part icular, the barriers to retrieving personal details (cases 5 and 6) were signif icant ly lower for those with more experience. This may be explained by the relat ively stable nature of the procedures which are used to access this information. In contrast, access to job-related information is more dependent on job-related experience. Experienced managers also had more authority of access over their own records but less authority over other managers' details suggesting that general experience demonstrates to managers that other managers' personal details are very sensitive in most organizations. Gen-eral experience also gives the manager some advantage over others in the processing of non confidential information required for the job (case 3). Both layout problems and errors were signif icant ly reduced with general - 150 -experience for this type of information. In the case of job-specific experience there are again reductions in access barriers to personal information. More s igni f icant ly , however, job specif ic experience reduces some of the barriers to job-related informa-tion. In particular,problems of the location of required information are s ignif icant ly reduced with on-the-job experience. It was previously shown that good location was very important for the accessibi l i ty of informa-tion (Appendix IV). Also, for non confidential information there are signif icant reductions in the number of problems with jargon for this kind of experience. Both of these points support the notion of high start-up "costs" of access in any new job (Arrow, 1974; Mechanic, 1962). Those who have had substantial job-specific experience have largely overcome them. 11. Other determinants In addition to those mentioned above the interviews suggested that other factors were influential in determining the direct and indirect regu-lation of access to information. These factors and their influence on access are summarized below. (i) Competitive environment Evidence gathered from the interviews seems to indicate that the com-petit ive environment of a company has a systematic impact upon access policies and practices. Generally, a company operating in a highly com-petitive market has more internal regulation of access to information. Such companies tend to keep managers to information that is needed for their jobs. In three of the companies where the markets were highly competitive, information access was handled carefully and throughout these companies - 151 -managers were supplied information on a "need to know" basis. In a l l other companies v i s i t e d , only certain areas of the business used information valuable to competitors and applied s t r i c t rules of access r e s t r i c t i n g i t s use. For instance, in a large crown corporation v i s i t e d certain costing information was very valuable to commercial competitors. This information was treated with greater security i n t e r n a l l y . Access was given to man-agers on a "need to know" basis. However, this rule did not apply with the same strictness to other internal information. In another company the divisions faced different levels of competition in the market place. For a highly competitive part of the business this meant that pricing information was very valuable and was treated accordingly. The company applied a level of regulation that was appropriate to each market. In a regulated monopoly the opposite situation was found for information about new product development. Whereas in a competitive market access to.infor-mation about new products would be kept secure u n t i l the l a s t moment, when there was no competition, the information was freely available both i n -t e r n a l l y and externally. In another regulated industry, where, u n t i l recently, prices f o r services had been fixed and competition between com-panies had been conducted largely on the basis of service, the new price system was expected by several managers to produce a more careful attitude towards information access i n t e r n a l l y . Another market factor influencing the internal access to information i s the type of business being conducted. F i r s t l y , the nature of the business influences the time period for which information has to be kept confidential i n t e r n a l l y . It also determines the level of p r o f i t margins available to the company. In the real estate industry the time span for confidential information i s often quite small as contracts can be signed - 152 -quickly. Additionally, the scope for huge profits for a l l the firms in the same real estate market creates l i t t l e incentive to poach confidential information. Problems with unauthorized access were not found in the firm vis i ted. Internally, managers had good access to valuable information even i f i t was not s t r i c t l y required for their jobs. As a contrast, in the food industry where competition is f ierce and profit margins are low, and the time span for confidential information is longer, security was found to be high internally. Managers were often restricted to informa-tion required for their jobs. Several cases of industrial espionage were reported in the company suggesting that s t r ic ter regulation of access may be just i f ied in such companies. ( i i ) External constraints Several companies visited operated with certain informational con-straints imposed upon them from outside. These constraints were reflected in the internal policy and practice of access to information. The five organizations that come under the Public Bodies Act are required by law to publish annual financial figures on expenses and sa l -aries paid to employees. This means that salary information, at least retrospectively, was available internally to those organizations, in direct contrast to commercial companies where salary details of other managers were kept from their peers. However, access to the document on salaries and expenses was not promoted by any of the organizations concerned and in practice an interested manager had to overcome considerable barriers to obtaining the information. Similarly, a regulated monopoly visited is required by law to publish its operating procedures. Although the public as well as the government is supposed to be given access to the document, - 153 -in practice the public must v i s i t the offices of the company to see them. These were examples of indirect regulation of access, or regulation by inconvenience. In another regulated monopoly, projects conducted by the company tended to be long-term ones because of the nature of the business. During the l ifetime of such projects demands are made upon the company by various external groups seeking what the company considers to be conf i -dential information. These demands, coupled with the complex nature of the work, produce a large amount of documentation for each project result-ing in some of this information being leaked to unauthorized persons. External regulations can also act as constraints upon commercial companies in Canada. For example, recent provincial regulations require companies to practice careful , progressive grievance procedures when dealing with their employees. In the view of one manager this has forced companies to become bureaucracies. Not only must records be kept but they must be made available to those who need them when a dispute arises. ( i i i ) Industry standards In four of the companies in which managers were interviewed, the internal handling of information was influenced by standards that prevailed in the industry. For example, in a branch of the transport industry there has been a long tradition of sharing information of a technical nature and information concerning the market place. However, the market-ing information fa l l s short of providing detailed information concerning routes as this is considered highly confidential to the companies. None-theless, the industry standard is clearly reflected in the way that access to these types of information was relat ively easy for managers internally. Two examples were found showing that industrial ethics can also i n -fluence the handling of confidential information. In one case a manager - 154 -was offered confidential promotional information about the major competitor but he would not take advantage of this access. It is a tradition among advertising personnel to refuse i l l i c i t access to advertising information. (iv) Union agreements In most cases examined, the collective agreement between the union and the company covered the general area of wages and benefits. However, for two organizations, the agreement also covered access to information. This clearly acted as a constraint upon the personnel and payroll systems. For instance, one agreement requires that grievance material be discarded after one year i f no further incidents arise. It also restr icts the amount of information that may be collected and prohibits the keeping of informal personnel records. Employees' rights of access to the f i l es are also written into the agreement. Interestingly, i t seemed that the rules of access for non-unionized management were identical to those used by the unionized employees. In companies where employees were covered by a col lective agreement, the details of the agreement were often available only to those who needed them and access was sometimes restricted for some managers. For example, in one company the published agreement could be obtained i f you were a member of the union but the payroll department would not release the details in the agreement to most managers. (v) Specific incidents Cyert and March (1963:48) documented a signif icant incident in a factory which led to changes in equipment and procedures in the work place. In this study ten signif icant incidents were recorded where infor-mation was leaked internally and externally causing embarrassment or economic loss to the company involved. These incidents led to changes in - 155 -the pol ic ies , practices, or both, of providing access to information. The changes recorded were always in the direction of greater control. In eight of the cases the leak involved confidential information and the recipient was a major competitor. The companies discovered the incident had taken place because the competition was using the leaked in -formation to their advantage. In one company, the advertising department was losing price information to a competitor before the o f f i c i a l day of release. This was apparent in the competitor's advertising. To combat th i s , the company tried several changes in procedure. F i r s t l y , false data were released occasionally but this proved ineffective. Secondly, the re-lease of the confidential information was only allowed at the last moment. Lastly, the physical area of the department was restricted to authorized employees only. In the same company, the research and development group was moved out of the main building after research details were leaked and f i l i n g cabinets were burgled. Internally, the leaking of sensitive information to employees can be an embarrassment to a company. A copy of a report on cost savings including planned redundancies was obtained by one of the off ice workers in a company (cf. Mechanic, 1962). This resulted in the company being picketed by the workers in protest against the recommendations of the report. This and other incidents led to a much more careful attitude on the part of management in providing of access to information for off ice workers. For example, the company began using non-union employees for typing sensitive or confidential reports in an effort to tighten the con-trol of access to this information. Several companies employ one person or a group of people especially to provide an outside channel for o f f i c i a l statements from the company. - 156 -The establishment of two of these positions was in direct response to problems experienced after employees had given confl ict ing statements to the press or the public. In such situations, i t is also necessary for the top management to ensure that al l employees are told about the new policy for i t to have any chance of being effect ive. The use of the latest computer technology to store and access infor-mation had caused five of the companies to review their policy on access. In each case the new systems triggered a more formal approach to regulating access to information as the possibi l i ty of enforcing predetermined access patterns became a real i ty . In one company the policy of enforcing access on a need to know basis is being implemented even though with the old system additional information was available to employees informally. In another company the computerization of personnel records forced the man-agement to state a formal access policy for this information and a manage-ment committee was established for this task. Summary 1. The results of testing specif ic hypotheses are summarized in table VI. 2. Further determinants of access were suggested by the interviews: (i) Competitive Environment ( i i ) External Constraints ( i i i ) Industry Standards (iv) Union Agreements (v) Specific Incidents - 157 -V. Fac i l i t ies and Strategies to Promote Access to Information The emphasis so far has been on the impediments to the access of information. It is of interest, however, to identify f a c i l i t i e s and strategies used by companies and individuals to promote access. For this purpose, questions were included in the questionnaire and evidence was gathered through the interviews. The results are presented below. 1. Organizational promotion of Access The responses made to the questions about the f a c i l i t i e s to promote access (section III of the questionnaire) were used in two ways. F i r s t l y , the number of f a c i l i t i e s was summed for each subject giving the values 0-5, and six or more. This number was used to investigate the relat ion-ships with the independent variables and with the questions on authority and access. Secondly, the analysis was repeated for each of the effect ive-ness measures of the f a c i l i t i e s used to promote access. (i) Independent variables a. Number of f a c i l i t i e s used to promote access (median = 3.792) Size of company +.2828 Size of department +.2186 Number of levels . +.2332 Age of information system +.2110 b. Effectiveness of each f a c i l i t y Newsletters (n = 134, median = 4.621) Staff meetings Cn = 152, median=5.547) Norms of data sharing +.1762 Norms of data sharing +.1988 Age of information system +.1803 Routineness -.2308 Routineness -.1263(p=.050) Trust and openness +.1946 Trust and openness +.1628 Management meetings (n = 161, Technical advisory (n = 100, median = 5.600) median = 5.350) Trust and openness +.1782 Trust and openness +.2196 % use of computers +.1837 Liaison personnel (n = 67, median= 4.765) Trust and openness +.2437 - 158 -Discussion: The number of f a c i l i t i e s is s ignif icant ly related to three measures of size for the organizations. This suggests that larger organizations can provide more f a c i l i t i e s because of the greater resources they have (money and manpower) and because they can exploit the economies of scale that would not be available to smaller firms. Furthermore, in larger organizations there is evidence that indicates a greater need for these f a c i l i t i e s . F i r s t l y , in larger organizations, the barriers to access are somewhat greater (see results in section IV). Secondly, other researchers have found that communication is positively related to the number of employees (Bacharach and Aiken, 1977). The increased use of f a c i l i t i e s to encourage access would be consistent with both these findings. The age of the formal information system is also positively re-lated to the number of f a c i l i t i e s . This may be a result of the increased demand for f a c i l i t i e s to promote the use of computerized systems. In three of the companies v is i ted , internal technical advisory services were provided to help managers develop sk i l l s to improve their access to com-puterized information. One of these companies had organized a special information centre. The centre trains managers in simple programming languages that enable them to write quick programs to extract special reports from the data f i l e s . The effect of this and similar f a c i l i t i e s in other companies is also supported by a strong positive relation between the effectiveness of technical advisory f ac i l i t i e s and the percentage use of computer f a c i l i t i e s . By analysing the results for the effectiveness of each f a c i l i t y , several consistent relationships can be seen. F i r s t l y , trust and openness is positively related to the effectiveness of every f a c i l i t y . Of course, - 159 -there is no suggestion of causality here and in fact trust and effect ive-ness may be related by mutual causality. Secondly, norms of data-sharing is positively related to the effectiveness of newsletters and staff meet-ings. Again, the same argument relating trust and openness to effect ive-ness may apply here. F inal ly , those whose jobs are more routine generally see less effectiveness in two of the f a c i l i t i e s to promote access to infor-mation. A possible explanation for this result is that managers whose information requirements are reasonably predictable or stable do not value f ac i l i t i e s that promote access to information (Comstock and Scott, 1977:171). ( i i ) Authority and access a. Number of f a c i l i t i e s used to promote access Authority to access own personal details +.1724 b. Effectiveness of f a c i l i t i e s Management meetings Authority to access confidential information, not required +.1664 Access to confidential information, required +.2050 Access to confidential information, not required +.1827 Access to non confidential information, not required +.1924 Technical advisory Access to confidential information, required +.1946 Access to confidential information, not required +.1872 Access to own personal details +.1849 Discussion: In those companies where there are more f a c i l i t i e s to promote access to information" there tends to be higher authority given to managers to access their own records. This has at least two possible explanations. F i r s t l y , the more f a c i l i t i e s used by an organization to promote access could indicate a desire to improve the attitudes to data sharing. This is supported by the previous finding that trust and open-- 160 -ness is positively related to the authority to access the manager's own records. Secondly, because size and the number of f a c i l i t i e s are strongly positively related, the increase in authority might be due to the size of the company and the accompanying formalization of rights of access for employees. However, size was not found to be related to the authority of managers to access their own records (section IV). Thus, the overall evidence seems to favour the f i r s t argument. The other main findings show a general increase in access to con-f idential information (required or not for the managers' jobs) is positively related to the effectiveness of management meetings and tech-nical advisory services. There is also an increased authority to access confidential information not required for the job where the effectiveness of management meetings is judged to be high. Again, the provision of effective management meetings and technical advisory services indicate a genuine objective of executives to provide a better "atmosphere" for information access. This is reflected in the kind of good access that is available to managers to confidential information even i f i t is not required for their jobs. This is supported by the,strong relationships reported previously between trust and openness and access to confidential information. In the interview sample, executives in at least three of the firms used management meetings to promote information exchange. The president of one company meets with al l his managers three times a year. At each meeting he promises to answer al l questions, which are submitted anony-mously. The president was clearly in favour of a more open approach to information handling and this open "style" was apparent from the comments of middle managers in the firm. - 161 -Many of the companies visited used at least one f a c i l i t y to promote access to company information besides the regular information given to managers to perform their jobs. As the above example demonstrates, some of these f a c i l i t i e s and their effectiveness were a direct result of a style of top management in exercising control over information flows (Forrester, 1965). Six of the companies visited used special units to handle internal access, external access, or both. Such groups as public relat ions, cor-porate communications, and information services were found in several of the larger firms. One of the functions of these groups is to promote access to specif ic and general company information. The main tool for doing this internally was often the company newsletter or a similar pub-5 l i ca t ion . These contain such details as general performance figures and art ic les often highlighting the human interest. Another large company used a series of movies to keep managers informed. However, newsletters were often judged to be of only modest value to managers. This is re-flected in the lower general effectiveness ratings that were given to this f a c i l i t y . Four specif ic cases were mentioned in the interviews in which the company promoted access to information for motivational purposes. As one manager expressed i t , " i f top management want managers to know something, they (the managers) get i t " . In two of the companies, managers were encouraged to access relative performance figures as a means to promote competition among them. In another f irm, a manager received a special In a large communications company a special telephone number was set aside by which employees could access the latest company information. - 1 6 2 -report that others at his level did not normally have access to. The manager's access to the report seemed to signify a reward from top manage-ment. 2. Individual coping strategies The major thrust of this study has been organizational regulation of access to information. In addition, some evidence was gathered on the ways in which individual managers develop strategies to cope with the pro-blems of access they face. (i) Developing interpersonal contacts among managers Apart from the information that they receive automatically, managers are le f t largely to their own devices to uncover the system of access to information when they f i r s t join a company. Companies rarely supplied any procedural manuals to the newly-arrived managers. In the two cases where the manuals were available, the managers had not found them helpful in discovering what information was available to them. One typical manager noted that access is governed in his company by the position the manager has (his authority) and the network of con-tacts which has been developed. One short-cut in this process is for managers to seek out the opinion-leaders (Rogers and Agarwala-Rogers, 1976). Three managers, each with a long record of service for their company, reported that junior managers often used them to find out i f i n -formation exists and how i t can be retrieved. The senior managers' expertise, authority, and contacts give them a clear advantage in access-ing information compared to other managers (Mechanic, 1962; Crozier, 1964). Many of the interpersonal contacts that managers had developed were exchange relationships where information.was exchanged for other - 163 -benefits. In one company, the finance department was given extensive cooperation by another department with the collection and interpretation of information. This was done so that the second department could ensure that the information was not misinterpreted by the f i r s t department. In another s i tuat ion, one manager passed information to another manager that he thought should have been using i t . The hope of the f i r s t manager was that he would eventually receive other, useful information in return. Two cases were cited by different managers showing that access to con-f idential information held by others occasionally has to be "earned". That i s , the managers had to demonstrate that they would not abuse the information obtained before further access to that source could proceed. ( i i ) Use of subordinates In addition to developing contacts among other managers, relat ion-ships were often developed by managers with subordinates or other employees inside and outside their department. One manager reported that he could by-pass the formal barriers to detailed information by using fa i r l y low-level contacts that he had developed in another department. In two companies, both operating in competitive markets, information was obtained about competitors informally from the central distr ibutors. The salespersons from the distributors are questioned informally to ex-tract information concerning "specials" that are being planned by the competition. The following comment from one of the questionnaires exemplifies the use of subordinates to overcome access problems: "In our company we have an unofficial policy of secrecy. It is possible to obtain information from other people providing you can convince them that you need i t (peers and superiors). People do not readily volunteer infor-mation. Superiors in possession of information vital to my own job do not, as a matter of course, pass i t - 164 -along unless I speci f ica l ly become aware that they have i t and ask for i t . As a result an informal infor-mation system is in place based, in part, on personnel relationships, i .e. My friends at lower levels pass on information i f their superior (my peer) is reluctant to do so!" ( i i i ) Informal record-keeping systems As well as using informal contacts to obtain information, there were strong indications in many companies visited that managers had con-structed their own informal system of personnel record-keeping. This was found in al l companies except those using decentralized personnel record-keeping. It seems that managers seek alternative paths of access to important information when the "costs" of o f f i c i a l access become too high to bear. The most frequently mentioned causes of managers using their own " i l l i c i t " record-keeping were the inconvenient location of the personnel department, the amount of "red tape" to overcome, and the lack of relevant information in the f i l e s . It is also a good indication that i f information access alternatives are available in other, work-related areas, the managers might choose to develop their own systems with access tailored to their needs. (iv) Information f i l t r a t ion The major complaints from managers about using the information once i t has been retrieved were about the high volume of the reports produced by computers (three examples) and the problems associated with interpret-ing them (two examples). One manager, rarely used the large computer reports that were sent to him. He would examine certain key figures to see i f any further investigation were necessary but otherwise he would f i l e them without looking further. In his case the reports were largely peripheral to his tasks and so his solution was to ignore them and not - 165 -process the bulk of material they contained (cf. Churchman and Schainblatt, 1965; Simon et a l . , 1954). Summary The results presented in section V can be summarized as follows: 1. Trust and openness is positively related to the effect ive-ness of every f ac i l i t y to promote access. 2. Managers performing more routine tasks believe that the effectiveness of newsletters and staff meetings is lower than managers who perform less routine tasks. 3. The size of the organization/department is positively related to the number of f a c i l i t i e s to promote access. 4. General access to confidential information is positively related to the effectiveness of management meetings and technical advisory services. 5. Four individual coping strategies were ident i f ied: (i) Development of Interpersonal Contacts ( i i ) Use of Subordinates as Intelligence Sources ( i i i ) Use of Informal Record-keeping Systems (iv) Information F i l t ra t ion. - 166 -CHAPTER FIVE PRESCRIPTIONS FOR DESIGN AND FUTURE RESEARCH I. Designing Organizations and Information Systems for Improved Managerial Access Table VII contains a summary of design prescriptions based on the findings of the study.'' While the prescriptions may improve managerial access to information, they may also have negative side effects that are costly to the organization. It is suggested that the balance of costs and improvements depends on the organization and the portfol io of changes that is being contemplated. Specific organizational differences are not treated expl i c i t l y in the detailed discussion which follows. 1. General prescriptions Some of the findings can be used to suggest general improvements to organizations and information systems that lead to better managerial access. The prescriptions are general in the sense that they may be applied to al l types of information. Specific prescriptions for different types of information are presented under part 2. (i) Direct and indirect regulation of access Apart from personal information, there was very l i t t l e direct regula-tion of managerial access in the form of rules or policy documents. The widespread use of the rule that allows access to information on a "need to know" basis was the major exception. The prescriptions are presented as i f the directions of causality are known. The problems associated with the research methodology and the limitations of correlational studies are discussed in the next section. - 167 -Table VII Designing Organizations and Information Systems for Improved Managerial Access 1. General Prescriptions Summary Area Findings Prescriptions Direct and in -direct regula-tion of access Access, re-trieval , and use Structural impact upon access Technology of access Attitudes to data sharing Regulation of access is largely by indirect means except for "need to know" rule. Access is inversely related to the problems of retrieval and is a l -most independent of the d i f f i cu l t i es in using in -formation. Cost of information was not related to access. Access is somewhat poorer in larger organ-izations. Better access was found in organizations with more levels of authority. Policy on access is made more expl ic i t where computer retrieval is employed. In organizations where i t is normal to share in-formation or where an attitude of trust and openness is developed, managers have improved access to information Regulate access direct ly wherever possible. Make managers aware of their authorization to access i n -formation Concentrate on improving the convenience of retrieving information (location, timing, d i f f i cu l t procedures, author-ity fa i lure , permission of superior, and embarrassment). Cost of information should not be used as an instrument to deter access to informa-tion . To compensate, provide more and better f a c i l i t i e s to promote access to informa-t ion. Volumes of report outputs should be kept to the minimum in organizations with few levels of authority. Policy of access should be decided by management. Those managers affected should be involved in the policy setting where possible Use organizational develop-ment techniques to improve managers' "styles" of data sharing. To improve trust and open-ness, f a c i l i t i e s to promote access to information should - 168 -Table VII (cont'd) Area Findings Prescriptions Individual coping strategies (direct and indirect regulation). Development of informal networks among managers. Frequent use of infor-mal record-keeping systems. 2. Specific Prescriptions Confidential information Non confidential information Personal information Access to confidential information needed for the job is generally more d i f f i cu l t than access to non confiden-t ia l information needed for the job. Managers had better authorization and access to non confidential i n -formation when i t was needed for the job com-pared with when i t was not needed. Managers' access to their own f i l es and those of their subordinates is generally regulated in organizations l ike access to non confidential information. be made available to managers. Make managers aware of their authorization to access infor-mation . If they are important, try to formalize the informal channels. Decentralize the record-keeping function wherever possible. Regulate access direct ly rather than indirect ly . Attempt to match the direct and indirect regula-tion of access to confidential information by reducing the inconveniences of access for those who need confidential information for their jobs. Managers should- be authorized to access any non confidential information within their department. Access should be made more convenient when the informa-tion is needed for the job. Use several f a c i l i t i e s to promote access to this i n -formation. Access to information about subordinates and other managers should be regulated on a "need to know" basis only. Information gathering should be restricted to what is reasonably necessary for per-sonnel tasks. Location and procedural barriers should be reduced for those with legitimate needs. - 169 -Table VII (cont'd) Area Findings Prescriptions a) Managers' own details b) Subordinates' details c) Other man-agers ' details Managers had limited rights of access. Some ignorance of access rights was found. Inconvenient access to centralized records was found. Inadequate records were often maintained. This information is treated with greater care than confidential information. Generally, managers are in favour of the secrecy that surrounds salary i nformation. Managers should be allowed to see their own personal details and be given the right to rebut them. Knowledge of these rights of access should be promoted among managers. Decentralize the record-keeping function but ensure that adequate security is available for the decentra-l ized records. Conduct an information requirements analysis to determine what the managers need to perform their per-sonal functions. Generally, maintain the status quo as a formal access policy. Ensure that the grading and evaluation procedures are well known by the managers. - 170 -It is generally suggested that organizations should use more formal rules to regulate access. The use of formal rules to regulate access to information may be economical and effective especially i f they are supported by organizational sanctions. One must also recognize that regu-lating access informally is a less precise form of control as i t is often accompanied by spi l lover "costs" in other areas of access. Access may also be encouraged among managers by using expl ic i t rules of access and making them known to the managers concerned. In the interviews there was a signif icant number of managers who did not know what their rights of access were. By helping managers to understand what information they are entit led to, organizations can help solve two problems of access. F i r s t l y , the use of rules wil l help new managers to eliminate some of the start-up "costs" they face. Secondly, some of the barriers to access experienced by managers having to just i fy their legitimate claims to information wil l be reduced. As a cautionary note, i t is further suggested that the use of expl ic i t rules to prevent managerial access to information has to be applied carefully and must be varied according to the type of information. The implementation of complex rules of formal access requires the managers to be familiar with those rules and i t is also l ikely to promote a "climate" that is contrary to trust and openness. In the case of non confidential information, not required for the job, where the demand from managers is probably low, i t is suggested that direct regulation or rules of access should not be used to deter access. ( i i i ) Access, re t r ieva l , and use In the MIS l i terature there has been an emphasis on the use of in -formation and the matching of individual "styles" of processing and reports - 171 -(Ackoff, 1967; Mason and Mitroff, 1973; Sorter, 1969; Benbasat and Taylor, 1978).- However, a consistent finding of the study was that a manager's ab i l i t y to access information is governed largely by the d i f f i cu l t y of retrieving information not by the d i f f i cu l t y of using i t . Exp l i c i t l y , access ab i l i t y was found to be inversely related to the barriers to retr ieval . Therefore, in order to improve managerial access to informa-t ion, the information system designer should concentrate on making the retrieval of information more convenient (Forrester, 1965). The following barriers to retrieving information were signif icant ly and negatively related to the managers' access ab i l i t i e s : location; timing; d i f f i cu l t procedures; authority fa i lu re ; permission of superior; and em-barrassment (Appendix V). It is suggested that reducing these incon-veniences wil l bring signif icant improvements to the ab i l i ty of managers to access information. It is interesting to note that several of the barriers are in administrative areas which information systems designers have tradit ional ly avoided (authority fa i lu re , permission of superior, embarrassment). The study indicates that these administrative barriers to information access may be of greater significance than many of the barriers to using information. Access to computer-based information can be improved by the use of intermediaries ( i .e . retrieval experts) or by the provision of technical advisory services which train managers to use retrieval languages themselves (Keen and Scott Morton, 1978; Zloof and de Jong, 1977). A second point of interest to note is the lack of impor-tance of cost as a barrier to the retrieval of information.'' Although The variable measuring cost was independent of most other barriers to access (Appendix III). The highest correlation (T = .258) was with the barrier that measures comparison problems. - 172 -both high maturity of the information system and decentralization were associated with an increase in awareness of information costs, i t would appear that cost plays only a minor role in the managers' perceptions of their access ab i l i t i e s . One manager, who was charged for his reports, told the researchers " i f that is what information costs then that is what i t costs". He had discounted the amount he was charged for information. Cost was treated as a fixed component in his formula for accessing infor-mation. ( i i i ) Structural impact upon access Generally the results support the Bacharach and Aiken (1977) f ind-ings for communication. Structural variables have only a limited impact upon the access to information of managers. Additionally, as many of these variables are normally f ixed, at least in the short run, the system designer normally has no control over them. Even i f we accept that these variables are f ixed, i t is s t i l l possible to suggest ways of improving managerial access to compensate for their effects. For example, in larger organizations there are s ign i -ficant increases in some of the barriers to using information, as hypothe-sized. One way to reduce some of these barriers is to promote the access of information by the provision of faci 1 it iessuch as newsletters, meetings, technical advice, etc. A second finding from the study suggests that information pathologies caused by ta l l hierarchies may not be as problem-atic as has been suggested (Wilensky, 1967). Furthermore, the study indicates that managerial access to information is somewhat improved by the absorption of uncertainty that takes place in hierarchies (March and Simon, 1958). One conclusion from this is that f i l t r a t ion of informa-- 173 -tion is helpful to managers in hierarchies. Therefore, in hierarchies with few levels of authority, systems designers can take advantage of this finding by designing outputs that can be processed by the manager (Ackoff, 1967). Even i f the level of detail in a report is reduced so as to l imit its apparent usefulness, the designer must be aware that managers frequently have alternative sources from which to gather and compare information (Simon et a l . , 1954). (iv) Technology of access Where computer technology is used extensively to retrieve informa-tion the findings indicate a trend to increasing formalization of access policies (see also Stewart, 1971:259). This often results in greater restrict ions placed on information access leading to access on a "need to know" basis. This has several implications for the design of infor-mation systems. F i r s t l y , the designers need to be aware of the potential impact upon access patterns that their computer-based information systems can have. Generally, i t is suggested that changes in access policy should not be decided by systems designers alone (Hedley, 1970). Policies concerning access should be decided at a suitably high level of mana-gerial decision-making. In at least one company v is i ted , a management committee had been formed to address this issue and to make recommenda-tions. This would appear to be a reasonable procedure. Secondly, i t is also suggested that those managers whose access patterns are i n f l u -enced by such changes should be involved in the design process. This follows the general wisdom in MIS which advocates "user involvement" as a means to improve information systems designs (Swanson, 1974; Keen and Scott Morton, 1978; Bjorn-Anderson and Hedberg, 1977). - 174 -Another related aspect of the impact of computer technology upon access is the increasing use of database management systems to administer information. The rationale behind these systems is the bel ief that in -formation is a corporate asset that can and should be made available to those who need i t regardless of where i t originated (Everest, 1974). However, in several- of the interviews i t was clear that the originators of data were not always prepared to pool information because by doing so they perceived a lessening of their control of i t (Argyris, 1971; Wilensky, 1967:182). This suggests that designers need to be aware of this problem as well. In fact , the ab i l i ty of computers to prevent un-authorized access to information may be helpful to designers here. New systems can be created that maintain traditional access patterns while 2 taking advantage of the data base concept. Once more, the cooperation of those departments who "own" the data should be considered as an essential part of any changes that are contemplated. F inal ly , the maturity of the computerized information system was reflected in the better access afforded to management. It would appear that when major systems problems have been resolved and a period of s tab i l i ty ensues, a genuine improvement in access is possible. Designers of systems should therefore concentrate their efforts on those areas which are judged to have major problems (see ( i i ) , for example). Date (1975:288ff) demonstrates that certain database designs can even apply access control at the level of individual " f i e lds " . Previously, the designer of file-based systems was restricted to controll ing access to whole records of data. This change in capability may also have implications for departmental control of access. - 175 -(v) Attitudes to data sharing One of the strongest findings of the study was the positive effect of accepting "norms" of data sharing and of more trust and openness on the direct and indirect regulation of access to information. Where man-agers normally shared information or where managers were trustful and open about information, there were signif icant improvements in both managers' authority and managers' ab i l i t i es to access information. To improve access to information, therefore, i t would seem desirable to improve and sustain a positive attitude towards sharing information. Eleven of the managers interviewed indicated that in their organizations i t is the executives who often set the "style" of sharing data. In some cases this style was one of encouragement which pervaded the whole of the organization. In other cases, top management discouraged data sharing and applied rules s t r i c t l y to prevent managers accessing informa-tion not required to perform their tasks. To encourage data sharing implies that both the attitudes of chief executive off icers and those of managers below may need to be changed. Clearly, many systems designers would not be competent in this area. It would require the use of organ-izational spec ia l is ts , perhaps working in a team with systems designers (Bjorn-Anderson and Hedberg, 1977). These special ists could employ some of the techniques used in organizational development (Miner, 1973:275ff). Addit ional ly, more f a c i l i t i e s to promote access to information could be instal led or improved i f they already exist. Using direct regulation to support managerial access is also suggested as a way to encourage managers to exercise their rights of access. Because access and trust and openness are thought to be related by mutual causality, by improving access to information as specified above, benefits of higher trust and - 176 -openness should accrue to the organization. (vi) Individual coping strategies The study uncovered some of the coping strategies employed by individual managers in order to overcome access deficiencies. They in -cluded the development of informal networks, the use of lower-level employees to col lect information, and the use of personal record-keeping systems. A fourth strategy, involving the f i l t e r ing of information, has already been dealt with above. Although in any social system there wil l always be informal access to information (Downs, 1967), an exces-sive use of informal channels is an indication of information pathologies (Wilensky, 1967). The unavailabil ity of information because of a man-ager's lack of authority or because of the inconvenience of access can result in a "loss of energy consumed in the struggle for information. Time is occupied by attempts to obtain and to hide information" (Forrester, 1965). In the case of informal record-keeping, there is a clear dupl i -cation of effort on the part of the managers which results from the poor ava i labi l i ty or low relevance of the formal system to provide the needed information. Al l of these points imply that where the formal information system is r ig id in i ts patterns of access or inadequate in its content or convenience then managers wil l expend time and resources in the pursuit of the information that they believe is required. Some of these pathologies can be reduced by the careful design of the formal information system. For example, one of the major reasons that managers used their own record-keeping systems was the inconvenience of using the centralized records. Where possible, central records should - 177 -be decentralized either physically or through the use of an appropriate 4 electronic communications system. Prescriptions for reducing some of the other types of pathologies include reducing inconveniences to access, p a r t i c u l a r l y r e t r i e v a l , and authorizing access to a wider selection of information sources. Again any changes should be designed with the cooperation of those managers who w i l l be affected. If informal channels of access are determined to be important then an additional strategy is to make them part of the formal system (Forrester, 1965). 2. Specific prescriptions Prescriptions are now presented from the findings according to the type of information. The types of information are those used throughout the study; c o n f i d e n t i a l , non c o n f i d e n t i a l , and personal. ( i ) Confidential information Three preliminary remarks need to be made. F i r s t l y , the amount of confidential information and the value ascribed to i t varies with the type of enterprise that i s being pursued and the competitive nature of the market place the company operates i n . Secondly, there appears to be j u s t i f i c a t i o n in c l a s s i f y i n g some other types of information as con-f i d e n t i a l even when, s t r i c t l y speaking, they are not. For example, in several cases internal reports or organizational information concerning redundancies were leaked to unions which exploited the information to _ Simon et a l . (1954:7) came to essent i a l l y the same conclusion in their study of contr o l l e r s . "The most important consequences of cen t r a l i z a - tion or decentralization of the records function have to do with the a c c e s s i b i l i t y of the documents and the r e l i a b i l i t y of the source records. Both of these c r i t e r i a point in the direction of r e l a t i v e l y great geographic decentralization." (their emphasis) 4 There are important cost/benefit considerations when computer data bases are distributed. See Champine (1977) for a good discussion of these. - 178 -their own advantage. It is suggested that such sensitive information should be c lass i f ied as confidential . Thirdly, the study indicates that confiden-t ia l information that a manager requires for his job has poorer access than non confidential information even when the latter type of information is not required for the job. This demonstrates the greater security and possibly the greater centralization afforded to confidential information. It appears that companies have some just i f icat ion in protecting con-f idential information from unauthorized internal access and supplying i t to most managers on a "need to know" basis. In general, this protection currently takes the form of indirect regulation as can be seen from the discrepancy in the results of the study between the orderings for authority and the orderings for the ab i l i ty to access confidential information. Un-fortunately, making information inconvenient to access in order to keep i t from those who do not need i t is often reflected in the d i f f i cu l ty of access for those who do need i t . It is suggested therefore that access to confidential information should be regulated direct ly rather than in -d i rect ly , and that access should be made more convenient for those who do need i t for their jobs. Additionally, i t is suggested that much con-f idential information can be made available to al l managers simply by changing i ts form (e.g. by summarizing information or using special codes) or by releasing i t after a c r i t i ca l date has passed. In this way the organization can protect confidential information from external 5 leakage while encouraging an open attitude to information sharing. This does not go as far as Forrester (1965) has advocated. He claims that "information is withheld from individuals inside the organization on the excuse that this keeps information from outsiders" and suggests that wider access should be available to employees. However, some in -formation is very valuable to organizations. It is reasonable that organizations treat i t with special care. The problems arise when too much information is incorrectly c lass i f ied as confidential . - 1 7 9 -Furthermore, i t is suggested that information system designers should periodically re-evaluate the c lass i f icat ion of information in an organiza-t ion. In this way the organization can avoid the unnecessary "costs" of access associated with information being misclassified as confidential . ( i i ) Non confidential information The findings indicate that non confidential information is both more authorized and easier to access when i t is required for the job compared with when i t is not. These distinctions may just ref lect the fact that managers have overcome many of the problems of access for non confidential information they need for their jobs. They have learned how to minimize the "costs" of accessing this information. Nonetheless, i t was clear from the interviews that many organizations do make distinctions between information needed and not needed and provide access to non confidential information using the "need to know" formula. In contrast to this attitude, i t is suggested that no distinctions should be made for non confidential information. It should be made ava i l -able to managers regardless of whether they need i t for their work or not. Clearly, where non confidential information is needed for the job then access should be made as convenient as possible for the managers. This can be done by the use of different f a c i l i t i e s to promote access to infor-mation and by the designer concentrating on reducing the larger barriers to access, particularly retr ieval . While i t is not suggested that such good access be given to non confidential information that managers do not require, i t should not be withheld from them i f they want i t . They should be both authorized to access i t and i t should be made available to them (Forrester, 1965). This should result in two advantages to organizations. - 180 -F i r s t l y , by cutting down on unnecessary restr ic t ions, the organizational attitude to data sharing should improve. Secondly, the greater freedom of access provides managers with greater scope for improving their decision-making (Pounds, 1969; Mintzberg, 1973). The one exception to the above suggestions is where interdepartmental access to information is concerned. Although some might advocate greater access as a general rule (Forrester, 1965), others have recognized some of the problems of allowing unlimited interdepartmental access (Ackoff, 1967; Argyris, 1971). This "problem was confirmed several times in the interviews. However, apart from the problems of database design (see ( i i i ) above), i t is suggested that departments form natural "walls" to the access of detailed information, particularly i f departments are re-warded for competing with one another (Ackoff, 1967). An effective way of removing these walls is for organizations to foster an attitude of trust and openness between departments and to stress the benefits of data sharing. At the same time a change in the incentive structure is also needed i f i t is not compatible with the sharing of data (Forrester, 1965). ( i i i ) Personal information The study showed that managers' own personal details and those of their subordinates were treated as non confidential information with respect to the direct and indirect regulation of access. Personal details about other managers, in contrast, were treated even more carefully than conf i -dential information not required for the managers' jobs. Access to per-sonal information should in general be carefully regulated not because the information would be damaging to an organization i f i t were leaked outside but because of the damage i t might cause to individual employees. - 181 -The employees' needs for privacy of their personal information must be considered along with the needs of the organization to gather and use such information (Sieghart, 1976; Great B r i t a i n , Parliament, 1972). Although in Canada there i s no l e g i s l a t i o n to govern the c o l l e c t i o n and access of personal information on employees i t i s suggested that companies should consider the r e s t r i c t i o n of personal information gathering to what i s reasonably necessary and adopt access procedures similar to those used by I.B.M. (U.S.) (Cary, 1976; Great B r i t a i n , Parliament, 1972). Personal i n -formation would then be made available to those who have a legitimate need for i t . Although currently there are few incentives to change, i t is suggested that improvements in access rights w i l l also be accompanied by a better "climate" of data-sharing (Zand, 1972). Changes in direct regulation of access to personal information must be matched by improvements in the convenience of access. The barriers that were associated with poor access to personal information were largely of the physical and a t t i t u d i n a l type (location, embarrassment, permission of superior, etc.). This was borne out by the interviews where the incon-venience of location and the need for managers to j u s t i f y their l e g i t i -mate requests for information were mentioned frequently. It i s suggested therefore that these barriers should be reduced wherever possible. Specif prescriptions are now presented for each of the categories of personal information. a) Managers' own details Managers generally appear to have good authority to access their own f i l e s . There were, however, greater problems in obtaining the information that respect this type of information was treated as confidential informa-- 182 -tion when i t is required for the job. The discrepancy between authority and access ab i l i t y is thought to be caused partly by the inconvenience of accessing such information (see above) and partly by the assumption of some managers that they have a right to access their own f i l e s . In the interviews, several managers had never tried to access their own f i l es and assumed that they had the necessary authority. To c lar i fy the posi-tion of those in this situation i t is suggested that managers be given the right to see their own f i l es and rebut them i f necessary and that this right of access should be promoted in organizations. Again, this is another move towards improving the attitude of data-sharing in a company. (b) Subordinates' details Managers often need information about subordinates on a daily basis for performing a variety of personnel tasks (interviewing, evaluations, grading changes, etc . ) . If access to personnel records is made too in -convenient or i f the content is inadequate, then managers will create their own record-keeping systems on subordinates. Both of these problems were frequently mentioned in the interviews. It is suggested, therefore, that wherever possible the personnel records should be decentralized to the departments and that a careful requirements analysis be undertaken to determine what information managers need for these tasks (Bariff , 1975). Some sensit ive, personal information (e.g. medical records) can be retained by the personnel department. This department should also be used to ensure that the decentralized records are being administered in accordance with the rights of access given to employees or negotiated by unions (Cary, 1976; Forrester, 1965). - 183 -(c) Other managers' details Because personal information should generally be available only on a "need to know" or "right to know" basis the prevailing practice of re-s t r ic t ing access to information on other managers is reasonable. It is acknowledged that organizations do prof i t from the current secrecy that surrounds middle managers' salaries (Forrester, 1965). However, i t would appear from the findings that many managers prefer the established system. •Nonetheless, some of the mysteries that currently surround salar ies, grades, and evaluation procedures could be removed economically and effectively by releasing expl ic i t statements on these subjects and by publishing sta-t i s t i ca l analyses on salaries by grade, for example. - 184 -II. Prescriptions for Future Research 1. Methodological overview The research methodology adopted in this study has been the sample survey or correlational design (Campbell and Stanley, 1966) where the object is to measure behaviours or perceptions in a "natural" setting (Runkel and McGrath, 1972). As with every design, there are drawbacks as well as advantages to this methodology. F i r s t l y , there is the problem with causal ambiguity (Bouchard, 1976). Str ict ly relational studies cannot answer the questions of causality. There may always be other, uncontrolled conditions that were responsible for the relationships observed. Campbell and Stanley note, however, that these studies are useful in "exposing hypotheses to disconfirmation" (Campbell and Stanely, 1966:64). Secondly, to what degree is the sample of managers selected representative of a l l managers? The more external val id i ty that can be claimed for the sample selection procedures the greater the possibi l i ty of generalizing the results (Kerlinger, 1967). Unfortunately, there are several factors that l imit external val idity in this study. The managers were chosen in one province of Canada. It may therefore be possible to say that the results are useful to managers in that province as a whole or to managers in similar provinces of Canada. However, because of the lack of heavy industry and the high frequency of company branch offices in Brit ish Columbia i t would be premature to generalize the findings to Canada as a whole. The remoteness of branch offices from headquarters may have a signif icant impact upon internal access to information. Furthermore, by dealing with one province i t was possible to control the differences in law that exist between provinces and, further a f i e ld , between countries. However, i t is - 185 -l ike ly that the results of a similar study of access in some Scandinavian countries, for example, would ref lect the different "climate" concerning rights of workers found in those countries (Sieghart, 1976). F ina l ly , the study concentrated largely upon the perceptions of the managers and not their behaviour. Managers' perceptions of access regulation may or may not resemble the real situation. It was assumed, however, that such perceptions are important to any subsequent choice of access a manager might make and this was confirmed somewhat by the interviews. Furthermore, the questionnaire was designed to allow the managers the opportunity to indicate those types of information they are barred from. It is suggested that this prevention of access is not just a perception of managers but indicates an organizational real i ty . The subsequent analysis of the access of these managers compared with the ratings of other managers who per-ceived their access as high, revealed dist inct differences in some of the independent variables (Chapter four). In a preliminary study such as this one the research methodology chosen seems appropriate (Campbell and Stanley, 1966:64). It enabled the collection of data that was broad in scope and high in complexity. While much research l i terature is relevant to access no attempt has been made to systematically investigate the topic. Therefore, the l i terature enabled hypotheses to be suggested but the study required a broad range of potentially relevant variables to ref lect our current state of knowledge concerning access to information. The use of both a survey instrument and interviews allows the results to be stated with more confidence than would be possible using a single methodology (Bouchard, 1976). Although the "costs" of the study were considerably increased by the use of dual methodologies, the benefits of such an approach were considered to easily - 186 -outweigh these. In retrospect, the interviews were invaluable for inter-preting many of the findings. 2. Prescriptions for research The findings of the study indicate that access to information is a f ru i t fu l area for research that has application in several related d i s c i -plines. The following suggestions for further research of access to in -formation are made partly to overcome some of the methodological weak-nesses of the study that were discussed above. In addition, the interview findings produced a number of potentially valuable factors which may be influential in determining access. These are also mentioned. By using alternative designs such as the naturally occuring f ie ld experiment (Runkel and McGrath, 1972) i t should be possible to explore the effects of changes in an organization upon the regulation of access to information. For example, access to information could be measured before and after the implementation of a new information system. Similarly, the effects of implementing employee rights of access regulation could be studied using the same methodology. The change in access would at least partly indicate the effectiveness of the implementation. Further longi-tudinal f i e ld studies could be effective in revealing the organizational processes that employees use to control the access to information for others (e.g. Pettigrew, 1973; Mechanic, 1962). Studies similar to the current one could be conducted for managers in companies across Canada and in other countries. Such studies would increase the representativeness of the sample. Additionally, the subject domain could be extended to include other groups of employees such as executives, unionized and non-unionized employees as well as middle man-agers. - 187 -Final ly , the structure for the study of access could be modified to incorporate variables potentially important to access. For example, variables to measure the competitive nature of the company's enterprise and managerial "style" could provide additional explanatory power for the regulation of access. Furthermore, a contribution to the study of organizations could be made i f companies or departments could be c l ass i -fied according to their regulation of access. Perrow, for example, has dichotomized organizations as mechanistic or human-relations (Perrow, 1973). Are such c lassi f icat ions reflected in the way organizations regu-late access? (cf. Athanassiades, 1973). At a department leve l , are the categories 'X' and 'Y' developed by McGregor (1960), useful in describing the different "styles" of regulating access to employees used by managers? This study has emphasized access as the dependent variable and the pre-scriptions have concentrated on suggesting ways of improving managerial access to information. From an organization's perspective the effect of more or less access to information upon a manager's performance is an equally important question. Does better access always lead to improved performance? (Forrester, 1965; Ackoff, 1967). Unfortunately i t is often d i f f i cu l t to obtain meaningful performance measures that can be related to the regulation of access. There would normally be many competing hypotheses that the relationship between access and performance could support. However, i f such measures were available then they should be incorporated into the structure as dependent variables wherever possible. - 188 -BIBLIOGRAPHY A c k o f f , R u s s e l l L . "Management M i s i n f o r m a t i o n S y s t e m s " . Management S c i e n c e 1 4 ( 4 ) , December 1967 . A g u i l a r , F . J . S c a n n i n g the B u s i n e s s E n v i r o n m e n t . New Y o r k : C o l u m b i a U n i v e r s i t y , 1967 . A i k e n , 0 . " C o n f i d e n t i a l I n f o r m a t i o n and t h e E m p l o y e e " . P e r s o n n e l Manage- ment 6 : 4 0 , J u l y 1974 . A r g y r i s , C . "Management I n f o r m a t i o n S y s t e m s : The C h a l l e n g e t o R a t i o n a l i t y and E m o t i o n a l i t y " . Management S c i e n c e 1 7 ( 6 ) : B 2 7 5 - 9 2 , 1 9 7 1 . A r r o w , K . The L i m i t s o f O r g a n i z a t i o n . New Y o r k : W.W. N o r t o n , 1974 . A t h a n a s s i a d e s , J . C . "The D i s t o r t i o n o f Upward C o m m u n i c a t i o n i n H i e r a r c h i -c a l O r g a n i z a t i o n s " . Academy o f Management J o u r n a l 1 6 ( 2 ) : 2 0 7 - 2 5 , 1 9 7 3 . A u e r b a c h P u b l i s h e r s . " Impac t o f P e r s o n a l P r i v a c y R e q u i r e m e n t s on Data Base A d m i n i s t r a t i o n " i n Da ta Base Management . P h i l a d e l p h i a : A u e r b a c h P u b l i s h e r s , I n c . , 1976 . A y k r o y d , C o l i n . " O c c u p a t i o n a l H e a l t h and S a f e t y U p d a t e " . L a b o u r R e s e a r c h B u l l e t i n ( B . C . ) 8 ( l ) : 2 7 - 4 2 , J a n u a r y 1980a . A y k r o y d , C o l i n . "OSHA Symposium f o r L a b o u r E d u c a t o r s on O c c u p a t i o n a l S a f e t y and H e a l t h " . L a b o u r R e s e a r c h B u l l e t i n . ( B . C . ) 8 ( 1 ) : 4 3 - 5 1 , J a n u a r y 1980b . B a c h a r a c h , S . B . and M i c h a e l A i k e n . " C o m m u n i c a t i o n i n A d m i n i s t r a t i v e B u r e a u c r a c i e s " . Academy o f Management J o u r n a l 2 0 ( 3 ) : 3 6 5 - 3 7 7 , 1977 . B a r i f f , M a r t i n L . " I n f o r m a t i o n R e q u i r e m e n t s A n a l y s i s : A M e t h o d o l o g i c a l R e v i e w " . Da tabase 7 ( 1 ) , 1975 . B a r i f f , M a r t i n L . and J a y R. G a l b r a i t h . " I n t r a o r g a n i z a t i o n a l Power C o n -s i d e r a t i o n s f o r D e s i g n i n g I n f o r m a t i o n S y s t e m s " . A c c o u n t i n g , O r g a n - i z a t i o n s and S o c i e t y 3 ( 1 ) , 1978 . B a r n a r d , C h e s t e r I . The F u n c t i o n o f t h e E x e c u t i v e . C a m b r i d g e , M a s s : H a r v a r d U n i v e r s i t y P r e s s , 1938 . B e n b a s a t , I z a k , and A l b e r t S . D e x t e r . " V a l u e and E v e n t s A p p r o a c h e s t o A c c o u n t i n g : An E x p e r i m e n t a l A p p r o a c h " . A c c o u n t i n g Rev i ew 5 4 ( 4 ) : 7 3 5 - 7 4 9 , O c t o b e r 1979 . B e n b a s a t , I z a k , and R o n a l d N . T a y l o r . "The Impact o f C o g n i t i v e S t y l e s on I n f o r m a t i o n Sys tem D e s i g n " . MIS Q u a r t e r l y June 1978 . - 189 -Benson, P.G. "Personal Privacy and the Personnel Record". Personnel Journal 57:376-80, 395, July 1978. Berry, A.J. and D.T. Otley. "The Aggregation of Estimates in Hierarchical Organizations". Journal of Management Studies 12(2), 1975. Bjorn-Anderson, Niels, and Bo Hedberg. "Designing Information Systems in an Organizational Perspective", in Prescriptive Models of Organiza- tions . Paul Nystrom and William Starbuck (eds.), North Holland/TIMS Studies in Management Science 5, 1977. Bouchard, Thomas J . "Field Research Methods: Interviewing, Questionnaires, Participant Observation, Unobtrusive Measures" in Handbook of Industrial and Organizational Psychology. Marvin Dunnette (ed.), Chicago: Rand McNally, 1976, pp. 364-414. Brady, Rodney, H. "Computers in Top-level Decision Making". Harvard Business Review July-August 1967. Burch, John G., and Felix R. Strater. Information Systems: Theory and Practice. Santa Barbara, Ca l i f . : Hamilton Publishing Co., 1974. Burke, R., and D. Wilcox. "Effects, of Different Patterns and Degrees of Openness in Superior-Subordinate Communication on Subordinate Job Satisfact ion". Academy of Management Journal 12:319-326, 1969. Burns, Tom and G.M. Stalker. The Management of Innovation. London: Tavistock, 1961. Campbell, Donald T. and Julian C. Stanley. Experimental and Quasi- Experimental Designs for Research. Chicago: Rand McNally, 1966. Canadian Task Force. Privacy and Computers. Ottawa: Information Canada, 1972. Carter, N. "The Executive and the Terminal", in J.D. White (ed.), Proceedings, Sixth and Seventh Annual Conferences of the Society for Management Information Systems. Ann Arbor: University of Michigan, Management Information Systems Research Center, July 1976, pp. 87-89. Cary, Frank T. "IBM's guidelines to employee Privacy". Harvard.Business .Review 54:82-90, September - October 1976. Champine, G.A. "Six Approaches to Distributed Data Bases". Datamation May, 1977. Chervany, Norman L. and Gary W. Dickson. "An Experimental Evaluation of Information Overload in a Production Environment". Management Science 20:10, June 1974. Churchman, C. West, and A.H. Schainblatt. "The Researcher and the Manager: A Dielectic of Implementation". Management Science 11(4): B69-B87, February, 1965. - 190 -Codd, E.F. "Seven Steps to Rendezvous With the Casual User", in Data Base Management. J.W. Klimbie and K.L. Koffeman (eds.), Amsterdam: North-Holland Publishing Co., 1974. Comstock, D.E. and W.R. Scott. "Technology and the Structure of Subunits: Distinguishing Individual and Workgroup Effects". ASQ, 22:177-202, June 1977. Crozier, Michel. The Bureaucratic Phenomenon. Chicago: University of Chicago Press, 1964. Cyert, Richard M. and James G. March. A Behavioral Theory of the Firm. Englewood C l i f f s , New Jersey: Prentice-Hall, 1963. Date, C.J. An Introduction to Database Systems. Reading, Mass.: Addison-Wesley, 1975. Dewhirst, H. Dudley. "Influence of Perceived Information Sharing Norms on Communication Channel U t i l i za t ion" . Academy of Management Journal 14(3):305-316, September 1971. Dhalakia, R.R., and B. Sternthal. "Highly Credible Sources, Persuasive Faci l i tators or Persuasive L i ab i l i t i e s " . Journal of Consumer Research 3(4):223-32, March 1977. Divorski, Stanley, Andrew C. Gordon, and John P. Heinz. "Factors Affect-ing the Accessibi l i ty of. Information Held by Public Agencies: The Case of the City of Chicago Building Department". Northwestern University Law Review 68(2):309-335, May-June 1973. Downs, Anthony. Inside Bureaucracy. Boston: L i t t l e , Brown, 1967. Dueltgen, Ronald R. "Regulations and Responsibility: The CSIN and the Private Sector". Proceedings of the ASIS Annual Meeting 16:363, 1979. Everest, Gordon C. "Database Management Systems Tutor ia l " , in Fifth Annual Midwest AIDS Conference Proceedings. Vol. 1, Norman L. Chervany (ed.), Minneapolis: University of Minnesota, College of Business Administration, 1974:A1-A12. Ewing, David W. "What Businesses think about employee r ights" . Harvard Business Review 55(5)..:81-94, Sept.-Oct. 1977. Feltham, Gerald A. Information Evaluation. Sarasota, Florida: American Accounting Association, 1972. Financial Statements. Brit ish Columbia: U.B.C., March 31, 1979 issued in compliance with Public Bodies Financial Information Act Chapter 51, Statutes of Brit ish Columbia, 1961. Forrester, Jay W. "A New Corporate Design". Industrial Management Review 7:5-17, 1965. - 191 -G a l b r a i t h , J a y R. D e s i g n i n g Complex O r g a n i z a t i o n s . R e a d i n g , M a s s . : A d d i s o n W e s l e y , 1 9 7 3 . G e r s t b e r g e r , P . G . , and T . J . A l l e n . " C r i t e r i a Used by R and D E n g i n e e r s i n t he S e l e c t i o n o f an I n f o r m a t i o n S o u r c e " . J o u r n a l o f A p p l i e d P s y c h o l o g y 52(4 ) : 2 7 2 - 7 9 , 1968 . G i b b , J a c k R. " C l i m a t e f o r T r u s t F o r m a t i o n " , i n L e l a n d P . B r a d f o r d e t a l . ( e d s . ) , T - G r o u p T h e o r y and L a b o r a t o r y M e t h o d . New Y o r k : John W i l e y , 1964 , p p . 2 7 9 - 3 0 1 . G i b s o n , C y r u s F . , and R i c h a r d L . N o l a n . "Manag ing the F o u r S t a g e s o f EDP ' G r o w t h " . H a r v a r d B u s i n e s s Rev iew J a n . - F e b . 1974 . G o l d s t e i n , R o b e r t C. "The C o s t s o f P r i v a c y " . D a t a m a t i o n O c t o b e r 1975 . G r e a t B r i t a i n , P a r l i a m e n t . R e p o r t o f t h e Commi t t ee on P r i v a c y . Cmnd 5012 L o n d o n : H . M . S . O . , J u l y 1972 . G r e a t B r i t a i n , P a r l i a m e n t . C o m p u t e r s : S a f e g u a r d s f o r P r i v a c y . Cmnd 6354 L o n d o n : H . M . S . O . , December 1975 . G r e e n e , R o b e r t J o h n . F a c u l t y A c c e p t a n c e and Use o f a Sys tem P r o v i d i n g Remote B i b l i o g r a p h i c and P h y s i c a l A c c e s s t o an A c a d e m i c L i b r a r y . U n p u b l i s h e d P h . D . D i s s e r t a t i o n , F l o r i d a S t a t e U n i v e r s i t y , 1973 . G r e g o r y , R o b e r t H . , and R . L . Van H o r n . A u t o m a t i c D a t a - P r o c e s s i n g S y s t e m s . B e l m o n t , C a l i f o r n i a : W a d s w o r t h , 1960 . Hage , J . , and M . A i k e n . " R o u t i n e T e c h n o l o g y , S o c i a l S t r u c t u r e and O r g a n -i z a t i o n a l G o a l s " . ASQ 1 4 : 3 6 6 - 3 7 6 , 1969 . Hage , J . , M . A i k e n , and C . B . M a r r e t t . " O r g a n i z a t i o n a l S t r u c t u r e and C o m m u n i c a t i o n s " . A m e r i c a n S o c i o l o g i c a l R e v i e w 3 6 : 8 6 0 - 7 1 , 1 9 7 1 . H e d l e y , R. A l a n . " O r g a n i z a t i o n a l O b j e c t i v e s and M a n a g e r i a l C o n t r o l s : A S t u d y o f C o m p u t e r i z a t i o n " , i n I n d u s t r i a l O r g a n i z a t i o n - B e h a v i o u r and C o n t r o l . J . Woodward ( e d . ) , L o n d o n : O x f o r d U n i v e r s i t y P r e s s , 1970 . H i c k s o n , D . J . , H i n n i n g s , C . R . , L e e , C . A . , S c h n e c k , R . E . , and J e n n i n g s , J . M . "A S t r a t e g i c C o n t i n g e n c i e s T h e o r y o f I n t r a - O r g a n i z a t i o n a l P o w e r " . A . S . Q . 1 6 ( 2 ) : 2 1 6 - 2 9 , 1 9 7 1 . H i n n i n g s , C . R . , H i c k s o n , D . J . , P e n n i n g s , J . M . , and R . E . S c h n e c k . " S t r u c t u r a l C o n d i t i o n s o f I n t r a O r g a n i z a t i o n a l P o w e r " . A . S . Q . 1 9 ( 1 ) , March 1974 . H i r s c h h o r n , E r i c L . " S u n s h i n e f o r F e d e r a l A g e n c i e s " . A m e r i c a n B a r A s s o c i a - t i o n J o u r n a l 6 3 : 5 5 - 5 8 , J a n u a r y 1 9 7 7 . K e e n , P e t e r G . W . , and M i c h a e l S . S c o t t M o r t o n . D e c i s i o n S u p p o r t S y s t e m s : An O r g a n i z a t i o n a l P e r s p e c t i v e . R e a d i n g , M a s s . : A d d i s o n - W e s l e y , 1 9 7 8 . - 192 -Kerlinger, Fred N. Foundations of Behavioral Research. New York: Holt, Rinehart and Winston, Inc., 1967. Kincaid, H.V., and M. Bright. "The Tandem Interview: A Trial of the Interview Team". Public Opinion Quarterly 21:304-312, 1957. Koester, R., and F. Luthans. "The Impact of the Computer on the Choice Activity of Decision Makers: A Replication With Actual Users of Computerized MIS". Academy of Management Journal 22:416-422, June 1979. Ladendorf, Janice M. "Information Flow in Science, Technology and Commerce". Special Libraries 61 (5):215-222, May-June 1970. Lawler, Edward E., Porter, Lyman W., and Allen Tennenbaum. "Managers' Attitudes Towards Interaction Episodes". Journal of Applied Psychology 52(6):432-439, 1967. Lawrence, Paul and Jay Lorsch. Organization and Environment. Boston: Harvard University Press, 1967. Lazarsfeld, Paul F., B. Berelson, and Hazel Gaudet. The People's Choice. New York: Columbia University Press, 1968, 3rd Edition. Learned, Edmund P., and Audrey T. Sproat., Organization Theory and Practice. Homewood, 111.: Richard D. Irwin, 1966. Lucas, Henry C. Jr.,. and Jimmy A. Sutton. "The Stage Hypothesis and the S Curve: Some Contradictory Evidence". Communications of the A .CM. 20:4, April 1977. Lucas, Henry C. Jr. "Unsuccessful Implementation: The Case of a Com-puter-Based Order Entry System". Decision Sciences 9(1) January 1978. Luthans, F., and R. Koester. "Impact of Computer Generated Information on the Choice Activity of Decision Makers". Academy of Management Journal 19:328-332, June 1976. McGregor, Douglas. The Human Side of Enterprise. New York: McGraw-Hill, 1960. March, James G., and Herbert A. Simon. Organizations. New York: John Wiley and Sons, 1958. Martin, James. Principles of Database Management. Englewood C l i f f s , N.J.: Prentice Hal l , 1976. Mason, Richard 0. Jr. "Basic Concepts for Designing Management Informa-tion Systems". A.I.S. Research Paper No. 8, U.C.L.A., October 1969. Mason, Richard 0. , and Mitroff , Ian I. "A Program for Research on M.I.S." Management Science 19(5), January 1973. - 193 -Mechanic, D. "Sources of Power of Lower Participants in Complex Organ-izat ions". ASQ 7 (3):349-64 , J % 2 . Miner, John B. The Management Process. Theory, Research, and Practice. New York: Macmillan, 1973. Mintzberg, H. The Nature of Managerial Work. New York: Harper and Row, 1973. Mintzberg, Henry. Impediments to the Use of Management Information. Hamilton, Ontario: Society of Industrial Accountants of Canada, 1975. Muchinsky, P.M. "Organizational Communication: Relationships to Organ-izational Climate and Job Sat isfact ion". Academy of Management Journal 20:592-607, December 1977. Nolan, Richard L. "Thoughts about the Fifth Stage". Database 7(2), Fall 1975. Nolan, Richard L. "Managing the Cris is in Data Processing". Harvard Business Review March-April 1979. Nord, Walter R. (ed.). Concepts and Controversy in Organizational Behaviour. Pacific Palisades, Ca l i f . : Goodyear, 1972. O'Rei l ly , Charles A. "Supervisors and Peers as Information Sources, Group Supportiveness and Individual Decision-Making Performance". Journal of Applied Psychology 62(5) :632-635, 1977. O'Rei l ly , C , and K. Roberts. "Relationships Among Components of Credi-b i l i t y and Communication Behaviors in Work Units". Journal of Applied Psychology 61:99-102, 1976. Pelz, Donald, and Frank Andrews. Scientists in Organizations. New York: Wiley, 1966. Perrow, Charles. "The Short and Glorious History of Organizational Theory". Organizational Dynamics 2(1):2-15, Summer 1973. Pettigrew, A.M. The Pol i t ics of Organizational Decision-Making. London: Tavistock, 1973. Pettigrew, A.M. "Information Control as a Power Resource". Sociology , pp. 187-204, May 1972. Pfeffer, Jeffrey. "Organizational Design" in Kenneth D. Mackenzie (ed.) Organizational Structures. Arlington Heights, II.: AHM Publishing, .1978. Poole, Marshall Scott. "An Information - Task Approach to Organizational Communications". Academy of Management Review 3(3) :493-504, July 1978. - 194 -Porter, H.O. "Legislative Experts and Outsiders: The Two-step Flow of Communication". Journal of Pol i t ics 36(3) :703-730, August 1974. Pounds, William F. "The Process of Problem Finding". Industrial Manage- ment Review 11 (1):1-19, Fall 1969. Pugh, D.S., D.J. Hickson, CR . Hinnings, and C. Turner. "The Context of Organizational Structure". A.S.Q. 14:91-114, 1969. Read, W. "Upward Communication in Industrial Hierarchies". Human Relations 15:3-16, 1962. Relyea, H.C "The Provision of Government Information: The Federal Freedom of Information Act Experience". Canadian Public Administra- tion 20(2):317-341 , Summer 1977. Riley, T. "Labours Interest in Freedom of Information". Canadian Labour 22(3):22-25, 1977. Roberts, K.H., and C. O'Reil ly. "Failures in Upward Communications in Organizations: Three Possible Culpr i ts" . Academy of, Management Journal 17:205-215, 1974. Rogers, E.M., and R. Agarwala-Rogers. Communication in Organizations. New York: The Free Press, 1976. Rosen, S. and Tesser, A. "On Reluctance to Communicate Undesirable In-formation: The Mum'Effect. Sociometry 33:253-263, 1970. Rummel, R.J. Applied Factor Analysis. Evanston: Northwestern University Press, 1970. Runkel, Phil ip J . , and Joseph E. McGrath. Research on Human Behaviour. New York: Holt, Rinehart and Winston, 1972. Shaffer, David R., Pamela V. Mays, and Karen Etheridge. "Who Shall Be Hired: A Biasing Effect of the Buckley Amendment on Employment Practices?" Journal of Applied Psychology 61 (5):571-575, 1976. Shneiderman, Ben. "Improving the Human Factors Aspect of Database Inter-actions". ACM Transactions on Database Systems 3(4):417-439, .December 1978. Siegel, Sidney. Non-parametric Stat ist ics for the Behavioral Sciences. New York: McGraw-Hill, 1956. Sieghart, Paul. Privacy and Computers. London: Latimer, 1976. Simon, H., Guetzkow, H., Kozmetsky, G., and G. Tyndall. Centralization vs Decentralization in Organizing the Control ler 's Department. New York: Controllership Foundation, Inc., 1954. - 195 -Simon H.A. The New Science of Management Decision. New York: Harper and Row, 1960. Simon, Herbert A. The Shape of Automation for Men and Management. New York: Harper and Row, 1965. Simon, Herbert A. , and Allen Newell. "Human Problem Solving: The State of the Theory in 1970". American Psychologist 26(2):145-159, February 1971. Smith, J .E . and N.P. Smith. "Readability: A Measure of the Performance of the Communication Function of Financial Reporting". Accounting Review 46(4) :552-561 , July 1971. Sorter, G.H. "An Events Approach to Basic Accounting Theory". Accounting Review 44(1): 12-19, January 1969. Steers, Richard M. Organizational Effectiveness A Behavioural View. Santa Monica, Ca l i f . : Goodyear, 1977. Stewart, Rosemary. How Computers Affect Management. London: Pan Books, 1971. Sussman, L. "Perceived Message Distortion, or You Can Fool Some of the Supervisors Some of the Time". Personnel Journal 53:679-682, September 1974. Swanson, E. Burton. "M.I.S.: Appreciation and Involvement". Management Science 21(2)-178-188, October 1974. Taylor, Frederick W. Sc ient i f ic Management. New York: Harper and Row, 1947. Thompson, J.D. Organizations in Action. New York: McGraw-Hill, 1967. Turney, Peter B.B. "Transfer Pricing in M.T.S.". M.I.S. Quarterly March 1977. Turoff, M., and Roxanne H. Starr. "User Behavior Patterns in the Electronic Information Exchange System". Proceedings of the A .CM. 12:659-665, 1978. Tushman, Michael L. "Work Characteristics and Subunit Communication Structure: A Contingency Analysis". A.S.Q. 24:82-97, March 1979. Weber, Max. "The Ideal Bureaucracy", in The Theory of Social and Economic Organizations. Glencoe: The Free Press, 1947. Weber, Max. "The Essentials of Bureaucratic Organization: An Ideal-Type Construction", in Robert Merton et a l . , Reader in Bureaucracy. Glencoe, 111.: The Free Press, 1952. - 196 -Weick, Karl E. The Social Psychology of Organizing. Reading, Mass.: Addison-Wesley, 1969. Westin, Alan F. Privacy and Freedom. New York: Atheneum, 1967. Westin, Alan F, "The Problem of Employee Privacy s t i l l Troubles Manage-ment". Fortune 99(11 ):121-126, June 4, ,1979. Westin, Alan F., and M.A. Baker. Databanks in a Free Society. New York: Quadrangle, 1972. Wilensky, Harold L. Organizational Intelligence. New York: Basic Books, 1967. Woodward, Joan. Industrial Organization: Theory and Practice. London: Oxford University Press, 1965. Wright, P. "The Harassed Decision Maker: Time Pressures, Distractions, and the Use of Evidence". Journal of Applied Psychology 59:555-561, 1974. Zand, D.E. "Trust and Managerial Problem Solving". A.S.Q. 17:229-240, 1972. Zloof, Moshe M., and S. Peter de Jong. "The System for Business Automa-tion (SBA): Programming Language". CACM 20:6, June 1977. Zmud, Robert W. "An Empirical Investigation of the Dimensionality of the Concept of Information". Decision Sciences 9:187-195, 1978. - 198 -SECTION I - QUESTIONS ABOUT THE RULES OF ACCESS TO INFORMATION In the f i r s t section we would l i k e to assess, what authorization your company gives managers i n your p o s i t i o n to access certain types of information. F i r s t l y , to allow us to focus on p a r t i c u l a r types of information, we would l i k e you to i n d i -cate, with a t i c k i n each case, which of the following types of information are considered c o n f i d e n t i a l by your company. For t h i s questionnaire, c o n f i d e n t i a l information i s information which i f released to other companies (or interested parties) would prove harmful to the performance of your company (or, i n the case of government agencies or u t i l i t i e s , would be an embarrassment to them). Some of the following examples may not apply to your company. If you cannot f i n d one type of information on the l i s t that would be considered c o n f i d e n t i a l by your company, please use an example that i s appropriate to your company and enter i t under the l i n e marked "other". Please t i c k i f c o n f i d e n t i a l Please t i c k i f c o n f i d e n t i a l Production figures Technological processes Chemical formulae Customer l i s t s P r i c i n g formulae Inventory l e v e l s Supplier d e t a i l s Research and development reports ************ Select one of the above examples of c o n f i d e n t i a l information that you are f a m i l i a r with and that you believe i s of most concern to your company. Please enter i t below: Market research reports Detailed sales reports Minutes of board meetings Sales invoices Minutes of managers' meetings Machine service reports Other (specify) Example of c o n f i d e n t i a l information Using t h i s example of c o n f i d e n t i a l information, read the following two statements and c i r c l e one of the numbers provided to indicate your assessment from: strongly disagree (1) to strongly agree (7). 1. In t h i s company, managers i n my p o s i -ti o n are authorized to get t h i s type of information as long as they believe i t i s needed for the performance of th e i r job c i r c l e one Strongly Strongly Disagree ( 1 2 3 4 5 6 7 ) A g r e | 2. In t h i s company, managers i n my p o s i t i o n are authorized to get t h i s type of information even though they believe i t i s not needed for the performance of t h e i r joh. Strongly Disagree c i r c l e one ( 1 2 3 4 5 6 7 ) Strongly Agree Next, assess the degree to which you think the following two statements apply to your company for one of the types of information you have not ticked ( i . e . , non-confidential information) and with which you are f a m i l i a r . If you cannot f i n d a suitable example from the l i s t , please use a more appropriate one. Enter your choice below: Example of non-confidential information 3. In t h i s company, managers i n my p o s i -tion are authorized to get thi s type of information as long as they believe i t i s needed for the performance of th e i r job Strongly Disagree c i r c l e one ( 1 2 3 4 5 6 7 ) Strongly Agree In thi s company, managers i n my p o s i t i o n are authorized to get thi s type of information even though they believe i t i s not needed for the performance of t h e i r job Strongly Disagree c i r c l e one ( 1 2 3 4 5 6 7 Strongly Agree .12 -.199;-Rules Concerning Personnel and P a y r o l l Information Companies c o l l e c t and maintain varying amounts of information concerning t h e i r employees. For the following items of personnel and p a y r o l l information we would l i k e you to ind i c a t e i f managers i n your p o s i t i o n can see these items where they concern: 5. themselves 6. t h e i r subordinates 7. other managers (peers) Please read the items and for each case c i r c l e one of the numbers provided to ind i c a t e your assessment from: strongly disagree (1) to strongly agree (7). MANAGERS IN MY POSITION ARE AUTHORIZED TO: 5. SEE 6. SEE SUB- 7. SEE OTHER Item OWN? ORDINATES? MANAGERS? Strongly Strongly Strongly Strongly Strongly . Strongly Disagree Agree Disagree Agree Disagree Agree Job Status (salary, grade, c i r c l e one c i r c l e one c i r c l e one job h i s t o r y ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) Job Performance and E v a l -uations (t e s t r e s u l t s , c i r c l e one c i r c l e one c i r c l e one evaluations, interview ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) d e t a i l s , references) General Correspondence and c i r c l e one c i r c l e one c i r c l e one Comments ( l e t t e r s of com- ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) mendation, complaint, etc.) Biographical D e t a i l s (address, telephone number, medical de- c i r c l e one c i r c l e one t a i l s , m a r i t a l and c i r c l e one family d e t a i l s , ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) q u a l i f i c a t i o n s , etc.) ./3 - 200 -SECTION II - THE PRACTICE OF ACCESS IN YOUR COMPANY In the previous section we asked you about your company's rules of access to information. In t h i s section we would l i k e you to assess your comp.-iny's p r a c t i c e of access f o r each of the d i f f e r e n t types of information. The types of information are mentioned i n the same sequence as Section I. In your company there may be c e r t a i n features that enhance or hinder access to information for managers i n your p o s i t i o n . Many of these features are represented by the statements below. Please read each statement and for each case c i r c l e one of the numbers provided to i n d i c a t e your assessment of how w e l l i t represents the current practices of your company for managers i n your p o s i t i o n , from: strongly disagree (1) to strongly agree (7). Cases 1,2 CONFIDENTIAL INFORMATION, where c o n f i d e n t i a l information i s information which i f released to other companies would be harmful to the performance of your company. Please use the same example you used on page 1 and enter i t i n the box provided: Example of c o n f i d e n t i a l information Can this type of information be obtained? Statements D i f f i c u l t or lengthy access procedures are required to get i t He never f e e l s embarrassed getting t h i s type of information He has to get the permission of h i s superior before he can get t h i s type of information The information contains too many errors It i s d i f f i c u l t to compare i t with other information (e.g. 4-weekly r e -ports when other reports are monthly) An e f f o r t i s made to locate i t i n a p o s i t i o n that i s normally convenient for the manager The existence of t h i s type of i n f o r -mation i s made widely known by the company Important d e t a i l s are missing An e f f o r t has been made to remove any heavy bias from i t The medium of the presentation i s normally matched with the p r e f e r -ences of the manager (e.g. typed rather than computer printout i s used i f preferred) The information i s badly l a i d out (e.g. graphical when tabular i s preferred) D i f f i c u l t language, symbols, or jargon i s present Others f a i l to recognise the mana-ger's legitimate authority to get t h i s information The company makes sure t h i s i n f o r -mation i s always a v a i l a b l e on time Most of the i r r e l e v a n t d e t a i l s have been eliminated Getting t h i s type of information i s c o s t l y ( i n $) Generally, i t i s easy to get and use t h i s type of information Case 1. Where i t i s needed for the managers' job Strongly Strongly Disagree Agree c i r c l e one ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) Case 2. Where i t i s not needed for the managers' job c i r c l e one YES/NO I f you c i r c l e d NO, leave that column blank Strongly Strongly Disagree Agree c i r c l e one ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ./4 - 201 -Cases 3,4 SON-CONFIDENTIAL INFORMATION. Please use the same example you used on page 1 and enter i t i n the box provided: Example of non-confidential information Can t h i s type of information be obtained? Statements D i f f i c u l t or lengthy access procedures are required to get i t He never f e e l s embarrassed getting t h i s type of information He has to get the permission of h i s superior before he can get t h i s type of information The information contains too many errors It i s d i f f i c u l t to compare i t with other information (e.g. 4-weekly re-ports when other reports are monthly)| An e f f o r t i s made to locate i t i n a p o s i t i o n that i s normally convenient for the manager The existence of t h i s type of i n f o r -mation i s made widely known by the company Important d e t a i l s are missing An e f f o r t has been made to remove any heavy bias from i t The medium of the presentation i s normally matched with the p r e f e r -ences of the manager (e.g. typed rather than computer printout i s used i f preferred) The information i s badly l a i d out (e.g. graphical when tabular i s preferred) D i f f i c u l t language, symbols, or jargon i s present Others f a i l to recognise the mana-ger's legitimate authority to get t h i s information The company makes sure t h i s i n f o r -mation i s always a v a i l a b l e on time Most of the i r r e l e v a n t d e t a i l s have have been eliminated Getting t h i s type of information i s c o s t l y ( i n $) Generally, i t i s easy to get and use t h i s type of information |Case 3. Where i t i s needed for the managers' job Strongly Strongly |Disagree Agree c i r c l e one ( 1 2 3 4 5 6 7 ( 1 2 3 4 5 6 7 ( 1 2 3 4 5 6 7 ( 1 2 3 4 5 6 7 ( 1 2 3 4 5 6 7 ( 1 2 3 4 5 6 7 ( 1 2 3 4 5 6 7 ( 1 2 3 4 5 6 7 ( 1 2 3 4 5 6 7 ( 1 2 3 4 5 6 7 ( 1 2 3 4 5 6 7 ( 1 2 3 4 5 6 7 ( 1 2 3 4 5 6 7 ( 1 2 3 4 5 6 7 ( 1 2 3 4 5 6 7 ( 1 2 3 4 5 6 7 ( 1 2 3 4 5 6 7 Case 4. Where i t i s not needed for the managers' job c i r c l e one YES/NO If you c i r c l e d NO, leave that column blank Strongly Strongly Disagree Agree c i r c l e one ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ./5 - 202 -Cases 5,6,7 PERSONNEL and PAYROLL INFORMATION Case 5. Personnel and P a y r o l l i n f o r -mation about himself Case 6. Personnel and P a y r o l l i n f o r -mation about h i s subordinates Case 7. Personnel and P a y r o l l i n f o r -mation about other managers Can at l e a s t some of t h i s type of information be obtained by mana-gers i n your position? c i r c l e one YES/NO Where you answered NO c i r c l e one c i r c l e one YES/NO YES/NO , leave the whole column blank Statements D i f f i c u l t or lengthy access procedures are required to get i t Strongly Strongly Disagree Agree c i r c l e one ( 1 2 3 4 5 6 7 ) Strongly Strongly Disagree Agree c i r c l e one ( 1 2 3 4 5 6 7 ) Strongly Strongly Disagree Agree c i r c l e one ( 1 2 3 4 5 6 7 ) He never f e e l s embarrassed getting t h i s type of information ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) He has to get the permission of h i s superior before he can get t h i s type of information ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) The information contains too many errors ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) It i s d i f f i c u l t to compare i t with other information (e.g. 4-weekly r e -ports when other reports are monthly) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) An e f f o r t i s made to locate i t i n a po s i t i o n that i s normally convenient for the manager ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) The existence of t h i s type of i n f o r -mation i s made widely known by the company ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) Important d e t a i l s are missing ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) An e f f o r t has been made to remove any heavy bias from i t ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) The medium of the presentation i s normally matched with the p r e f e r -ences of the manager (e.g. typed rather than computer printout i s used i f preferred) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) The information i s badly l a i d out (e.g. graphical when tabular i s preferred) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) D i f f i c u l t language, symbols, or jargon i s present ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) Others f a i l to recognise the mana-ger's legitimate authority to get t h i s information ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) The company makes sure t h i s i n f o r -mation i s always a v a i l a b l e on time ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) Most of the i r r e l e v a n t d e t a i l s have been eliminated ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) Getting t h i s type of information i s c o s t l y ( i n $) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) Generally, i t i s easy to get and use t h i s type of information ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) ( 1 2 3 4 5 6 7 ) - 203 -SECTION I I I - GENERAL QUESTIONS ABOUT ACCESS TO INFORMATION 1. The amount of c o n f i d e n t i a l information We would l i k e to f i n d out how much work-related information i s c o n f i d e n t i a l to your company. Please read the following statement and c i r c l e one of the numbers provided to i n d i c a t e your assessment from: strongly disagree (1) to strongly agree (7). A s i g n i f i c a n t proportion of information r e l a t e d to the work of managers i n my p o s i -t i o n i s c l a s s i f i e d as c o n f i d e n t i a l i n t h i s company (where, as before, c o n f i d e n t i a l information i s information which i f released to other companies would be harmful to the performance of your company). i c i r c l e one _^ . Strongly ( 3 Strongly Disagree Agree 2. F a c i l i t i e s to promote access to information The following i s a l i s t of f a c i l i t i e s to promote access to information which may be pro-vided by your company. Please t i c k the ones your company uses ( i f any), and f o r each one so marked, indicate i t s effectiveness i n promoting access to information for i t s current l e v e l of use. C i r c l e one of the numbers provided to i n d i -cate your assessment from: low (1) to high (7). _. , Eff e c t i v e n e s s i n Tick . . ^ i t promoting access e x i s t s Features low high ( 1 2 3 4 5 6 7 ) Meetings with your s t a f f ( 1 2 3 4 5 6 7 ) Management meetings . . ( 1 2 3 4 5 6 7 ) Technical Advisory (e.g. s t a t i s t i c s ) • • ( 1 2 3 4 5 6 7 ) L i a i s o n personnel . . . ( 1 2 3 4 5 6 7 ) Other (specify) ( 1 2 3 4 5 6 7 ) ************* SECTION IV - QUESTIONS ABOUT YOUR COMPANY, YOUR JOB, AND YOURSELF As w e l l as questions on access to information we would also l i k e to know some of the more important features of your company, your job, and yo u r s e l f . The questions asked i n t h i s section are extremely important because they allow us to study access to information i n comparative terms. Complete data are required i n order to make the most meaningful i n t e r p r e t a t i o n s of the other responses you have given us. 1. Size of your company Estimate the number of employees working i n your l o c a l company. I f you work i n a sub-s i d i a r y or a d i v i s i o n of a larger corporation, your answer should r e f l e c t the subsidiary or d i v i s i o n , not the whole organization. T i c k one box only Number of Employees Up to 200 201-400 401-600 601-800 801-1000 1001 or more (specify) ' Estimate the number of employees i n your department: The numher of employees i s 2. Shape of your company Give the number of departments i n your com-pany, subsidiary, or d i v i s i o n , where a department consists of at l e a s t two persons and two l e v e l s . The number of departments i s : c i r c l e one 1 2 3 4 5 6 7 8 9 10 more than 10 (specify) Give the number of l e v e l s of authority i n your department by counting from the department head to the lowest category of workers. The number of l e v e l s of authority i n my department i s : c i r c l e one 1 2 3 4 5 6 7 8 9 10 more than 10 (specify) ./7 - 204 -2. Shape of your company (continued) 4. Decentralization of Authority The number you have c i r c l e d on the l a s t page corresponds to the number of l e v e l s i n your department. Using the number "1" to correspond to the department head, c i r c l e which number applies to your l e v e l . If you are head of your department or at a higher l e v e l , c i r c l e the number "1". Please read the following statements and for each one indicate whether or not you can usually make decisions i n these areas without reference to a superior. C i r c l e one Decision My l e v e l i n the department i s : c i r c l e one 1 2 3 4 5 6 7 8 9 10 3. Routineness of manager's work Read the following s i x statements and for each statement assess the degree to which you think i t applies to the work of managers i n your p o s i t i o n i n your department. C i r c l e one of the numbers provided to i n d i c a t e your assessment from: strongly disagree (1) to strongly agree (7) ( i ) There i s something d i f f e r e n t to do every day c i r c l e one Strongly ( ! 2 3 4 5 6 7 ) S t r o n « l y Disagree Agree ( i i ) Managers i n my p o s i t i o n do the same job i n the same way every day c i r c l e one Strongly ( x 2 3 4 5 6 7 ) S t r o n 8 l y Disagree Agree ( i i i ) In my company we need to learn more than one job c i r c l e one Strongly ( x 2 3 4 5 6 7 > Strongly Disagree Agree (iv) The same steps must be followed i n processing every piece of work c i r c l e one Strongly ( x 2 3 4 5 6 7 ) s " « 8 l y Disagree Agree (v) For almost every job a manager i n my p o s i t i o n does there i s something new happening almost every day c i r c l e one Strongly ( x 2 3 4 5 6 7 ) Strongly Disagree Agree (vi) The work of a manager i n my p o s i t i o n i s very routine c i r c l e one Strongly ( x 2 3 4 5 6 7 ) ^ c n g l y Disagree Agree YES/NO The promotion of lowest-level super-v i s o r s YES/NO Promotion of nonsupervisory s t a f f YES/NO D i s c i p l i n i n g of lowest-level supervisor YES/NO D i s c i p l i n i n g of nonsupervisory s t a f f YES/NO Decisions about whether or not people work overtime YES/NO Procedures used i n personnel s e l e c t i o n YES/NO Determination of the number of lowest-l e v e l supervisory positions YES/NO Determination of the number of nonsupervisory p o s i t i o n s YES/NO Decisions about budget a l l o c a t i o n s YES/NO Determination of the budget f o r your department YES/NO Determination of new programs and a c t i v i t i e s YES/NO Determination of new objectives and projects YES/NO Creation of new sections or departments YES/NO Creation of new positions YES/NO Handling of p u b l i c r e l a t i o n s outside the company YES/NO Giving o f f i c i a l information to someone or a group outside the company YES/NO Choosing suppliers f o r materials YES/NO Decisions about accounting procedures ./8 - 205 -5. R e t r i e v a l Technology 6. A t t i t u d e s towards data sharing In t h i s part we would l i k e you to assess what types of technology are most frequently used by managers i n your p o s i t i o n i n obtaining information to do with t h e i r job ( c o n f i d e n t i a l or non-confidential information). The following statements r e f e r to the d i r e c t use of a p a r t i c u l a r r e t r i e v a l tech-nology. I f managers t y p i c a l l y use t h e i r s e c r e t a r i e s to r e t r i e v e computer stored data then e f f e c t i v e l y the managers obtained the data through the i n d i r e c t use of technology. The managers' d i r e c t use was th e i r s e c r e t a r i e s ( i . e . they used no technology to r e t r i e v e the information). Read the four statements below and for each statement c i r c l e one of the numbers pro-vided to in d i c a t e your assessment from: strongly disagree (1) to strongly agree (7). (1) ( i i ) In t h i s department, a manager i n my posi-t i o n often uses computer technology (e.g. computer terminals) to r e t r i e v e job-r e l a t e d information c i r c l e one Strongly ( x 2 3 4 5 6 7 > Strongly Disagree Agree In t h i s department, a manager i n my posi -t i o n often uses manual technology (e.g. mailing system, manual f i l i n g system, jo u r n a l s , newspapers) to r e t r i e v e job-r e l a t e d information Strongly Disagree c i r c l e one ( 1 2 3 4 5 6 7 ) Strongly Agree ( i i i ) In t h i s department, a manager i n my posi-t i o n often uses electromechanical tech-nology (e.g. telephones, conference c a l l s , microfilm) to obtain job-related information c i r c l e one Strongly Strongly Disagree Agree (iv ) In t h i s department, a manager i n my posi-t i o n often uses no technology (face-to-face personal contacts, meetings, use of s e c r e t a r i a l or cl e r k s ) to obtain job-r e l a t e d information Strongly Disagree c i r c l e one 1 2 3 4 5 6 7 ) Strongly Agree Estimate the percentage frequency of d i r e c t use fo r the above categories of technology for job-r e l a t e d information (to nearest 10%) Computer Technology % Manual Technology % Electromechanical Technology. . . % No Technology _. % ,100% We would also l i k e your view on how easy i t i s generally to obtain information of any type i n your department. Read the following f i v e statements and fo r each statement c i r c l e one of the numbers provided to i n d i c a t e your perception of what you think the a t t i t u d e towards data sharing i s i n your depart-ment from: strongly disagree (1) to strongly agree (7). ( i ) I t i s normal i n t h i s department f o r people to share information with one another c i r c l e one Strongly ( x 2 3 4 5 6 7 ) Strongly Disagree Agree ( i i ) The information a manager i n my p o s i t i o n receives i s often inaccurate c i r c l e one Strongly 1 2 4 y Strongly Disagree Agree ( i i i ) I t i s often necessary f o r a manager i n my p o s i t i o n to go back and check the accuracy of the information he has received c i r c l e one Strongly ( x 2 3 4 5 6 7 ) Strongly Disagree Agree (iv) (v) I t i s easy to t a l k openly with most mem-bers of th i s department c i r c l e one Strongly ( ! 2 3 4 5 6 7 ) Strongly Disagree Agree I t i s easy to ask advice from most mem-bers of t h i s department c i r c l e one Strongly ( x 2 3 4 5 6 7 ) Strongly Disagree Agree 7. Maturity of the Information System Estimate the number of years that computers have been used i n your company for t e c h n i c a l or administrative purposes: Tick one Up to 5 years (specify) _ 6-10 yrs 11-15 yrs 16-20 yrs 21 or more years (specify) ./9 - 206 -7. Maturity of the Information System (cont'd) ( v i i ) Indicate with a t i c k your type of company Estimate the number of years since most of the o r i g i n a l problems of computer usage were overcome and the computer was accepted into the normal procedures of your company. Tick one S t i l l major problems Up to 5 years (specify) _ 6-10 yrs 11-15 yrs 16-20 yrs 21 or more years (specify) Manufacturing Service Government Educational Voluntary D i s t r i b u t o r P u b l i c U t i l i t y Other (specify) Forestry Mining F i n a n c i a l Realty Engineering Department Store F i s h e r i e s Thank you f o r your time. I f you have any further comments on access to information not covered by the questionnaire, please use the space below. 8. Biographical D e t a i l s In t h i s l a s t part of Section IV we ask you to supply us with some information about yo u r s e l f . These questions allow us to com-pare your data with those of other respondents. NOTE: Your answers to t h i s questionnaire w i l l not be seen by anyone except the researchers. Please do not sign the questionnaire. (i ) What i s your present age i n years? ( i i ) What i s your sex? Male Female ( i i i ) Indicate your highest attained l e v e l of formal education some high school high school graduation some college college degree some graduate study advanced degree (iv) (v) (vi) Job T i t l e Number of years i n present p o s i t i o n Indicate with a t i c k your present depart-mental a f f i l i a t i o n Marketing Sales Finance Accounting Purchasing Personnel Customer Service Engineering Labour Relations General Administration Building Real Estate Law A c t u a r i a l Medical Computers/D .P. Transportation Other (specify) - .207 -T H E UNIVERSITY OF BRITISH COLUMBIA FACULTY OF COMMERCE AND BUSINESS ADMINISTRATION Ju l y 22, 1980 Dear : We would l i k e to i n v i t e you to p a r t i c i p a t e i n a research study we are conducting at the Faculty of Commerce. As you can see from the enclosed questionnaire, we are looking at how a manager such as yourself sees access to information within his/her company. Your answers to the questions w i l l help us to understand how and why access to information varies across departments and companies. Your name i s one of only 200 that has been selected from a l i s t of managers i n B r i t i s h Columbia. As our sample i s designed to include a v a r i e t y of companies and departments, we would very much appreciate your responses as they w i l l help to provide a comprehensive p i c t u r e of access to information. The members of the research team are the only people who w i l l see your responses. Any reports produced from the study w i l l be i n an aggregated form i n which there w i l l be no way of i d e n t i f y i n g an i n d i v i d u a l or a company. The questionnaire takes between 25-35 minutes to complete and a return envelope i s enclosed f o r your convenience. We would l i k e to send you the f i n a l report comparing access across companies. Please include t h i s l e t t e r with the completed questionnaire i f you would l i k e to receive a copy. Thank you for your cooperation. I l a n Vertinsky Michael Newman Professor and Chairman, P o l i c y D i v i s i o n Project D i r e c t o r , Access Faculty of Commerce and Study Business Administration #203 - 2053 MAIN MALL, UNIVERSITY CAMPUS, VANCOUVER, B.C., CANADA V6T 1Y8 - 208 - EXECUTIVE > PROGRAMMES Facul ty of C o m m e r c e and Bus iness Admin i s t ra t ion R o o m 103 - 2053 M a i n Ma l l T h e Univers i ty of Br i t ish C o l u m b i a July 30, 1980 Dear As you know, Executive Programmes works closely with the Academic Divisions within the Faculty of Commerce and Business Administration. This close working relationship results in challenging and informative seminars delivered by highly competent Faculty. From time to time, Executive Programmes is approached by one of our Faculty to co-operate in another important aspect of executive education — doing research. Attached to this letter you w i l l find a questionnaire that relates to such a piece of research. Dr. Vertinsky and Mr. Newman are investigating the question of managerial access to company information. On reviewing this research, I f e l t that the study might be of interest to many managers who have participated in Executive Programmes seminars. Therefore, I have authorized the selection of approximately 300 managers from our mailing l i s t as potential participants in the study. I hope you agree with me that the subject under investigation is important to the management process and on that basis w i l l agree to co-operate in providing the information requested. The 300 managers selected have been chosen to represent a variety of organizations in order to provide a comprehensive picture of the managerial problems in access to information. The researchers have given me their assurances that they w i l l be the only people to see your individual responses (as you w i l l note on the enclosed return envelope, the completed questionnaire is returned directly to the research group). The researchers further assure me that any reports produced from this study w i l l contain only aggregate data in which there w i l l be no means of identifying individual respondants or even individual companies. I am informed that the enclosed questionnaire takes about 25 to 35 minutes to complete depending upon the nature of your individual situation with respect to access to managerial information. As implied above a - 210 -Appendix II Barriers to Access (Medians) Case 1 Case 2 Case .3 Case 4 Case 5 Case 6 Case 7 D i f f i cu l t Procedures 1.688 3.385 1.348 1.447 1.465 1.333 4.833 Embarrassment 1.375 3.192 1.305 1.682 1.673 1.458 4.571 Permission of Superior 1.632 3.136 1.203 1.355 1.262 1.187 5.125 Errors 1.374 1.350 1.444 1.435 1.284 1.303 1.405 Comparison of Di f f icu l t ies 1.925 2.143 1.992 1.900 1.364 1.415 2.429 Location Problems 3.083 4.083 . 2.383 3.596 4.025 3.500_ 6.000. Ignorance of Existence 5.344 4.850 3.227 3.620 4.417 4.024 5.792 Missing Details 1.976 1.850 1.776 1.760 1.595 1.500 -1.944 Biased 3.516 3.375 3.150 3.222 2.773 2.955 3.300 Presentation Problems 4.615 5.000 4.150 4.367 4.538 4.607 4.412 Layout Problems 2.069 2.176 1.696 1.711 2.265 2.222 2.500 Jargon 1.438 1.500 1.523 1.806 1.476 1.464 1.719 Authority Failure 1.465 2.269 1.303 1.479 1.431 1.408 3.222 Timing 2.893 3.667 2.556 3.357 2.636 2.250 4.455 Irrelevant Details 2.611 2.682 2.337 2.429 3.139 2.864 3.633 Cost ($) 2.426 2.875 2.842 2.921 1.897 1.850 3.536 Appendix III Intercorrelations of Barriers for Case 3 (signif icant at p < .05) i | 2 | 3 1 ^ 1 5 1 ^ | 7 1 8 | 9 1 1 0 1 1 1 ' 1 1 2 1 1 S f 1 4 1 1 5 1 16 1. D i f f i cu l t Procedures .233 .453 .456 .399 .275 .167 .415 .121 .207 .287 .234 .428 .216 1 1 .139 .166 2. Embarrass-ment .250 .229 .197 .162 .126 .233 .180 .227 .165 .165 3. Permission .341 .315 .268 .230 .266 .239 .157 . 209 .168 .470 .255 4. Errors .526 .170 .550 .164 .412 .294 .307 .271 .233 5. Comparison .196 .205 .496 .450 .345 .404 .243 .171 .258 6. Location .383 .249 .171 .275 .192 .220 .232 .313 .165 7. Ignorance .189 .160 .204 .150 .217 .250 8. Missing details .146 .505 .412 .367 .240 .167 .266 9. Bias .148 .151 .166 .258 10. Presenta-tion .139 11. Layout .507 .262 .198 .197 .197 12. Jargon .256 .193 .246 .189 13. Authority fai lure .238 .162 14. Timing .279 15. Irrelevant details 16. Cost ($) - 2 1 2 -Appendix IV Kendall's- Coefficient of Concordance Variable W Significance Decentralization .219 .000 Routineness of Technology .030 .000 Trust .356 .000 Retrieval : Case 1 .136 .000 Case 2 .054 .000 Case 3 .161 .000 Case 4 .123 .000 Case 5 .140 - .000 Case 6 .147 .000 Case 7 .077 .000 Use: Case 1 .154 .000 Case 2 .184 .000 Case 3 .106 .000 Case 4 .125 .000 Case 5 .243 .000 Case 6 .256 .000 Case 7 .213 .000 Case 5: Authority .077 .000 Case 6: Authority .051 .000 Case 7: Authority .079 .000 "A high or signif icant value of W may be interpreted as meaning that the observers or judges are applying essentially the same standard in ranking the N objects under study" (Siegel, 1956:237). - 213 -Appendix V Access and the Barriers to Access (Kendall correlations) Case 1: Confidential information Case 2: Confidential information not required required R 1. Location -.3957 R 1. Permission of superior -.4272 R 2. Timing -.3652 R 2. D i f f i cu l t procedures -.4059 R 3. D i f f i cu l t pro- U 3. Presentation -.3241 cedures -.3468 R 4. Location -.2770 R 4. Authority fa i lure -.3262 R 5. Timing -.2180 R 5. Embarrassment -.3165 R 6. Embarrassment -.2127 U 6. Missing details -.2180 R 7. Existence not known -.2026 R 7. Permission of R 8. Authority fai lure -.2003 superior -.2134 U 8. Errors -.2081 Case 3: Non confidential , requi red R 1. Di f f i cu l t pro-cedures -.4093 R 2. Location -.4021 R 3. Timi ng -.3830 R 4. Existence not known -.3509 U 5. Missing Details -.3457 U 6. Jargon -.3288 U 7. Comparison -.3255 R 8. Authority fai lure -.2911 Case 4: Non confidential , not required R 1. Di f f i cu l t procedures -.4532 R 2. Location -.4007 R 3. Permission of superior -.3795 R 4. Existence not known -.3688 R 5. Timing -.3329 U 6. Missing details -.3237 R 7. Authority fa i lure -.3222 U 8. Comparison -.3067 Case 5: Own personal details R 1. Timing .3892 R 2. Embarrassment .3137 R 3. D i f f i cu l t pro-cedures .2983 R 4. Authority fai lure .2655 R 5. Permission of superior .2401 U 6. Bias .2170 R 7. Location .2143 •R 8. Costly .1743 Case 6 : Subordinates' personal detai1s R 1. Timing .3709 R 2. Authority fa i lure .3607 R 3. D i f f i cu l t procedures .3194 R 4. Location .2792 R 5. Embarrassment .2608 U 6. Comparison .2141 R 7. Permission of superior - .2064 U 8. Missing details .1994 - 214 -Appendix V (cont'd) Case 7: Other managers' details R 1. Permission of superior -.4659 R 2. D i f f i c u l t procedures -.3771 R 3. Embarrassment -.3598 R 4. Location -.2880 R 5. Failure of authority -.2773 R 6. Timing -.1934 "R" i s used to indicate a barrier that s i g n i f i e s r e t r i e v a l while "U" s i g n i f i e s the use of the information. - 215 -Appendix VI Techniques to Prevent Access There were several techniques employed by the companies in the interview sample to regulate access directly and indirect ly . In some cases the techniques were only used to prevent unauthorized access to con-f idential information. In other situations the control would apply to both confidential and non confidential information. Techniques used to regulate access are accompanied by certain benefits and costs to the com-panies. These are discussed for each technique. 1. Direct regulation (i) Directives from management Three cases were found in which management's policy was to issue directives concerning the access to confidential information. One manager told employees that what they see and hear while at work has to remain in the building. They are also required to sign a waiver that notice has to be given i f they wish to perform transactions related to the busi-ness. This is done in order to prevent a conf l ict of interest arising where an employee might exploit their inside knowledge or early notice of materially valuable information. In another, privately held company, the staff are instructed on arrival that because of the special status of private companies they are not allowed to disclose company information to outsiders, particularly where i t concerns financial detai ls . In two other cases, top management issued statements to the effect that no manager could discuss company information with the public or press. Man-agers in these cases are required to channel requests from the public or press for information to the public relations group (see below). - 216 -This method is an example of companies making their policies on access to information quite expl ic i t for employees. It was not clear how effective i t was in preventing leaks of confidential information. In one case the policy was reinforced by discipl inary measures. ( i i ) Channeling of external statements In six of the companies visited a policy was in effect that restricted public statements to one department, usually public relations. In at least one organization this was clearly done in order to present a consistent voice to the outside world. Five incidents were related to the researchers demonstrating that confl ict ing statements had been made in the past lead-ing to an embarrassing situation for top management. Both newspaper reporters and pol i t ic ians had telephoned employees (not always managers) direct ly to seek internal information or opinions. In one crown corpora-t ion , most of the attempts to stop leaks of internal information fa i led apparently because of the pol i t i ca l a f f i l i a t ions of some of the employees who were wi11ing to leak unfavourable information. Using a single voice to broadcast company information to the press and public can only be effect ive, of course, i f the employees are made aware of the policy normally through a directive issued from above (see (i) for a d is -cussion of this point). ( i i i ) Use of sanctions In companies, the policy of access to confidential information can be used to convey a reward or a punishment to managers. In other s i tua-tions threats can be used against employees to deter them from breaking the rules of access. In one company the manager seemed to be in possession of a special - 217 -report that other managers at a similar level did not receive. The impl i -cation seemed to be that his access to this report was granted as a special favour from the managers above (the executives in this case). This impression was reinforced by the fact that some of the other managers thought he should not receive i t . In the same company those workers who have at least twenty five years service are part of a special "club". The club is occasionally addressed by the president who reveals to them some of the plans of the company, information that would not normally be available to other employees. Another subject interviewed related that the company she worked in emphasizes loyalty in i ts employees. When an employee resigns they are normally asked to leave immediately, partly to prevent leaking of informa-tion to competitors, particularly i f the employee is moving within the industry. After one employee refused a move to another city the person was ostracized by the company. The employee was le f t off mailing l i s t s and memos were not answered. Although the employee was not f i red and the problem turned out to be a misunderstanding, i t demonstrates the effectiveness of using access to information as a deliberate weapon for punishing an employee. Examples are also available in the l i terature on the use of organizational gatekeepers who control the access of informa-tion for others (e.g. Pettigrew, 1973). Lastly, an example was given to the researchers where employees were threatened by management with discipl inary action i f they were caught accessing information outside their functional area. This applied to off ice workers in the case cited and the information was largely com-puterized and operational. No examples were given where managers had been threatened with similar sanctions i f they strayed outside their functional - 218 -areas and accessed information that was not needed for their jobs. (iv) Special markings Eight managers in f ive companies in the sample gave examples of the use of special markings when confidential information was being sent from one part of the company to another. Usually the documents would be sent in sealed envelopes stamped "private and confidential" or just simply "con-f ident i a l " . In one case, different coloured folders were used by a com-pany to distinguish between publicly available information and internal , confidential information. Although the practice of putting special markings on documents is not very expensive i t does have at least two drawbacks. F i r s t l y , i t tends to be abused by those who control i t . Three cases were cited where.the managers thought that too much information was stamped confidential and that much of the information was in practice readily available in the company. This leads to the term "confidential" being devalued. Secondly, stamping documents in special ways tends to draw attention to their potential value rather l ike postal thieves are attracted to registered mail. 2. Indirect regulation Retrieval (i) Keeping managers ignorant A manager's lack of knowledge about information available in the organization is a natural by-product of the "need to know" formula for determining access. For example, in one company, information about promo-tional items required two to three months of confidential ity during which time the discussions were confined to executives and the marketing managers on the basis of their "need to know". The material was kept in the heads - 219 -of those involved and not recorded on paper. Other managers and employees were formally excluded from the discussion and did not know that the meetings were taking place. In another company one group of managers in a functional area had their salaries tied to those of the unionized maintenance workers. This resulted in generally higher salary levels compared with other managers. The policy was not well known within the company. Another effect of the "need to know" rule concerning access is that managers may not be aware of what others are doing in the company. This can result in a duplication of e f for t , a situation recounted by two managers in different companies. Three cases were found in companies where the policy of ignorance was applied by managers to their subordinates and to unions. In one industrial board, reorganization plans were kept from the employees until the last moment. This was done to reduce the speculation and anxiety over redundancies that might follow i f an early announcement was made. Other cases involved unions in two of the com-panies. In the accounting department of one of these companies, i t was standard policy to keep cost accounting information from the union i f the information concerned the subcontracting of non-union staff . Clearly, many companies use ignorance as a method of restr ict ing access, intentionally or otherwise. Ten of the managers sampled gave expl ic i t examples of i ts use. For confidential information i t forms a cheap means of security as long as trusted people are involved. If the information is not documented but is retained and verbally communicated by those entrusted with i t , then the risk of unauthorized access is further reduced. When the information requires documentation at an early stage, as is the case for complex projects or studies, for example, then the risk of leakage is higher. Documents tend to be copied and leaked. - 220 -Several examples of this latter situation were related to the researchers, one of which ended with a company being picketed by the workers after a consultant's report on staffing was leaked to the union. However, i f the method of keeping managers and others "in the dark", as one manager put i t , is applied to non-confidential information then the result may be the duplication of effort that was found in several situations. Additionally, i t may result in the promotion of a negative attitude to information sharing. ( i i ) Non-recorded information Nine cases were cited by managers where information was not recorded in document form. In each case the information was of a confidential nature. For instance, one manager noted that most information of a con-f idential nature is given in his company by word of mouth only. The executives had learned through experience not to put confidential infor-mation on paper. In another company, policy information was afforded the same treatment in that i t was not committed to paper but was retained in the heads of those few people concerned with the discussion. For low volume information, verbal access to data would seem par-t icu lar ly effective in the treatment of highly confidential information for short time periods. The method does not require elaborate security as the information is not kept in document or computerized form. It would seem to find particular application in the preliminary discussion of strategic planning questions or promotional issues, and situations related to the researchers verify th is . Of course, the lack of docu-ments concerning policies could be disconcerting to those middle managers seeking direction from top management, but that is another issue. - 221 -The method of not recording information breaks down under several related conditions. F i r s t l y , the volume of information could be so high or the content so complex that the people concerned could not reasonably be expected to r e c a l l the details with any accuracy without the support of documentary evidence. Secondly, the time of discussion could be s u f f i c i e n t to preclude the sole reliance on the memories of the people concerned. Lastly, the project may involve relationships outside the company that cannot be t o t a l l y controlled. In one very large company the time span of any project was measured in years, the issues were highly complex, and they had to keep several diverse external groups supplied with information about the project. Clearly, in these circum-stances the company had to document the project extensively in order to meet these c r i t e r i a alone, besides the more technical ones. ( i i i ) Visual access to information Only one example of this method was found during the interviews. As a way of protecting c o n f i d e n t i a l , five-year plans from unnecessary copying, the executives circulated i t to middle management with the i n -structions that no copying was allowed. This i s an indirect way of r e s t r i c t i n g access to confidential i n -formation, employing the assumption that the fewer copies made the less l i k e l y unauthorized access w i l l occur. It i s an effective method of keeping managers informed of highly confidential information. At the same time i t demonstrates a certain trust in managers not to abuse their access to highly confidential information. It also implies a higher status for those on the c i r c u l a t i o n l i s t . - 222 -(iv) Physical security The use of physical security to protect confidential and non-confi-dential information varies not only between companies but between d i f fe r -ent functions within companies. For vis i tors to the companies the physical barriers faced in seeing personnel ranged from high security, involving guards, signing i n , and the issuing of passes, to no security whatsoever. In some cases the high degree of physical security was associated with the value of the / equipment being maintained and in other cases i t resulted from the value of the information being kept in the building. Inside the companies there were usually one or more areas to which special physical security was applied. In one company v i s i ted , the advertising department was a "secure" area with a sign warning that only authorized staff were allowed to enter. This restr ict ion on entry was established as a direct result of information leaks to competitors. In the same company, the research department had been moved to another loca-tion outside the main building in order that special security could be applied to the information contained within the group. Another special area of high security was the computer operations room in many of the companies v is i ted. Security was usually accomplished by the use of keys and/or magnetic badges. In one insta l la t ion, staff wore badges with different coloured dots to distinguish between the different areas of security within the computer department. It is standard practice in most computer departments to keep copies of important data f i l es at a separate location. In the case of theft, f i re or other damage, the f i l es can be recovered. In this way, many computer instal lations are better protected than those using manualt records for their operational data. - 223 -As an i l lustrat ion of the contrast, an industrial board maintains manual records for part of its operational information on c l ients. There is no protection from loss in the case of f i re or theft. No back-up copies are taken of current records. Most of the physical security found in companies applies to the bulk of operational information, including some of a confidential nature. However, as most of i t is not confidential , the security is largely to protect the company from a total or partial loss of records which might occur in the case of a computer room f i r e , for example. In the case of most management or executive of f ices , where much of the information used is confidential , however, the security is normally much less rigorous. In several cases in the study confidential information was clearly le f t on managers' desks and sometimes was not even locked away at night. Although executives were not part of the study, they too would seem to have a similar attitude towards information security. However, although their offices are not usually protected by security systems, executives are normally located in a special suite of rooms with a separate secretary to screen v is i tors . In this way they have better protection from theft of information than most middle managers. Physical security of information can be a very expensive form of protection and most companies visited have seen f i t to apply i t to specia l , largely operational, areas of information handling only. Gen-eral ly computer records are better protected from unauthorized access than manual records. It was found in the study that most information used by managers did not have great security applied to i t and where security was used i t was in the form of locked drawers or off ice doors. - 224 -It is not clear that most managers would be wi l l ing to change the way they handle information, in which case other methods of protecting confidential information from unauthorized access might need to be examined. (v) Di f f icul ty of retrieval One way organizations can make confidential information secure from unauthorized access is to ensure that i t is d i f f i cu l t to obtain, and several examples of this were given during the interviews. Some of these problems of retrieval follow from the level of security applied to certain kinds of data. The need for employees to surmount physical obstacles to gain access to data is an effective barrier for most workers. In other cases, the information was not necessarily well secured; i t was just held in a location at a distance that ensured limited access. One company had set up a separate f irm, off premises, partly to restr ict employees from gaining access to confidential infor-mation. In a different example, the confidential information was held local ly in the accounting department without much security but i t was so scattered throughout the department that the physical d i f f i cu l t y of assembling al l of the pieces would prevent unauthorized persons from accessing i t . This is an example of what someone has called adminis-trative ineff ic iencies and their impact upon access (Canadian Task Force, 1972). Seven other examples given by managers concerned their need to just i fy to other employees why they should have the information. Often the information was confidential. In one company the board minutes are kept in the president's of f ice . In order to get them the manager had - 225 -to j u s t i f y his need for them. If others do not recognize managers' authority to access the information then they may have to c a l l on a higher authority to obtain them and three examples of this were related to the researchers. In one situation the manager was refused access to technical information on the grounds that he might misinterpret i t . He had to negotiate with those who held the information u n t i l a compromise solution was found. In the computer department of one company, one person was given charge of the administration of passwords. When an employee requests a password, the request i s v e r i f i e d by the person in charge to make sure that the employee's need is legitimate. In a l l of these cases the need to j u s t i f y a request for information was an effec-tiv e barrier to unauthorized access. But in any company the method can be taken to extremes. For example, one manager had to j u s t i f y to an employee why he should have access to f i v e years of f i n a n c i a l data even though the same information was publicly available. The creation of inconvenient procedures of access i s usually not d i f f i c u l t to implement and from the comments made by managers they seemed to be quite effective in preventing access to information. In companies where the need to protect confidential information i s acute the limited use of physical barriers or, a l t e r n a t i v e l y , "gatekeepers" who administer such information, seems j u s t i f i e d . However, for infor-mation that i s non-confidential, erecting unnecessary barriers and making i t awkward and time-consuming to retrieve needed information could seriously i n h i b i t a manager's performance. In addition, the need to j u s t i f y a request for access to another employee i s subject to the vagaries of the employee's interpretation of the manager's rights of access. A widespread use of this method in a company to control access - 226 -to non confidential information would be an indication of a poor a t t i -tude to information access and the sharing of data. (vi) The use of technology to restr ict access patterns Almost al l of the companies studied had been using computer technology to store and process vast quantities of operational data for several years. Seven of the companies used systems that have built-in f a c i l i t i e s for restr ict ing certain f i l es and programs from unauthorized use, employing special terminals, passwords, or both. These companies were exploiting the new capability by using i t to keep employees to data that they "need to know". This method of protecting information, confidential or non-confi-dent ia l , is a by-product of the technological f a c i l i t i e s available with modern computer systems. For confidential or sensitive information, such as payrol l , the method is an effective means of protecting tt from unauthorized access as long as the administration of passwords etc. is also carefully handled. For non-confidential operational information the enforced access patterns could also be helpful in overcoming some fears expressed by different departments concerning the ownership of data. Some managers were reluctant to pool information with other departments: i f the system could support the traditional patterns of access some of this reluctance may be overcome. (vi i ) Timing of release In four of the companies, policies were used which governed the timing of release of confidential or sensitive information. In some cases i t was done to maintain a competitive advantage; in other cases i t was used for organizational reasons. - 227 -An example of the f i r s t kind was found in one company where adver-t is ing details were considered confidential but were finding their way into the competition. In order to combat this leakage the timing of re-lease of the advertisement was lef t until the last moment. The same was true in the transportation industry where schedules of departure times were not released until i t was absolutely necessary. One company in -volved in property deals only makes the information available to managers not involved after the contract has been agreed upon. Two examples of the second type were found where companies would keep reorganization plans from those to be affected until the last moment. The judicious use of timing of release is a very effective way of protecting some confidential information from unauthorized access. For certain types of information where an important decision is to occur or a deadline has been set, i t can defuse some of the problems associated with the early release of confidential information while s t i l l allowing interested parties to be informed of the post-decision results. Use (i) Coding of information Four instances were recorded where confidential information was coded. In some cases this was a deliberate policy of the company to pro-tect confidential information from being leaked. In other cases i t appeared to be more of a by-product of the information and not part of the company's policy. An example of the f i r s t kind was given in a company which had acquired land for future development of a commercial enterprise. As the item was an asset of the company i t had to be shown in the accounting - 228 -f igures, but in order to prevent the competition from finding out the location of the s i te , i t was given a code that is known only to a few senior managers in the company. The second use of coding was found at a branch off ice of a large company. Planning information was not given any special security by the manager concerned even though i t was conf i -dential as he believed that i t would require about ten years experience to read and understand the information. He was the only one in the off ice who had that experience. Coding is a cheap and generally effective method of protecting con-f idential information. It does not require special security and i t can be used for high volumes of data. It has the potential disadvantage that should the information find i ts way into the possession of unauthorized persons i t could be misinterpreted. In practice i t would seem that few companies employ coding as a deliberate policy to protect confidential information. (x) Use of summarized or interpreted information In order to protect confidential information, companies or depart-ments within companies sometimes employed summarized or interpreted infor-mation. The detailed information is retained by the originator, and the recipient gets a summarized or interpreted version. Nine examples of this method were given by different managers. Recorded confidential information can be protected within and outside the company simply by removing some of the sensitive detail thus making i t non-confidential. Minutes of meetings are a typical example of this process. One manager commented that although he receives minutes of board meetings they are so summarized that many of the important (useful) details are missing. The same manager repeated the process when he sent memos to - 229 -his staff to inform them of on-going act iv i t ies raised in the board minutes. Sensitive material was removed before the information was circulated. Often financial documents are summarized before being made readily ava i l -able internally. Because detailed cost and sales figures are confidential in many companies they are often condensed to the next or higher levels of generality. An example of this was found within a large company where one division was limited to the access of summary costing information of another div is ion. In another large company broad cost figures were widely available internally. In fact they were shared with their main competitor. However, detailed cost figures were neither available to the competition nor to the majority of managers internal ly. In a health care organiza-t ion , the budget was often unknown well into the f iscal year. This was one reason that led the finance department to be very cautious about releasing detailed costing statements internally. As another manager in the same organization put i t ; the finance department practices selec-tive dissemination of information in summarized form. In another case, a manager requested from another department a quotation for a particular service. The manager was given the quotation but he was not allowed access to the detailed methodology by which the quotation was determined, ostensibly because the department thought he might misinterpret the pro-cedure. The use of summaries and interpretations is a widely used method of protecting confidential information and some examples have been presented. For non-confidential information the method is also applicable in reducing the volume of documents sent to a manager (e.g. exception reporting). However, the technique does have some drawbacks as Sorter (1969) has indi-- 230 -cated. One of the main disadvantages of the method is that summarized or interpreted information is prone to misinterpretations. As one manager interviewed noted, the problem in his company was not retrieving infor-mation, but interpreting i t .
UBC Theses and Dissertations
Managerial access to information Newman, Michael 1980
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.
- 831-UBC_1981_A1 N49.pdf [ 10.71MB ]
- JSON: 831-1.0095306.json
- JSON-LD: 831-1.0095306-ld.json
- RDF/XML (Pretty): 831-1.0095306-rdf.xml
- RDF/JSON: 831-1.0095306-rdf.json
- Turtle: 831-1.0095306-turtle.txt
- N-Triples: 831-1.0095306-rdf-ntriples.txt
- Original Record: 831-1.0095306-source.json
- Full Text
Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url: