1.1111111 I, IIIIIIIII 1‘11 1“ II 111111111“ 11111 11m 11111111111111 1111 III a 11111 "1111111111” 1 111:1 11111I I11111II11II111 I1 1711111111111“ 111.11.1‘111II1 11: IIIIfiINIm“ 1111 11111111111: mgk 1IIIWSI1111I1121111 d11I1III1 1;. 1111 1; 11m 2: 131' e. ' 1111111111.. I11 1'1'1111111111111 I 11' I11111111 11 11 "H11 .1 1111111 11111 11‘. 1 .1 11111311111 11 1 I "I .I1II 111’ II. 11 II 111111111111 11 111111: .I—II '+ . 111111 111 1111,1111 111111 11 I1 1111 111111 11111 111111111311 I 1 111141 111., 1111111 . 1111.111111111. 111111111111, 1111111111111 ‘ 1 11 11 1’ 1 WI 111111111111111.1111l 11111 1111’ 11111111 1111111111“ 11” 111111111}I1 1111.. 1 1 .".11II_: I 1 {1:111 -; m om— Wu. . ’M. mum—d“ . . ‘ F W? - ~ .-_ I -_ *- ’: I >- .1 . --.._ <. .. -_. -. . . . V .0 u. - A .- o. n .mwm .~ 14 ”W .—‘ - .- .. fl-.— l—--' a. Q ' m- - - “I « r n . - v- . - 1 ‘l - .‘- - D 0 > , -... . ' _ O .u ' - a ‘ ‘ *~. . v o - , u - u . . ' '-~m—_nm - _. .I- .. .. ‘. . . _,, - - .n on -. no ». . ,3." ‘- ' ~0— . -- .W. o.-. .. ‘~h- — I ‘ n. I- . -... -c .. go _ .- . A ’ ht- ' ~ ' v ' ..- . _' - 49¢ - . -.- . *., _ I ‘ n ' ' - n... . .- . . w-l - - - ~ I . u . . H o- .‘ 9‘ ‘ 4... _ 4 ' - u 111* 111 111 1111 ._..4- “m—m -— H.- ’37.... 1 1 J:I'1I 12111 111111111 11:11I 111:1 I1IIII1 ”mfg, —- Mum u. _ M_" —< n a n I . o. . - 4— ‘v‘A-m.‘ .. M ”w. .m—v—d‘ ..._~—.—- - . -- . .-.¢...~—. -- ‘ - a”. W-- - v .- — hn't.-. u.-<-V - u. ”a..." w‘ ..—~-n».’~.=.4 . o .—. ._ . " “ -... 1 'IIIII‘I11IILI11T15I1' I1 111 1511 §I1111 1I11 ' 11: 1111111111 .'. ‘1 ”'31 1111'1I1 + +1 1 : ' 1' 111111111 1' '1 ‘ . 1:111 "11111111111111 1.1 11- 111111111 1 t" ‘1. '1 *1 1 ; 11 1111 1 “wt—‘— :2. ‘1 i1; a .‘ .r .1 ¢ ' l' . ‘1‘...“ 'W'oo. , -. . ' —~ < . m ".0 . .3 I : ‘4; IL: . 1. ‘ 1 l“: i ‘ . .1 a b ‘ , o '1' 1.11., ‘. ‘. 1: I ‘1" I} 2 ,1, : ‘ " 11 :11 o' ‘1 ' 1' ‘t ”1' 1111111111 1‘ 1-11 ~':"1' u f ’5 . 1 I: ‘1 '1 -' ‘L: :1 11'. ‘1 w W1 . .311 1’ 1:5; I~ ‘I::: 1 11'EI'11’11‘11I1’ '2 ;. '." 11r' r” i 111111 ’ 1 ' " ' ' " . .19 ' v I 1111- '1v111F31' '1 1 1' '1 4 1 11 1II " . '1. I ' " 1 ' x . 11?, . 1' 1 l .. .I-I‘ a A. A..- '- 111 1111111 -.—. —...- -- . _. gm’ H. 1%."-.. ‘ —‘::.3.‘~‘ u - '_ 4‘. if: ‘6'“: Ilium-9'61!» _.- . ( . .nu.ns~:‘- "H I .1 .. - A. - . ' 1......» __- A a F - t. .q t .<—. “-4, f.- . ~-l".. .- J's..- -‘D-vo -—-——] ‘— I&‘-y— d~u '. fl 3 .- 3.-- .._..?.e:__’ ."——u '1 ‘__‘_._£."' This is to certify that the dissertation entitled A’QH—(gifg‘ugfi'll FY WORK k-Hb \NbxdibuA—L T—wamwmq m COFlPorU—TE Wmma: THE CM? 0Y- Mkweruem W LUKE—Ur, ZWBIA presentedby Pam pglani mu tum/a has been accepted towards fulfillment Q \N of the requirements for ' \r degreein (0 C—\D\C)3KéY .J .\_____) \ A- Majo professor Date amp”: Hfl fl MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 bV1ESI.] RETURNING MATERIALS: Place in book drop to LIBRARJES remove this checkout from .ggslxgagln. your record. FINES will - . be charged if book is returned after the date stamped below. “”2. ) thwfii 792 2/7 2:,“ A15; ACHIEVEMENTS AT WORK AND INDIVIDUAL FUNCTIONING IN CORPORATE HIERARCHIES: THE CASE OF MANAGERS IN LUSAKA, ZAMBIA BY Pempelani Mufune A DISSERTATION Submitted to Michigan State University in partial fulfillment Of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Sociology 1986 ABSTRACT ACHIEVEMENTS AT WORK AND INDIVIDUAL FUNCTIONING IN CORPORATE HIERARCHIES: THE CASE OF MANAGERS IN LUSAKA, ZAMBIA BY Pempelani Mufune This field study investigates the effects of achievement (primarily rank, promotion, and power) on self- esteem, self-investment (SJLJ in work and work orientations (flexibility, participativeness, and universalism) in corporate hierarchies. We have tested a working model linking achievements to the moderating variables of 3.1. in work and self-esteem and to the dependent work orientation variables. The subjects consisted of 101 managers in parastatal, private, and multinational firms in Lusaka, Zambia. Data was collected through personal interviews completed from March to May, 1985. Random sampling procedures were used to select respondents. We wanted to confirm the relationship between achievements, self-esteem and 8.1. in work. A second purpose was to confirm the relationship between achievement and work-orientations. A third purpose was to confirm the relationship between SJL.in work, self-esteem, and work orientations. Findings of the study were that promotions do not relate to 8.1. in work while rank has a low but positive correlation with SCI. in work. Power is moderately and positively correlated to 8.1. in work. A similar pattern emerges in the relation of achievement variables to self— esteem. Promotions related to orientations in a small and insignificant way. Rank has a major impact on work orientations. Power significantly affected participation but not flexibility nor universalism. These data also confirmed the significant relationships between 8.1. in work and work orientations; however, many of these relationships are moderated by the effects of rank. These relationships seem to hold more for people who have been in their jobs for at least five years. They are also moderated in part by the type of control exercised in firms. Thus powerful managers in firms with a high total amount of control tend to be more participative than their counterparts in firms with medium and/or low total control. Not all achievement factors are of equal import in individual managers' orientation toward work, self-esteem, and 8.1. in work. It may be important to specify the different circumstances under which different achievements become consequential. What affects peoples orientations at one level may not at another level; however, empowerment through increasing total amount of control in the firm may assist managers at different levels to adjust to the firmfs central values. ACKNOWLEDGEMENTS I wish to express my sincere thanks to Professor David Wiley who as Chairman of my guidance committee has provided support, inspiration and skills to my intellectual development as well as the completion of this dissertation. I am particularly thankful to Dr. William Faunce for his gifted ideas and skills which saw me through this dissertation. Although he was not my dissertation advisor he agreed to read and extensively comment on the content of this dissertation. The courses I took with him kindled my interest in the sociology of work and industry. I am thankful to Dr. Jay Artis and to Dr. John Gullahorn for being on my committee and for their commentary from which I have greatly benefited. I sincerely thank Carey Draeger for her skillful typing of this dissertation, and Marian Sakala for being my Research Assistant in Zambia. Special thanks go to the Nwankwos, Malambos, Tembos, Mundendes, Malilwes and Ngategizes for their optimism and encouragement. This work would have been difficult to complete without encouragement from my family in Zambia. iv CONTENTS LIST OF TABLES O O O O O O O O O O O O O O O O 0 LIST OF Chapter 1 FIGURES O O O O O O O O O O O O O O O O O 0 Introduction - Research questions and explanations Introduction . . . . . . . . . . . . . . . Research Questions and their explanations . . . . . . . . . . . . . . Research Setting . . . . . . . . . . . . Literature Review Promotional and Rank Effects . . . . . . . Consequences of Self-esteem and 8.1. in Work . . . . . . . . . . . . . . . . . . Power Effects. . . . . . . . . . . . . . Methodological Considerations The Sample . . . . . . . . . . . . . . Problems Encountered During Field Research . . . . . . . . . . . . . . . . Scaling and Analysing Items. .‘. . . . . . Findings Hypotheses Set I . Hypotheses Set II. Hypotheses Set III Hypotheses Set IV. Hypotheses Set V . Summary, Discussion and Conclusion Summary of Problems Encoutered . . . . . . Summary of Findings. . . . . . . . . . Discussion . . . . . . . . . . . . . . APPENDICES A- Letters of introduction to the firms in which interviews were carried out . . . B- a) List of firms in the Zimco directory from which parastatal firms were selected . . . . . . . . . . . . . . b) A typical example of organizational chart in a parastatal firm . . . . C- Interview Schedule. . . . . . . . . D- Additional descriptive statistics . . BIBLIOGRAPHY O O O O O O O O O O O O O O 0 0 O Page iv 11 26 37 66 7O 79 96 101 113 136 145 150 157 172 175 184 193 198 204 205 212 216 LIST 0 Tables 1:1 F TABLES Page Skilled occupations by racial origin in maHUfaCture (1965.766 ) o o o o o o o o o o o o 33 Distribution of labor according to skills in Zambian mines . . . . . . . . . . . . . . 33 Item analysis of self-esteem and 8.1. in work scales. . . . . . . . . . . . . . . . . 106 Item analysis of powerfulness and job security scales. . . . . . . . . . . . . . . 109 Item analysis of flexibility vs rulemindedness and universalism vs favoritism work orientation scales . . . . . 110 Item analysis of participation vs authoritarianism work orientation scale. . . 111 Regression coefficients of promotion effects on work orientation. . . . . . . . . 115 Zero-order correlations for achievement variables, 8.1. in work, Self-esteem, and work orientation variables . . . . . . . . . 116 Regression coefficients of rank effects on work-orientations. . . . . . . . . . . . . . 126 Zero-order correlations for upper level managers 0 O O O O O O O O O O O 0 0 O O O O 131 Zero-order correlations for middle and lower level managers . . . . . . . . . . . . 132 Percentages showing relationship between age and rank . . . . . . . . . . . . 134 Regression coefficients of promotion effects on self-esteem holding for S O I O in work 0 O O O O O 0 O O O 0 O O 0 O o 146 4:8 Zero-order correlations of work orientation variables for managers with 2 years or less in present job . . . . . . . . . . . . . Zero-order correlations of work orientation variables for managers with 3-4 years and with over 5 years in present job . . . . 8.1. in work levels of managers with differing degrees of powerfulness. . . . Correlations of power on 8.1. in work and work orientation variables according to type of control in firm. . . . . . . . . Distribution of responses on five key scales . . . . . . . . . . . . . . . . . Multiple regression models of achievement and work orientation variables . . . . . Correlation of promotion and 8.1. in work according to group of age at given rank. vii 155 156 154 159 212 213 215 LIST OF FIGURES 1a 1b Page Working model of the effect of achievements and power on individual functioning . . . . . . . . 8 Effects of promotion and rank on 8.1. in work. . 119 Effects of promotion and rank on 8.1. in work. . 120 viii CHAPTER 1 INTRODUCTION RESEARCH QUESTIONS AND THEIR EXPLANATIONS Introduction This is a study aimed at determining whether work achievements (isen rank and promotions) and exercising power in the formal hierarchy of work organizations have significant consequences for individual social psychological functioning (iie. work value priorities guiding individuals' choices in the organization). The problem itself is rooted in the larger one of how social structure affects and is affected by an individual's orientations. Put this way, the issue has attracted a lot of attention from a variety of sociologists. The work of Kohn et al. (1983), Kanter (1977), etc., stand in this tradition. The theoretical basis of this study is rooted in the works of many authors who have linked structural aspects of work organizations to the behavioral dispositions of individuals at different levels of the work organization hierarchy. The subjects of this research effort are man- agers. Managers are defined as individuals exercising super- visory functions within hierarchies of work organizations. The diversity of theoretical orientations informing this study becomes clear as we briefly consider the major variables of interest. The primary variables being considered in this study are career achievement (promotions and rank), power (control), self-esteem, self investment in 'work, and work orientations. In the literature, careers have been discussed under a variety of topics and for this reason their defination is often obscure. Thomas (1983), following Hall and Hall (1976), identified five distinct meanings of the concept “career". These are 1. Career as advancement, iJL, vertical mobility — the movement upward within the organizational hierarchy. Directionality is important here, for moving up the hierarchy is normatively judged to be good while moving down is judged as bad. II. Career as profession, ine., certain occupations are deemed as careers by virtue of having identifiable lines of advancement whereas others insofar as they are dead end jobs are not. III. Career as a life long sequence of jobs - that is career is taken to mean an individual's Job history. This definition refers to positions regardless of their level or type, hence to be employed or to have been employed is to have a career. ‘L. Career as a sequence of role-related (including personal) experience - that is an individual's movement in and out of any social (familial, marital etc.) role constitutes a career . V. Career as a lifelong sequence of work attitudes and be- haviors. This conception of career combines the lifelong process of definition III and the subjective experience embodied in definition IV. In this study we take definition III as constituting what we want to capture by the term "career." Following Spilerman (1977), Spenner et al. (1981, 1982), and others, we distinguish between "career" and "career line." "Career refers to an individual’s job history." (Spilerman, 1977, p.551) "Careers are individual level phenomena: a person's employment biography, a particular individual's job history" (Spenner et a1. 1981, p.3). Careers are to be distinguished from career lines by levels of reference. Whereas careers refer to individual—level phenomena, career lines are so- cial-structural in nature. Career lines refer to work histo- ries that are common to a portion of the labor force. That is, career lines are determinable regularities in job se- quences that a group of people follow over time. They are age graded jobs that are common to a given portion of the labor force. In this study our concern is with career rather than career line. More specifically, we limit ourselves to career mobility. By career mobility we are mainly referring to promotions a given individual has experienced within and outside his or her present work organization. Our focus is on promotions because we believe they lie at the intersection of cognitive and motivational behaviour of those in work organizations. The factors we see as important in our conception of careers and promotions include: length of stay in a given job; the age at which one is in a given job; the hierarchical level at which ones job is; the amount of lateral transfers the individual experiences; the number of demotions the individual is subjected to; and the amount of job security or insecurity one is subjected to. Apart from career and promotions, another independent variable we consider is that of power; Power is the "ability to get things done, to mobilize resources, to get and use whatever it is that a person needs for the goals he or she is attempting to meetJ'(Kanter, 1977,1L166) Put in this way, the concept of power may be treated both at the level of system (i.e., organization) and at the level of the individual. With regard to the treatment of power at the level of the individual, we are informed by the writings of Parsons (1960) Kanter (1977), and Hamilton and Biggart (1985) among others. Our conception of power at the level of the organization is mainly derived from the extensive sur- veys made by Tannenbaum and associates (1968,1974), who conceptualize individuals as differing in the amount of power they exercise within organizations; however, the total amount of power within organizations differs from firm to firm. We are precluding a zero—sum interpretation of power. The work of Faunce (1982, 1984) informs our conception and treatment of self investment (85L) in work and of self- esteem. Self-esteem refers to personal judgements of worthiness expressed in attitudes which individuals hold toward themselves. Self investment in work refers to a commitment towards work based upon work's relevance for self-esteem (Faunce 1984).£LI. in work is related to such concepts as central life interest (Dubin, 1956, 1974) and job involvement (Lodahl and Kejner, 1965). 8.1. in work can be seen as a process varying over time. The individual may be said to enter the organization with a concern for achievement (herein talked of in terms of promotion and its accompaniments of status and power). People typically enter the labor force with relatively high 8.1. in work, which, depending on their work experiences, either declines or is enhanced. As Faunce (1984) says the system for assigning status which reflects an individual's recognized achievements in a given work organization is crucial to the process of'S.I..in work. It is not just the knowledge of one's status (which reflects success or failure) but the frequency of being reminded of that status which is important to the process of investing oneself in work. As a sidenote, it is important to remind ourselves that for managers it is very difficult to avoid being reminded of one's failures and successes (sometimes seen in terms of status) in the work organizations. The type of office a manager has, the number of people at his or her command, the number of people able to command him or her, the fringe benefits, etc.,la11 serve to remind the manager of his or her achievements. 8.1. in work is, however, not the same as self-esteem. An individual's 8.1. in work is not lowered to the extent they fail at a task they have self—esteem in, but their self-esteem may be lowered to the extent they fail at something in which they have high self investment. The rela- tionship between SJL in work and self-esteem is rather com- plex. For an individual with an initially high 8.1. in work, a high frequency of positive evaluation of that individualTs work helps to maintain or enhance his or her level of 8.1. in work; A high frequency of negative evaluation of one's activity may lead to lower self—esteem if it is not checked. We assume individuals seek a positive self-esteem and will therefore try to alter the situation in the case of low self-esteem. They can do this by challenging the criteria of evaluation and/or by trying harder to succeed (Faunce 1984). To the extent their "challenge" and/or "try harder" response is a success, they are positively evaluated and high.S.I. in work is maintained. To the extent they fail, low self-esteem continues, and they may eventually withdraw their 8.1. in work from that activity. Withdrawal of 8.1. in work is only possible, however, where there are low reminders of onews failure. Where there are many reminders of one's failure (and therefore high negative evaluations of one's activity) withdrawal of SJL.in work is unlikely, and continued low self-esteem may be more probable. We are conceptualizing both self investment in work and self-esteem as aspects of an individual's functioning (broadly seen in terms of orientations toward work). We see these as largely dependent on promotions or rewards (which, in our model, define evaluations of success or lack of success in work activity) and on powemu We, however, also see 8.1. in work and self-esteem as having independent effects on work orientations. Managerial work orientation is thus another concept which relates to individual's functioning in the work organization. We define work orientation as behavioral dispositions individual managers have toward relationships bringing them in contact with others at work. We express managerial work orientations as a set of paired alternatives on a continuum. These are the comparison‘between: (a) Authoritarianism and Participativeness: this refers to a disposition to allow subordinates to contribute to decisions made in the firm as against shutting them out of the decision-making process, believing that they are best fitted to carry out decisions already formulated. (b) Rule-mindedness and Flexibility: this refers to a dis- position on the part of the individual whereby rules are followed to the letter regardless of the situation as against allowing for one to interpret rules according to contingencies . (c) Universalism and Favoritism: this refers to a dispo- sition to follow objective standards in the distribution of rewards and/or sanctions as against selective treatment according to standards an individual sees appropriate. These are necessarily brief notes about the major vari- ables of interest in the present research. The variables are derived from the research questions of our interest. We have a model in which these various variables are linked. From the model we have derived a number of hypotheses we seek to test. Figure 1.1 Working model of the effects of achievements and power on individual functioning Achievements Moderators Work orientations at work LI. »-~«—~>- [PromotiOHL-*~*[Job security} I.. - w-->»Participation vs 1 } :4 Authoritarianism ‘i +___._/ [Rank _*L—~—7~7+C General Self— -esteem"*'Flexibility vs *< “\\\;1 Rule mindedness 1171\2 \ . \ «.Afifr_;[§. I. in work >EUniversalism vs Power 4_}~»~—~~77717' ” Favoritism main effects - - - - subsidiary effects The above diagram summarizes the main relationships of our interest. The concern is with the social psychological effects of achievements at work (i.e., promotion, rank and power) in work organizations. The diagram causally links achievement at work to aspects of individual functioning in work organizations. There are two general linkages this model suggests: the extent to which career experiences directly affect work orientations and the extent to which work experiences affect work orientations as mediated by, mainly, self-esteem and S.I. in work. We derive several hypotheses from the two general linkages which we seek to test. 0n the face of it, we could test the two general hypotheses by estimating the magnitude of the relationships involving direct effects vis a vis those involving mediated effects. This could require path analysis. "Path analysis is basically concerned with estimating the magnitude of the linkages between variables and using these estimates to provide information about the underlying causal processes" (Asher, 1983, p. 30). The path coefficients may be obtained by using regression techniques. For path analysis to be utilised, certain assumptions have to be met (Pedharzur 1982). These are: l- the relationships in the model have to be linear, additive and causal. Relations are said to be causal if: a) there is a concomitant variation or covariation between the two variables in question, b) there is a temporal asymmetry (time ordering) between the two variables and c) other factors which may possibly cause the variation observed in the relationship between the two variables are removed (Asher 1983). The matter of eliminating confounding variables is largely theoretical. 2- each residual is not correlated with variables that 10 precede it in the model. 3- The model must be recursive - that is, it must involve mathematical definitions in which the first cases are given and each nth case is defined in terms of one or more preceding cases, especially the one immediately preceding it. In simple language there must be one-way, causal flow in the system or model. This precludes reciprocal causation between variables. 4- Variables are measured on an interval scale and without error. Many of these conditions are met by our model. The relationships involved are causal and linear in the sense of changes in one variable affecting changes in another. In terms of measurement we used among others the Likert scale which we assume to be an interval scale. However, the model is non-recursive. Relationships among work achievement variables, self—esteem, S.I. in work and work orientations are theoretically specified as bidirectional. Secondly we do not pretend to be testing the whole model as would be required in path analysis. For these reasons we cannot use path analysis. The arrows in the model show relationships which are later on depicted as hypotheses. One such arrow shows how S.I. in work affects the relationship between work achievements and self—esteem. This relationship may be seen as definitional. It is definitional in the sense that S.I. 11 in work is defined as a commitment to work based on the extent to which achievements affect self-esteem. S.I. in work also affects the relationship between achievement and self-esteem causallyu Any particular achievement, e4y, a promotion, may affect self-esteem because of a high level of SLI. in work. An individual may have many promotions or high rank but with low S.I. in work. Where this level of S.I. in work is raised, a noticeable change may be observed in the relationship between achievement and self-esteem. Research Questions and Their Explanations A. Hypotheses Set 1: Career mobility and S.I. in work 1. The more a manager sees chances for promotion and/or attaining higher rank in the organizational hierarchy, the higher his/her S.I. in work. 1a Related hypothesis: The longer a manager stays at one particular level, the lower his/her S.I. in work is associated with number of promotions. 2. The more promotions and the higher the rank a manager has had in a given organization, the higher his/her S.I. in work. 2a Related hypothesis: The older a manager is relative to others at the same organizational level, the lower their 8.1. in work is associated with promotion. The above set of.hypotheses are in recognition of the fact that in work organizations promotions and higher rank 12 are among the major rewards representing the path of achievement. Promotions constitute the most sought after reward package in work organizations because all other rewards such as salary increases, fringe benefits, authority, greater autonomy, greater chances for exercising creativity, and indeed status and the like are all accompaniments of promotion. As such, promotions are almost universally recognized as very important among managers (Sofer, 1970; Thomas, 1983). On the other hand, one's promotions are synonymous with one's achievements and re- flect that individual's performance in work as evaluated by relevant others according to the norms of status assignment present in the hierarchy of the given work organization. Insofar as promotions are seen as achievements in work orga- nizations, they are fundamental to the self investment pro- cess."The self investment process is a selective process through which activities become differentially imbued with significance for the maintainance of self-esteem" (Faunce 1984). There are, as Faunce (1984) points out, potentially many activities which can serve as bases for self— evaluation, but not all activities have the same impact upon an individual's self-esteem (i.e., how individuals judge their own worth. What makes an activity relevant for one's :self-esteem is the importance one (and relevant others) arttaches to that activity. Activities concerned with promotions are very important in managerial circles and, as 13 such, are very relevant to the S.I. in work process. Moreover in the work organization there are no shortages of reminders of one's promotions or lack of promotions. How many people you are able to give directives to, how many people tell you what to do, the number of conflicting de- mands placed on your job, what office you occupy, and where it is located etc., all remind individuals of their status which may be a direct result of promotions they have re- ceived over time. Hypotheses 1 and 2 essentially posit that the more one is promoted (or is promised promotion) the more he or she invests him/herself in activities which bring about promotion. Promotion represents success in work - an important area of activity likely to bring recognition and, therefore, frequent positive evaluation by relevant others. This frequent positive evaluation (as represented by the number of promotions, ranks and the perquisites associated with them) in an important activity is likely to enhance an individual's self-esteem, since self esteem is dependent on the evaluations of relevant others (Coopersmith, 1967). What is more, since promotions increase nonmaterial benefits such as autonomy and creativity, they provide a favourable climate for one's self-confidence and hence a sense of personal worth. The related hypotheses deal with failure to achieve (i.e., get promoted). An individual manager who stays in a 14 job longer than others, just like one whose age is consider- ably above that of others at a given level of the organization, is one who is either experiencing no or little promotion. Our basic argument is that although the kind of work managers do is such that it is very difficult for them to avoid frequent evaluation, it is still possible for some lowering of S.I. in work. The longer a manager stays at a job or the older he or she are relative to others in similar positions, the more likely that his or her efforts to succeed (get promoted) according to existing criteria or to change the criteria upon which promotions are based have not come to much. In the event that such efforts have not succeeded the individual manager may try to underplay the activities associated with promotions. Thus we expect some kind of lowering of the relevance of promotion for self- esteem. However, it is important to note that the related hypotheses will only be true where managers avoid frequent evaluations based on promotions. To this end managers can have some control over the frequency and nature of evaluation. Thus, for example, they might seek positive evaluation for how far they have come rather than.for how far they might go. This may tend to modify the influence of length of stay in a job and age on S.I. in work. A more specific consideration of related hypothesis la is in order here. The idea is that the number of promotions an individual manager has had will be less associated with S.I. 15 in work where one stays too long in present job. However, it is by no means suggested that such an individual will withdraw their S.I. in work. There are alternatives to promotion which may help sustain one's level of S.I. in work. Rank may be such an alternative. Individual's career is carried out in organizations and are in most cases colleague defined (Hall, 1977, p. 279) "How a person feels about his current position depends on how he measures himself. He can look backward; he can look around; he can look forward" (Silk, 1971 quoted in Hall, 1977, p. 279). This relates to Dubin and Tausky‘s (1965) idea of career anchorage. According to Dubin and Tausky, people anchor their assessments of the progress they have made by either focussing on their career origins or by focusing on the level of ultimate possible achievement. Those who focus on ultimate possible achievement, evaluate success in terms of how close their current rank is to the peak of the organization. Those who focus on career origins evaluate success in terms of how far they have come. For those with a downward anchorage it is important for colleagues to agree that they indeed have come far in terms of their careers. For both types of individuals the relevance of achievement for self-esteem is maintained if "success" as differentially defined is reached. 16 B. Hypotheses Set II: Achievements and work orientations l. Participative, universalist, and flexible work orientations positively vary with number of promotions and rank. 1a Related hypothesis: The longer a manager stays at one level of the hierarchy, the less participative and flexible he/she is likely to be 2 . Participativeness, flexibility and universalism posi- tively vary with job security. Promotions and length of stay in a job are related but we see them as different. Promotions refer to actual number of movements in the hierachy, whereas, length of stay refers to time spent in a given job. A person who stays long in a job is one not experiencing promotion.but may have a high number of promotions due to past promotions. The above hypotheses link careers to work orientations. With regard to the first hypothesis, a manager who experi- ences promotion is likely to have been in jobs requiring quite different tasks. In order to adequately perform in them he or she needs to be flexibleu Such a manager need not apply a range of behaviors specific to earlier jobs to present or future ones.‘Mobile managers are bent on going as far as they can in the hierarchy (Jennings, 1971). This commitment to mobility makes them prepare for tasks they expect to assume sometime in the future. To do this they must be flexible. To the extent mobile managers anticipate 17 future promotion, they must prepare subordinates who must assume their present activities in the event they are promoted. Otherwise promotion may be denied on the grounds that there is no one to do their current job. Training others for one's job does not only require flexibility but also participativeness. It is very difficult to train others without letting them in on what you do. Promotions are also a way to control the behaviors of those at the bottom (Granick, 1976; Thomas,1983L.To follow the directives of those at the top, even when they conflict with.onefls'best judgements, requires flexibility. It also requires the involvement of others in a kind of collective effort to carry out the wishes of those at the top and, hence, is a participative process. Mobility also affects work orientations, indirectly, through self-esteem. Self-esteem is dependent on other peoples' judgements of our worthiness. Promotions, in part, imply favorable judgements of worthiness and may, hence, lead to, high self-esteem. Those with high self-esteem have a more accepting attitude towards others. .As such they may be more flexible and participatory in their dealings. In this related way promotion may, through self-esteem, affect work-orientations. Conversely a manager who stays in a job too long is not being rewarded. Such a manager has little incentive to exert all his efforts on the job. Such a manager may choose to work according to rules, which preclude flexibility. Such a 18 manager may not be interested in letting others do what he does lest they be judged as better and hence may not be as participative. An insecure manager is not interested in a superior's review of his/her work. To the extent such a manager relies on what seems to be impersonal and objective standards (i.e., rules) he/she is protected from superior's arbitrary review of his/her efforts. Rules also serve to remove per- sonal conflicts an insecure manager may have with ambitious subordinates. Sticking to rules removes personal risk and serves to protect one from making tough decisions for which he/she may be blamed. Insecure managers may not be partici- pative since this involves making others share power and job activities - things which are rarely inscribed in organiza- tional rules. C. Hypotheses Set III: Promotions, self-esteem and work orientations 1. There will be a lower relationship between promotion and self-esteem for managers with a lower S.I. in work than for managers with a higher S.I. in work. 2. Self-esteem is inversely related to authoritarianism, favouritism, and rule-mindedness. The first hypothesis is addressed to the relationship between promotions and self-esteem. There is necessarily no reason why promotions should enhance one's self-esteem. Self-esteem is dependent on other peoples' judgements of an 19 individualfs worthiness, but the individual must acknowledge these judgements. The more other people hold us in high esteem, the higher our self-esteem” Thus self-esteem may be said to be predicated upon interpersonal relations. Other people form the frame of reference from which we derive our self-esteem. Other peOple form reference groups which we actually use in comparing ourselves to others.'This frame of reference may exist at different levels. It may be at a societal level, emy, others belonging to the same class; it may be at a more middle level (e.g., other working with us in the same organization); or it may be at a micro level (e4n, close friends, family, members of a work group etc.L Which reference group affects our self—esteem may depend on the importance we attachlto given reference groups. Self— esteem is more likely to be affected by the views of members of reference groups we regard as very important. What is implied in hypothesis:l is that promotion will not affect our self-esteem unless we regard the people relevant to our being promoted or demoted as a very important reference group. While we have argued that promotions may lead to high S.I. in work, it is not necessarily the case that S.I. in work will likewise lead to promotion. A person with relatively low S.I. in work may be promoted in the hierarchy provided he/she does work well, has the right connections, and/or plays organizational politics well. Such an individual may prefer a non-work area of activity (e4L, 20 family life). To this extent such a person is less likely to have failure at work (lack of promotions) affect her/his self—esteem, unless such failure interferes with this external area of importance to that person. Hypothesis 1 in hypotheses set III is not merely true by definition, S.I. in work is not merely the extent to which promotion affects self-esteem. Promotion affects self-esteem as an indicator of achievement. Promotion is not, however, the only kind of achievement. There are alternatives such as rank which may serve to affect self-esteem. For hypothesis 1 the lower relationship is expected between promotion and self-esteem where S.I. in work is low because, in part, alternatives to promotion come to be relied upon in the maintainance of esteem and S.I. in work. This relates to our earlier argument that it is hard for managers to reduce their S.I. in work as they are constantly being evaluated and reminded of their achievements. Where one achievement standard does not adequately serve the purpose of maintaining esteem others come to be relied upon more. Hypothesis 2 stems from the fact that self-esteem is dependent on other people's evaluation of one's worth. An individual with authoritarian, favoritist, and rule-minded orientations is one who harbors threats and hostility to- wards others and, as such, has a less accepting attitude to- wards them. A less accepting attitude toward others may translate into a less accepting attitude toward oneself. 21 More importantly, low self-esteem leads to greater rigidi- ties in dealing with others. Low self-esteem individuals may seek (or need) greater personal control of situations in or- der to reduce the risk of failure. D. Hypotheses Set IV: S.I. in work and work orientations 1. Participative, flexible, and universalist work orientations are positively related to S.I. in work. 2. The longer an individual stays in one job, the less their flexibility is associated with S.I. in work. In one way, low SJL.in.work may be seen as an aspect of alienation. It implies some kind of disjunction and/or isolation from some central values of work and the work organization. One of these values is that of mobilizing individuals for attainment of goals - that is, to suit different people to cooperate towards the achievement of a particular goal. Insofar as one has low S.I. in work, this value will not be central to his/her actions, and as such the individual will be less likely to involve others in decision making (iJL, such a manager is likely to have a less participative orientation). A less participative orientation may also be seen as an attempt at reducing the number of people involved in the (and by implication the frequency of) evaluation of a given individual's work. Individuals with low S.I. in work, in order to decrease frequency of evaluation of their work, may find it necessary 22 to follow bureaucratic routines rigidly, thereby calling less attention to themselves and to their work. Low S.I. in work is difficult to sustain in the face of frequent evaluation. Insofar as we see low S.I. in work as reflecting some kind of isolation, it implies absence of social integration, which is central to a participative orientation. Beginning with Weber it has been widely recognized that one of the central values of work bureaucra- cies is impersonality (i.e., absence of favoritism in work practices). An individual manager who has low S.I. in work is one with a disjunction from such values and may hence be unconcerned about universalist (non-favoritist) work prac- tices. The second hypothesis is in recognition of the view that the S.I. process is an affective process. An individual who stays too long in one job may be so emotionally attached as to find it difficult to extricate him/herself from it. Tasks and practices specific to that job may interfere with performance in other jobs and, hence, negatively affect flexibility. Staying too long in one job may lessen an indi- vidual's S.I. in work (as stated in sub-hypothesis 1 of Hypothesis Set I). Hypothesis 2 (in Set IV) goes beyond this by linking S.I. in work to flexibility. Not only does staying long in one job affect S.I. in work, but it may also annul any positive effect S.I. Ln work may have on flexibility. 23 E. Hypotheses Set 31; Power effects 1. S.I. in work is inversely related to powerlessness among managers. 1a.Related hypothesis: Powerful managers in organizations with high total control invest themselves in the job more than powerful managers in organizations with low total control. 2. Authoritarianism, favoritism, and rulemindedness are positively related to powerlessness among managers. 2a.Related hypothesis: Authoritarianism, favoritism, and rulemindedness will be more positively associated with powerlessness among managers in firms with low total control than in firms with high total control. We expect that holding high levels of power will lead to high S.I. in work among managers (hypothesis 1) because it promotes autonomy. We assume that an environment which encourages autonomy is favorable ground for S.I. in work to occur. More importantly, powerfulness appears to be an assurance of positive appraisal or evaluation from relevant others in the work organization. Most individuals in the organization will appear supportive of a powerful manager, largely because of the way work organizations are structured. They are pyramidal structures in which rewards and sanctions flow from the top to the bottom. To be promoted and to acquire autonomy and status there must be a 24 group of individuals supporting individuals efforts at the top (iwe. sponsors). Such support from people who are powerful will be reciprocated. Even where such support is not forthcoming, fear of sanctions will still make individuals supportive of those with power. Such supportiveness for a powerful manager translates into more or less high regard for the individual and his/her activities. Moreover there are many powerful reminders of the achievements of those who are powerful. They have sta- tus, they are able to command more of the organizations' re- sources, they have better placed offices, etc. The related hypothesis (1a) is in recognition of the fact that firms differ as to the total amount of control. In a firm with high total control there is increased participation and mu- tual influence within and between hierarchical levels. More- over in organizations with high total control the greater integration provides consensus as to what is to be recog- nized and evaluated as achievement. Less integration pro— duces less consensus and may reflect multiple standards of evaluation.lfl:is for these reasons that we expect to have more positive evaluations between individuals in firms with high total control than for those in firms with low total control. In regard to hypothesis 2, it is worth noting that pow- erful managers are the ones with the ability to help subor- dinates secure personal goals in the work organization. This 25 help may be reciprocated by subordinates in the form of willingness to follow suggestions made by powerful managers. It is important to note that we are dealing here with the direct path from power to work orientations as depicted in the Model 1. The indirect paths whereby power affects work- orientations through self-esteem and S.I. in work have al- ready been touched upon. "The more the leader (or any mem- ber) helps other members achieve their goals, the greater will be the members acceptance of him." (Pelz, 1952, p.50). With such willingness and acceptance there is little need for a powerful manager to apply heavy handed approaches in relations with subordinates. On the other hand, a powerless manager cannot sponsor a subordinate for promotion and, in many cases, his/her suggestions as to what needs improvement may not be taken seriously by colleagues at the top. At- tempts at helping subordinates may lead to frustration on the part of the powerless manager. The inability to acquire an optimal amount of organizational resources for subordi- nates is not easy to hide and may foster resistance among them. An authoritarian style may be necessary to handle in- dividuals who may be prepared to resist or Challenge the powerless manager. To handle many who are prepared to resist, a policy of divide and rule (favoritism) may be necessary. For a manager in this situation, rules may come to represent an organizational resource by which he/she rewards those who comply with him/her and by which those who 26 do not comply are punished. The strategy'of the powerless manager may be to be rigid or flexible according to subordinate's compliance. Again the related hypothesis (2a) is in recognition of the differential existence of amounts of power in work organizations. For organizations with high total control, there is greater integration, which means an enlargement of common interests between managers and subordinates. Here one's capacity to invoke rules is not so much needed in or- der to secure compliance. In firms with high total control, power will more likely yield status. The converse is also the case. In so far as a higher total control means a greater association between power and status we would expect a greater association between high total control and S.I. in work (hypothesis 1a) and a lower association between high total control and authoritarianism, favoritism and rule- mindedness (hypothesis 2a). Research Setting In general, it is important to consider the context of any study because it illuminates factors which may possibly confound the operationalization of the major variables under consideration. The setting does provide theinterrelated conditions under which the variables of a researcher%5 interest operate. More specifically, in our case, the variables of major interest (i.e., promotion, rank, power, self-esteem, S.I. in work and work orientations) as they 27 exist in industry are of relatively recent origin to Zambia. Not until the 1950'S on self-esteem were high on conservatism (which was equated with authoritarianism). The underlying traits of authoritar- ianism, according to Boshier, make it incompatible with high self-esteem. Acceptance of others is associated with accep- tance of oneself. Authoritarianism is a hostile attitude to- ward others and is ultimately a less accepting attitude to- ward.oneselfu The point is that we expect people who have high self-esteem to be participative, flexible, and univer- salist in orientation (Hypothesis 2 in set III) because self-esteem breeds self confidence. A person who is confi- dent in his/her abilities will allow others to do what he does, and this requires flexibility on his/her part. To the./“ extent he/she is participative, authority is relied on less _, ._-_._ ._ __._-__ .___ . .._ _. - 70 than cooperation in achieving certain organizational ac- tions. POWER EFFECTS A. Powerfulness and S.I. in work Power is an interactional concept and belongs to deci- sion making variables. It can only be defined in relation to the actions of individuals in group settings. For Pfeffer (1981), power is synonymous with politics in organizations. Organizational politics involve those activities to ac- quire, develop and use power and other resources to obtain one's preferred outcomes in a situation in which there is uncertainty or dissensus about choice." (Pfeffer, 1981, p.7) Kanter (1977) defines power in a similar way; For her, power is "the ability to get things done, to mobilize resources, to get and use whatever it is that a person needs for the goals he or she is attempting to meet." (Kanter, 1977, ILIOO) For one to be powerful at work he/she must be credi- ble. Credibility is an interactional attribute assigned to individuals based on their willingness and ability to tend to role obligations.. .. in this sense power in organizations derives from performance of obedience, that is from staging a convincing demonstration that one's actions are appropriate to one's role. (Hamilton and Biggard, 1985, ELIZ) An act of power must be evaluated as correct role performance. Power is always evaluated by relevant others in light of normative prescriptions. Being powerful rests on being included in calculated strategies and decisions, being included in strategies and decisions rests on their evaluation of one's 71 accountability, which in turn rests on one's willingness to obey group standards of behaviour, which include the norm for exercising power." (Hamilton and Biggard,1985.;L13) Our argument is that being powerful is an evaluative process. No one can be powerful unless he or she evaluates her/himself and to a lessor extent is evaluated by relevant others in the organization as successful at "getting things done." A person who is powerful is one who has positive self-evaluation. In this sense one can have power irrespective of whether others admire or accord honor to him/her or not. Our argument, leading to the first hypothesis in hypothesis Set V, is that this frequent posi- tive evaluation leads to greater SCI. in work. B. Powerfulness and work orientations For an individual to secure power in the organization, occupying an authoritative position is not enough; he or she must secure enduring cooperation, ile”. expand the "zone of indifference" (Barnard, 1939) or expand subordinates "zone of acceptance" (Simon, 1976). Parsons (1962) largely agrees with the view of power as command of loyalties leading to orientations conducive to efficiency in organizations. Since enduring power is a question of commanding loyalties, it cannot be secured by being authoritarian or rule-minded or by favoritism. It is the search for legitimate power that produces less authoritarian leadership in organizations. One must be flexible, participative and universalist in 72 orientation. Kanter does not see power in terms of command of loyalties. For her, people who control important contingencies are the ones with a strong basis for personal power. They are able to do this because they can obtain for the subordinates resources, opportunities, and rewards available through the organization. For this, they tend to foster high morale, and they are less rigid, directive, and/or authoritative. Earlier on we advanced the argument that self-esteem has consequences for work orientations. More specifically we argued that self-esteem breeds self confidence which allows individuals to let others participate in activities they engage in. A self confident manager is less likely to enforce administrative rules in order to secure compliance. Self confidence may breed flexibility. Power in an organizational context may positively affect self-esteem. The less power an individual has the lower they evaluate themselves in certain ways. To the extent this is the case power may affect work orientations in the same manner self-esteem does. Empirical studies testing these ideas have been scarce. They tend to generally support the thesis that powerfulness is related to a style which is participative, flexible, and universalist, although the correlations are low. A case in point is the study by South et a1. (1982) in which they tested Kanter's hypothesis that powerlessness among 73 supervisors is positively related to favoritism and a close, controlling style of management; and powerlessness among supervisors is negatively related to job satisfaction and group morale. They obtained questionnaire data from 152 female employees and 10 female and 36 male supervisors in a federal bureaucracy. South et a1. (1982) found that correlation between gender of supervisor and power was strongly negative (iJL, -.67) indicating that male supervisors have more influence than female supervisors. The supervisory style of females was found to be more close (r = .22), more favoritist r = .14), fostered lower morale (r = -.12), and engendered less job satisfaction (r = -.l9) than the style of their male counterparts. The researchers then tested the hypothesis that female style differed from male style because of differences in power. They found that when power was controlled differences between males and females on all these variables were greatly reduced. In general, their results support Kanterfls (1977) assertions. The correlations they obtained were low. We believe this is due to the zero- sum interpretation of power which characterizes Kanter“s (and by implication South et aLJ conceptions. Power is seen as a constant quantity inside an organization. Certain managers have as much power as others do not have, and although there may be a Change in the distribution of power, the total amount remains the same. 74 It is these misgivings we have about zero conceptions of power in organizations which lie at the base of our re- lated hypothesis in hypotheses set V. The extensive research undertaken by Tannenbaum and associates since the 1950's in- dicates that power is an expandable and contractable resource in organizations. Tannenbaum et al.(1968, 1974) have argued that organizational theorists have unduly emphasized relative distribution at the expense of total amount of power. In studying power effects at the work place, it is necessary to look at both its relative distribution and at the total amount of power in the organization. Kanter (1977) and South et a1. (1982) look only at the relationship between a manager‘s power and manager's style without looking at how this relationship is mediated by the total amount of control existing in the organization in which action takes place. Tannenbaum et al. (1974) related power to management styles in a comparative study of the kibbutz in Israel and firms in Yugoslavia, USA, Italy, and Austria. They found that the more highly controlled the organization.(ilen. the higher the total amount of power in the firm), the more par- ticipative were the styles of managers in that organization. Yugoslav and kibbutz plants were found to have a relatively high amount of total control and their managements more par- ticipative in style. Among the three private economy coun- tries, American firms were found to have the highest amounts 75 of total control and subsquently more managers with partici- pative styles. Tannenbaum et al. (1974) say that the differ— ences between Italy and the USA are especially revealing because they show how organizations that are basically alike in the distribution of control (power) may nevertheless differ in total amount of control and hence in consequences. Research conducted by Bartoke et a1. (1982) in ten West German firms similarly indicates that total amount of power in organizations is associated with a participative orientation. A study by Farris and Butterfield (1972) in Brazillian banks also found a positive link between amount of control and effectiveness. The higher the total amount of control a bank had, the more effective it was. We assume that organizational effectiveness is positively associated with flexibility, participativeness, and universalism. If this is so, then the total amount of control should be associated with the afore-mentioned work orientations. The related hypotheses are in recognition of the fact that organizations differ as to their total amount of control. High total control reflects increased participation and mutual influence within and between hierarchical levels. This also implies greater integration of individuals in the organization. Our argument is that such an organization provides an unfavorable climate for authoritarianism, favoritism, and rule—mindedness. The greater integration means an enlargement of common interests and/or value 76 orientations between managers and their subordinates. In such a situation a manager does not need to invoke rules and authority or divide and rule the subordinates in order to secure compliance. AN OVERVIEW Here we recapitulate the main themes emerging from our survey of the literature concerning the effect of work achievements on work orientations. Kanter (1977) contends that achievement at work (i.e. rank, promotion, and power) lead to flexible, participative, and universalist work orientations. She is supported by Martin and Strauss (1959), Jennings (1971), and Granick (1976) who in part argue that microeconomic behavior is regulated by the nature of achievements open within the work organization. However empirical evidence seems to refute these views. Kessin (1971) found few systematic effects of mobility on work orientations Jackson and Curtis (1972) failed to find achievement effects, and so did Seeman (1977) in his comparative study of American and French workers. All of these empirical studies, however, are not concerned with work achievements rather they deal with intergenerational mobility, and they dubiously assume that the self-investment process occurs similarly across occupations. Few take into account such life course events as career stage. A study by Halaby and Sobel (1979) which took some of these factors into account found that achievements at work were 77 consequential for work orientations. Another theme is that S.I. in work relates to achievements at work. Clelland et a1. (1976) compared the ELI. levels of nurses at four levels. They found that the higher the nurse's position in the hierachy the more involved she was in her job. Similarly Newman (1975) found that location factors LIJL job level, department and work group) account for more variance in job involvement than do personal factors. Davis (1966) also reported this to be the case. Mitchel et a1. (1975) contests the supposed positive association between achievement and SJL.in.work, however their study was only limited to auto workers and had a low questionnaire return rate. All these studies equate job level and work achievement. This may be limiting in that achievements may be more encompasing. In the literature there seem to be little evidence to support the idea that work achievement lead to enhanced self-esteem. Studies by Jacques and Chason (1977), Kohn and Schooler (1973), and Bachman and O'Mally (1977) either do not find any correlation or find very little correlation between achievement and self-esteem. This may be so because this relationship is moderated by other factors, notable S.I. in work. For an activity to affect one's esteem it must be perceived as important by that individual. Theorists of power in organizations indicate that it promotes S.I. in work. Kanter (1977), 78 Pfeffer (1981), and Hamilton and Biggard (1985) are unanimous in claiming that power results from positive evaluation by others in the organization. Similarly positive evaluations lead to greater S.I. in work, and since it promotes autonomy it leads to flexibility, universalism and participativeness. South et al. (1982) tested these views. They found low correlations between power and orientations promoting flexibility, universalism, and participativeness. We believe these low correlations reflect the zero-sum conception of power prevalent in this literature. Studies by Tannenbaum and associates indicate that power is both an expandable and contractable resource. Both its relative and its total distribution affect S.I.in work and work- orientations. CHAPTER 3 METHODOLOGICAL CONSIDERATIONS The Sample The method of this research was the personal interview. Personal interviews were carried out with managers in their work organizations. Most of these interviews were completed by this researcher in March-April, 1985. After 50% of the interviews had been completed a research assistant was hired to assist with the interviews due to time pressures. The use of a research assistant raises one methodological issue - that of the type of training the assistant had to conduct the interviews. Fortunately, the research assistant had extensive experience in large scale research projects. She had the experience of working with the Rural Development Studies Bureau, a research arm of the University of Zambia, and had been recommended to the researcher by people in the Department of Sociology there. The research assistant studied the interview schedule for a week, during which time the researcher made sure the assistant understood the definitions, organization, and indeed the concepts in the interview schedule. The research assistant accompanied the researcher to various interviews during which she observed the procedures. Then in the presence of the researcher she 79 80 conducted interviews with five managers. After each of these interviews the researcher extensively commented on the conduct of the interview to the assistant. Managers were operationally defined as any person in the work organization assigned supervisory duties over the work and work-conduct of others in the firm. A total of 106 managers were interviewed in 14 work organizations. These organizations,for reasons stated lateron, were of three types: parastatal (or publically owned) firms, multinational firms, and indigenous private corporations. We utilized a two stage sample design.in thislcross- sectional design survey. In the first stage of our sample, the aim was to draw a sample of firms to be included in the study. We used a simple random sampling procedure in selecting parastatal firms in which we carried out in- terviews.‘Lonrho firms (MNC) were also chosen with such a procedure. Private indigenous firms and two MNCs were chosen without a simple random sampling procedure. In these firms we were assigned an authority, usually a research officer or a personnel or public relations manager, by the chief executive officeru‘We were to work with the assigned person. Our typical procedure was to sit down with the assigned person and complete a list of managers and their distribution within the firm. In some cases we were even provided with organizational charts. In this second stage, we used lists of supervisors and managerial staff so drawn 81 as our sampling frame. We then used a simple random sampling procedure to obtain the approximately ten (10) managers to be included in the interview. Parastatal Firms All the parastatal firms in Zambia are consolidated (ine., subsidiaries ) under the Zambia Industrial and Mining Corporation (ZIMCO). ZIMCO is the holding firm of all the companies in which the state has an interest. As a parent company, ZIMCO may be among the 500 largest in the world. It came into being with the Zambian government's Mulungushi economic reforms of 1969 designed to acquire some measure of economic independence for the country. Prior to the 1969 nationalizations, the economy was entirely'foreignmowned. ZIMCO presently controls nearly 80% of economyu This control is sometimes direct, but it is mostly through its wholly owned subsidiaries such as the Industrial Development Corporation (INDECO). We used a table of random numbers (see Mendenhall, 1971) to select the parastatal firms included in our sample. We used the ZIMCO directory (1984) as a list of the population elements (N):fnom which the sample was drawn. The ZIMCO directory listed a hundred and thirty one (131) corporations. We decided on a simple random sample which could allow us to draw one in every 26 firms from which we could interview managers with equal probability of selection. Using the table of random numbers we drew 5 82 elements(firms). These were the: Zambia State Insurance (ZSICD, Zambia Steel and Building Supplies (ZSBS),.African Farming Equipment (AFE), Metal Marketing Corporation (MEMACO), and the National Milling Corporation (NMC). MEMACO's main business consists in marketing Zambian metals, principally copper and cobalt abroad. On a very limited scale, some of the marketing is done locally. ZSIC is a firm which almost monopolizes the insurance business of the country. It is the insurance company of most of the parastatals. NMC is a firm dealing in cereal products such as the processing and.marketing of'wheat, maize, rice, and stockfeeds. Some of the products which NMC deals with (e4L, wheat and some ingredients in stock-feeds) are at best produced in Zambia in very limited quantities and therefore must be imported. AFE is a firm dealing in farm implements. It imports and services tractors from the United States and other countries. 2838 is a firm dealing in building products like steel, cement, roofing materials and construction equipment. Indigenous Private Firms These firms constitute the smallest sector of the econ- omy. It is much more difficult to gain entree in private indigenous firms. First, they are not grouped under a cen- tral organization as parastatals are under ZIMCO. They are quite scattered, and negotiating entree is a matter of 83 negotiating with individual firms. Most of these firms are small and in many cases cannot yield the sample of 10 managers from each firm which was our target. Furthermore, at the time we were conducting this research, SITET, the government organization overseeing corrupt and/or illegal practices in industry was very active. Proprietors of these firms felt that they were being unfairly singled out and were very apprehensive about having outsiders (let alone researchers) on their premises. The listing of private industries (14L, directorate of Zambian industry, Falcon Press, 1971) which was to serve as our sampling frame was clearly outdated. For these reasons,‘we were unable to apply a simple random sampling procedure to this group of firms. We were only able to gain entree to two private Zambian firms: Mazembe Tractors and International Chemicals (Interchem) which fall under the Chibote Croup of Companies. It is generally believed that the Chibote Group is typical of private indigenous firms in Zambia. Mazembe Tractors Corporation is involved in importing earth-moving equipment for use in building roads, rails, and in clearing fields. Mazembe imports most of its earth moving equipment from such international companies as Caterpillar, Walmet, and Scania. It sells a wide range of services to peOple who buy this type of equipment in Zambia. Interchem is engaged in the production of pharmaceuticals, toiletries, cosmetics, and the like. These human drugs are especially 84 sold to hospitals, drug stores, and chemists throughout the country; Interchem seems to be the largest and one of the most successful concerns involved in pharmaceutical production in the country. Multinational Firms By far the biggest multinational group in Zambia is Lonrho. Lonrho is a holding company with diverse interests in Africa. During the colonial period, it was mainly a Rhodesian mining and land company. With the coming of inde- pendence in the Africa of the 1960's Lonrho transformed it- self into a MNC. It started operations in Zambia around 1961, when it acquired Consolidated.Motors,‘whose business was spare parts. After independence, it acquired many more firms and became the largest MNC group Operating in Zambia. Today, it has about fifty two (52) firms under its control. The number of Lonrho's holdings in Zambia could have been larger had some of its companies not fallen to nationaliza- tion. We used a simple random sampling procedure to choose the three Lonrho firms for interviews. These Lonrho firms are: Commercial Motors, Power Equipment, and Star Motors. Commercial Motors and Star Motors are in the business of im- porting, assembling, and servicing motor vehicles. Star Motors imports cars from Japan.and Europe into Zambia. It also imports spare parts to service the cars of its Zambian clients. Among its Chief competitors is Commercial Motors whose specialty is in importing Peugeot cars..Although these 85 two companies fall under Lonrho, they are fierce competitors - which reflects the decentralized structure of Lonrho itself. In relative terms the companies which fall under Lonrho have considerable autonomy to do as they please. Lonrho does not seem to be a vertically integrated company. The third Lonrho company we dealt with was Power Equipment, whose specialty is selling and servicing electrical machinery for use in diverse settings. We also conducted interviews in two other multinational corporations: Mobil Oil Zambia Limited and International Computers Limited (ICL). These exist as independent entities, and we did not use a simple random sampling procedure to include them in our sample. Mobil is an American subsidiary, and ICL is a British subsidiary of the same name. Mobil Oil imports and sells petroleum products in the country. It is one among many doing so. Its chief competition comes from British Petroleum and AGIP, an Italian firm. ICL is in the computer business. It sells and provides a range of services for computers from England. Its main competitor seems to be a subsidiary of IBM. Wherhree Different Types of Firms? This question is important in the light of the fact that we do not, in our hypotheses, directly refer to the different types of (MNC,parastatal, and private indigenous) firms. This is a small sample survey. There is a possibility 86 that the range of values on SJL.in.work, self-esteem, and work-orientation scales may not be large enough to produce some of the hypothesized relationships. It is possible that some of the hypothesized relationships may require more range in S.I. in work than that existing among managers. This is a limitation we may have to accept. We hope that the use of three different types of firms may serve tolextend the range on some - especially work orientation — variables included in the hypotheses. The items in our construct are meant to capture the different experiences individuals undergo at the work place. The use of three different types of firms may help in a more complete coverage of these experiences. We also utilize Likert-type items in our schedule. These items assume an underlying attitude dimension to which each item monotonically relates. Using the three types of firms may ensure a fuller coverage of the attitude choices coming with each Likert item. We believe MNCs, parastatals and indigenous private firms make for differing work experiences, and by sampling within these strata we are reducing the sampling error. Sampling variability is usually related to variability existing among the basic units of a population. By dividing firms into three types, sampling error becomes a function of within- stratum variabilityu This serves to reduce sampling error where stratum variability is less than the overall variability. Sampling error is, however, primarily a result 87 of size of sample. The Managers This is a two stage sample design. In the second stage we wanted to convince managers to be interviewed. Our pro- cedure was to contact the chief executive (iJL, managing director or general manager) of the firm of our interest. This was relatively easy to do because we obtained cover letters from various organizations and individuals (Appendix I). The chief executive officer usually referred us to the company's director of personnel and/or public relations. If the firm had a research department (ety, ZSIC), we were referred to the director of research. We sat down with the assigned authority and formulated a list of all the people with supervisory duties or, in some cases, a list of supervisory positions in the firm. Lists of supervisory positions were generally easy to obtain from parastatal companies, for they have organizational charts. These charts included most people with supervisory duties, but they do not include those at the lower end of the management hierarchy. (An example of such a chart is shown in Appendix 1;) After we obtained these lists, by simple random sampling procedure we chose 10 managers to be interviewed. Through the officer assigned to work with us we would arrange to have one-hour appointments with the managers selected. We insisted that these appointments be at the convenience of 88 the individual manager. Where a position was vacant or the manager was unavailable, we requested the assistant or the person acting in the position as the substitute. Pretest The purpose of the pretest was to discover the extent to which the interview schedule would be comprehended by the managers. We sought to design questions which carry the same meaning for the respondents as for the researcher, and thus avoid haphazard responses. The pretest was on five in- dividuals. Two of these were administrators at the Univer- sity of Zambia, one taught at a college, another was a re- search officer in ZSIC, and the fifth was an officer (a management position) in the Zambia National Commercial Bank. All these individuals had at least completed high school. Following Schuman (1966), this researcher asked the five re- spondents to "explain a little what they understood by the questions in the schedule and what made them arrive at the alternative they chose in case of closed items." It was made amply clear to repondents that the intention was to discover if the questions on the schedule were readily understood by managers and not to test the five respondents. We divided the explanations as given by the pretest respondents into four categories: (1) The explanation is clear and very near to that intended by researcher. (2) The explanation is not that clear - it touches on some of the meaning intended by researcher. (3) The explanation is very unclear and bears 89 little ralation to that intended by researcher. And, (4) respondent simply repeats the questions as stated in the schedule. We did not do anything if the respondent's explanation was classified according to category 1 or 4. This is because category 1 was what we were seeking, while category 4 did not allow us to judge how adequately the respondent comprehended the question. Where the pretest respondent's explanation was completely unclear (code 3), we aimed at determining whether the question was: too vague, too obscure in meaning, and/or misunderstood because it involved unfamiliar/technical wording. We asked respondents what they understood by certain terms in the question and whether they thought the question was too general or abstract. On the basis of the pretest we found it necessary to drop certain.questions and alter others. We dropped items 110 and 111 from the schedule as all the five pretest re- spondents doubted whether we could get such confidential in- formation as the number of people promoted/demoted in the firm in the last 10 years. The items were important to us so despite their admonition we included them, but after we had conducted about one-third of the research and obtained either very vague answers or outright refusals, we dropped the two items. We changed items 74 and 86 on the schedule because, in the pretest, we discovered they contained terms unfamiliar or too technical to permit comprehension. Item 74 90 was a Likert type item which read, "I have a positive attitude toward myself" with reponse alternatives ranging from strongly agree to strongly disagree. The problem with item 74 seemed to be the term "positive attitude" which was interpreted in a variety of ways. We replaced item 74 with the statement, "I have confidence in myself and in the actions I take." Item 86 which read "A manager should bargain with subordinates to obtain results," was misunderstood in a similar fashion. The problem seemed to be that the word "bargain" which was given the connotation of compromising oneus principles or those of the organization. Thus one of the pretest respondents said that he saw no reason why he should give money to or "pamper" (his meaning of "bargain") subordinates to do what they are paid for or supposed to do. Item 86 was changed to read, "A manager should consult with subordinates to obtain resultsJ' In the pretest we also discovered that items 29, 30, 83, 98, and 104 on the schedule were misunderstood because they appeared vague or obscure in meaning. Items 29 and 30 were Likert type items which read, "If performance went down I may be demoted/transferred." We discovered that the response to this item in part depended on the reasons for poor performance and to vnunn such performance was attributed. Individuals saw no reason to be demoted or transferred if performance dropped due to factors beyond their control, e.g. lack of foreign exchange or raw 91 materials. Demotion or transfer made sense when a drop in performance was linked to the individual. On this basis we changed items 29 and 30 to read, "If it was generally believed I contributed to a drop in performance, I would be demoted/transferred." Item 83, also a Likert type item, seemed to be too abstract for the respondents. It read, "All in all, I am inclined to feel that I am a failure." One respondent's comments were revealing - "All in all, I am breathing and living." We changed this item to read, "In terms of what:I do in life, I am inclined.to feel I am not so successful." Item 98 also suffered from unclarity. It came after an item which asked respondents about the number of children they had. Item 98 asked about age. It was not made clear whose age - the respondent's or the children. We changed item 98 to read "What is your agefl'Item 104 was apparently unclear because it was not too informative. Item 104 required an individual to state his/her income bracket. It was not clear whether we wanted monthly or annual income. We changed the question to include the term annual. The pretest revealed that we could expect problems with the content and not the format of Likert type items with our subjects. The format did not present a problem because of the relatively sophisticated education of our subjects. However on the basis of our experience with the interviews, the pretest alterations were adequate to ensure comprehension of the questions by the managers. We based 92 this judgement on the unsolicited comments we obtained from many respondents with regard to various items. We present a sample of such comments of selected items in the following paragraphs. For instance, we:had.expected some difficulties with the comprehension of items 48 to 59 because they were ques- tions which we borrowed, almost verbatim, from Tannenbaum (1974). They dealt with how individuals rated the amount of control of various groups in the firnfls hierarchy. They were supposed to do the rating using the scale: very little control, little control, some control, quite a lot of control, and a great deal of control. One person who thought workers have some control made this comment: "They have control when they work hand in hand with the union, other- wise they don't." Another individual who thought the chief executive officer (CEO) should ideally have a great deal of control while the rest should have quite a lot of control justified his choice by saying, "the CEO should be different for he takes responsibility for failures in the firm)‘ Item 62 on the schedule is a Likert type item reading, "An individual who wants to make his own decisions can be quickly discouraged hereJ'Those who chose the undecided option defended their choice by such comments as - "it all depends on the type of decision or it depends on whom you are dealing with."‘Those who chose to disagree justified themselves by saying: "We give ample room for individual 93 discretion here" or "This is only so when that decision is clearly inimical to the interest of the firhu" Those who agreed had justifications such as: "The alternative is chaos" or "PeOple like to concentrate authority hereJ' Item 69, a Likert type item, read, "The most important thing to teach workers is obedience to managers/employersfl' Typical comments among those who were undecided included: "You have to motivate workers too."letc. One person who agreed with this statement said, "This is so going by the Japanese experience." Most unsolicited comments came from those who disagreed. Typical of these were: "Not obedience but mutual respect is importantf‘"lt is not always the case that the manager is right;" "What is important is not mere obedience but communication;" "Obedience alone may be counterproductive for it inhibits fresh ideas;" etc. Item 80 read, "Even.if one does not like a rule, they should obey it." Typical comments from those who agreed were: "or else the whole system will collapse;" "If you want to be efficient this is necessary;" "It is nice to obey first and complain later;" and "Rules are for positive things like safety." Those who disagreed typically said: "It is not good to obey even stupid rules;" "A firm is not a military camp;" "At work it is important to have an independent mind;" etc. Typical comments from those who chose the undecided option included: "It depends on the type of the rule;" "If one has good reasons for disliking a rule, 94 then they should not obey it, but if the reasons are flimsy they should." These various unsolicited comments produced by the re- spondents seemed, to this researcher, to be rational reasons for choosing certain responses over others. What is more important, they also indicated that individual managers had understood what was being asked. Operationalization of Concepts An interview schedule (Appendix II) was utilized. The schedule was organized so that items which dealt with in- dividual feelings and private life came last (14%, items 92 to 105). Below we state the concepts of our interest and illustrate their operationalization. Career mobility (promotion): Questions here dealt with an individual's jobs and job history. For each job the indi— vidual had occupied, we asked whether the job was a promotion, a move down from previous job, or a move about the same as previous job. Similarly, the individual was asked to state in summary fashion the jobs which constituted promotions (item 43); the jobs which were lateral transfers (item 44); and the jobs which were demotions (item 45). On the schedule, career items are from 1 to 47 (excepting for items 26 to 30). 95 Hierarchy or Job level: Item 46 on the schedule was designed to gauge job level. This was a closed item in which individual was required to choose from five explained job levels. ELI. in work was measured by Likert-type items: 67, 68, 76, 77, and 88. For instance, item 88 read, "I feel I am a suc- cessful person to the extent I do my work well," with five responses ranging from strongly agree to strongly disagree. Self-esteem: Likert-type items adapted from Rosenberg, e4L, item 82 which read, "I feel I have a number of good quali- ties." There were five such items: 73, 74, 82, 83, and 89. Job security or Insecurity: were also Likert-type items like "If performance went down, I may be demotedfl'(item 29), "There is much job security in my position." (item 26). Job security items on the schedule are from 26 to 30. Work orientations: The rest of the Likert-type items on the schedule (excepting items 60 to 66) were designed to measure various work orientations. These consisted of paired alternatives: participation vs. authoritarianism; flexi- bility vs. rulemindedness; and favoritism vs. universalism. Examples of these work orientations are respectively: "The most important thing to teach workers is obedience to man- agers," "Workers who question old and accepted ways of doing things cause trouble," and "As a manager, it is impossible 96 not to tend to favour one or a few subordinatesJ' Powerlessness: Measured by Likert-type items (60 to 66). The aim was to gauge an individual's influence on the job. An example of such an item is, "Others make conflicting demands of me on the jobJ'(item 66) Total amount of control: This is an organizational level concept derived from summing individual managers' responses. Managers were asked to indicate how much control is held by particular groups in the firm (and how much control they should have). Responses ranged from very little to a great deal of control. Items 48 to 59 on the schedule deal with this concept. PROBLEMS ENCOUNTERED DURING FIELD RESEARCH 1. Nonresponsiveness One individual in a private firm refused the interview outright. This researcher had an interview appointment with him at 10:00 a.m. on Tuesday, April 16, 1985. This manager had insisted that we give him a copy of the schedule prior to the interview. We agreed on condition that its contents were not discussed with anyone. We went through the interview as planned, he told us that he thought the questions were not appropriate for him. We pressed him to elaborate, but he refused. We also treated two other interviews as spoiled. One of these interviews was with a purchasing manager in a parastatal. We met with him five 97 times, but each time after answering a few questions he would say he was very busy and tell us to see him at a future date. As a result, our interview with him never was completed. Another was with an expatriate manager from India. This man worked as chief accountant for one parastatal. After we had obtained responses of the first 76 questions on the schedule, he said the questions we had been asking him were similar and told us to fill in the rest of the answers on the schedule on our own. Our efforts to make him change his mind were unsuccessful. We had a different type of nonresponse from several individuals. This was in the form of refusal to answer given questions on the sched- ule. Eight managers refused to answer this or that question on the schedule. One of these nonrespondents was a general manager. Five (62.5%) of these managers were in level 2 (upper) management, one (12.5%) was a middle level manager, and the last one was a lower level (level 5) manager. The majority (62.5%) were in parastatal firms. Half of these in: dividuals (4) refused to answer the question on income (item 104 on the schedule). Of these, three were upper level managers, and one was in middle level management. Two upper level managers refused to answer self-esteem questions. one of these explained that "These questions have nothing to do with science, they are questions of feeling, and I am not answering them." He was a British expatriate. The other individual who refused self-esteem questions was a Zambian 98 general manager. He told me that after college graduation he attended a job interview with an American firm called Diacarb in which such type of questions were asked - They asked me such questions as have you ever lied, stolen, etc. in your life. They really know that everyone has done things like that in their life. If you say yes, then you will not get the job. If you say no, then you will be lying. So I refused to answer those questions, and I knew I had lost the job. One other upper level manager refused to answer the question about the ideal level of control of various groups in the firm. He said - "This workers' participation is the thing confusing Zambian industry." One upper level manager and one lower level manager refused to answer the question about parents' education. This researcher did not worry very much about these re- fusals to answer some questions. Those who refused to answer some questions consisted of only 7.9% of the total sample. Their number was too small to have any significant impact on the results. 2. Sample distribution In terms of our expectations we were working on some kind of pyramidal distribution of management positions in the various firms. The sample we obtained was clearly dif- ferent. Instead of a pyramidal distribution A, we obtained a pentagonal distribution like <::>». We were dealing with five hierarchical levels of jobs. We managed to interview 3 of level 1 (chief executives) managers; and 32 99 people at upper level (level 2) management. This is understandable in view of the fact that many chief executives claimed they were too busy to talk to us and referred us to their immediate subordinates. Referrals were accepted because they were closest to executives in terms of hierarchical position. We had 37 individuals from middle- level management in our interviews, 19 were from junior level (level 4) management positions, and the rest (10) were mere supervisors (level 5 managers). Thus there were not more managers at the bottom of the hierarchy than at each successive higher rung of the hierarchy. Rather more managers were in the middle, than at the bottom level, or than at the upper level. We cannot determine whether this is the actual distribution of managers in Zambian firms or is due to other factors. According to Bhagavan (1977) 1.3% of the employees in manufacturing are upper level managers, 9.5% are middle level managers, and 14.9% are lower level managers indicating that there is some kind of pyramidal distribution of management in Zambian manufacturing industries. The point may be that this pyramidal distribution does not occur in all firms. It is also possi- ble that the distribution we obtained is an artifact of the way we posed the question. We asked individuals to indicate into which hierachical position their job falls. Studies of stratification indicate that people tend to put themselves in the middle levels more than in any other levels. There is 100 also the possibility that since we were formulating lists of managers in the firm (our sampling frame) with middle level managers they could have inflated their numbers at the ex- pense of lower levels, especially since most lower level managers do not carry the title of managers. The problem here may be that we may have oversampled from certain managerial levels. We do not aim at generalizing this study to all managers, so we do not see the need to develop weights on the basis of a subset of managers which we interviewed at the chief executive and lower levels of management to make proper adjustments. 3. Field administration As already indicated, we worked with assigned authori— ties in the firm, usually people in personnel or research, to construct our sampling frames and to make appointments for interviews. In one parastatal company, the training and manpower officer agreed to contact the various managers in the sample to fix interviews at a time convenient to them. He sent a time table of interviews to the managers involved without explaining who we were and why we wished to conduct the interviews.'This caused suspicions when we arrived at the designated time. We were forced to contact those managers by phone and explain our business. On many occasions, we missed interviews through no fault of our own. The interview generally lasted an hour. If we were doing research in a firm, we could arrange for an 101 interval of one hour between interviews, but with some managers we could sit in the office for more than one hour before we started the interview. As a result we were necessarily late for the next interview. We always explained to the people involved in the next interview why we were late. Many accepted these explanations, but a few did not. We also had some difficulty in dealing with individuals who had their own ideas about the research. They insisted on telling us what they thought we wanted to hear and seemed to have little interest in our questions. One of these had a doctorate in chemistry. We found that for most questions he would somehow end up talking about how frustrating it was to work in parastatals where every upper level manager consid- ered a person with such education a threat. It also did not help our morale that quite a substantial portion of our in- terviews were postponed more than twice. As already indicated, gaining entree to private firms was much more difficult than gaining entree into public firms. Some firms had even taken steps to confirm or au- thenticate my status as a researcher. Mobil Oil Zambia Limited went so far as to seek confirmation of this status from the director of the African Studies Center at Michigan State university. Scaling: Analyzing Items Our concern here is with some of the measurements we employed in our interview schedule. We used Likert type 102 items to measure some behavioral dispositions of individuals at work. We had several items designed to measure a given behavioral disposition. For instance, there were six items each designed to measure self-esteem and self-investment in work. The aim of doing this is to develop a scale on which individuals can be differentiated. The problem we face is - can we presume the items are measuring the same thing? When the responses for each individual are combined do we have reliable scale which can differentiate individuals on various dispositions (McIver and Carmines 1981). Put more simply - are the items sufficiently related to be measuring the same thing? Our aim is, in part, to obtain items which when combined will best measure the attitude in question. We will eliminate items which do not appear to differentiate people on a given attitude or disposition. The following analysis is patterned after that of McIver and Carmines (1981). The aim is to compute scale reliability. Following McIver and Carmines (1981) an approach that most researchers feel is preferable is based on interitem correlation matrix (Anderson, Barsilevsky and Hum 1983) This is defined as Np/[i+p(N-l)], where N, is the total number of items and p, is the mean correlation among the N, items. If the correlations are high the items are presumed to be measuring the same thing" (Anderson et al., 1983, p. 254). For the purposes of shedding light on the questions raised above, we have utilized the S.P.S.S. subprogram "Reliability," using reliability estimates and factor 103 analysis. The following Tables (3:1 to 3:4) contain corrected item to total correlations, means, and factor loadings of the various items in the scale. The various items had responses with values from five to one ranging from strongly agree to strongly disagree. The means in the tables indicate the average score individuals had on a given item. Item to total correlation is corrected. This means that it does not include the item's value as a component. We do this to avoid artificially inflating the score as where this correction is not made both the item and the total scale scores contain the item's variance. The corrected item to total correlation is designed to measure the interelatedness of the scale in use. Guttman criticized Likert type scales because it is not clear whether the items included belong on a unidimensional scale. Factor analysis is an effective technique specifically designed to answer the question of dimensionality of multiple summative items (McIver and Carmines 1981). The factor loadings for the various items in Table 3:1 are included so that we have a basis to judge whether the amount of variance shared by the items is suf- ficient to justify the assumption that they are measuring the same thing. The factor solution we utilize here employs the usually accepted convention that factors in the related solution should be determined by an inner (eigen) value level of 1.0. The implication is that "only components 104 accounting for at least the amount of total variance of a single variable will be treated as significant" (Nie et a1. 1975 p. 479). We have several decision rules with regarding the interpretation of Tables 3:1 to 3:4 and especially the inclusion of items on the scale. These tables show item total correlations, communality , and factor loadings. "Communality" is the variance of an observed variable accounted for by the common factors in a factor model. Factors refer to hypothesized. unmeasured. and underlying variables assumed to be sources of observed variables, while factor loadings refer to coefficients in a factor matrix. It is the usual practice that loadings are considered salient and allowing for interpretation when above or equal to 0.3 (Suarez 1982 and Gorsuch 1984). The rule of the thumb is to consider factor loadings less than 0.3 as not substantial (Kim and Mueller 1978, p. 70). Following Kim and Mueller (1978) our interest is in factor based scales rather than factor scales. Factor-based scales are scales utilizing only some of the information obtained from factor analysis. Kim and Mueller (1978) provide two justifications for doing this: 1) some of the particular values obtained in a factor solution are subject to sampling errors. Thus researcher "may ignore specific variations in the factor loadings and consider only one type of information as relevant: either a variable loads on a given factor or it does not” (Kim and Mueller 1978, p. 70). When this is done only variables with 105 substantial loadings are summed.for the scale and those with minor loadings are ignored. 2) In simple index construction the factor analytic model is not expected to fit the data completely because non-random measurement errors in variables and minor factors not related to what is of interest may account for some of the observed correlations which in turn may affect the weights obtained. Thus, "particular numbers obtained contain substantial 'noise,’ it may be wise to ignore minor distinctions and differences" (Kim and Mueller 1978, p. 72). With these comments in consideration we base our judgement as to which items we include in our scales on three rules: (1) the item to total correlation (a measure of reliability using the method.of internal consistency) must be near 0.5 or above (2) an item with above 30% contribution to the variance (communality) in the factor solution will be taken as contributing significantly to the scale, and (3) we will be interested in items with substantial (above 0.3) factor loadings. We will include an item in the scale where two of the above rules are met. For self-esteem items in Table 3:1, all of the six items show high (i.e. above .300) item to total correlation. Items SESl and SE84 are the only one with insubstantial factor loading values. However for both, the factor loadings like the item-to-total correlation (a reliability measure) are substantial. For these reasons we have decided to retain both items SESl and SE54 in the self-esteem scale. 106 The item we decided to drop with regards to the S.I. in work scale is SIWl. This item again shows a non-substantive factor loading and its item to total correlation is low, the contributions of SIWl are indeed very low. It is an item which read "The major satisfactions of my life come from work". The rest of the items on the S.I. in work scale dealt with individual's judgment of the importance of work for his/her psychological existence. SIWl may be more a measure of job satisfaction than of S.I. in work. Table 3:2 shows how items on the powerfulness vs power- lessness scale and job security items fared. In general the self-esteem and S.I. in work scales were much more interrelated than those shown in Table 3:2. Many of the items on these scales had to be removed. Table 3:1. Item Analysis of Self-Esteem and Self-Investment in Work Scales (N=101L CORRECTED ONE FACTOR SOLUTION USING ITEM-TO- PRINCIPAL FACTOR WITH ITERATION ITEM MEAN TOTAL CORRELATION FACTOR LOADING COMMUNALITY SE81 2.43 .478 .500 .250 SE82 3.06 .645 .761 .579 SE83 3.15 .698 .771 .595 SE84 2.32 .406 .424 .180 SE85 1.83 .702 .781 .610 107 Table 3.1 (continued) SE86 2.74 .714 .680 .680 v7.3.3}? "3:93- "mTBZSm" """T63§""" ”"129 '''' SIW2 2.54 .468 .528 .322 SIW3 2.84 .573 .702 .498 SIW4 2.73 .473 .644 .433 SIWS 2.31 .564 .727 .535 SIW6 2.55 .388 .520 .283 SE8 = Self-Esteem, SIW = Self Investment in Work. Negative loadings correspond to negative wording of the items on the questionnaire. * = removed from the scale as a result of its low values. Self-esteem SE81 = I feel I do not have much to be proud of. 8E82 = I have confidence in myself and in the actions I take. SE83 = I feel I have a number of good qualities. SE84 = In terms of what I do in life, I am inclined to feel I am not so successful. SESS = I certainly feel useless at times. SE86 = I feel I do have much to be proud of. S.I. in work SIWl = The major satisfactions in my life come from work. SIW2 = I do what my job description requires, no firm has a right to expect more. SIW3 = Most things in my life are more important than work. SIW5= I feel I am a failure as a person when I fail at something to do with work. SIW6 = I feel I am a successful person to the extent I do my work well. 108 Thus we removed two of the five items on the Power scale» Pow4 and Pow5 have item to total correlations, factor loading and communalities which.sre not substantial. The exception.is the factor loading for Pow4. The other items Powl, 2, and 3 have at least two of the measures which are substantial. Thus for Powl, while the item to total correlation is low, its factor loading is very high (.87), and its contribution to the variance in the factor solution is also high (76%). Both the item to total correlation and the factor loading for Pow2 are substantial. Its communality is negligible. The item to total correlation like the factor loading for Pow3 is high (respectively .5 and .7). The contribution of item Pow3 to the variance is also substantial (i.e. 56%). One job security item (Josl) does not meet our decision rule for inclusion in the scale. None of the job security items show very high item to total correlation, although that of Jos3 is substantial.We have included items JosZ, Jos3, Jos4, and JosS in the scale because they have significant factor loadings and show a high contribution to total variance (i.e. communality). Item Josl shows too low scores across the board to be interrelated. 109 Table 3:2. Item Analysis of Powerfulness and Job Security Scales (N=101) CORRECTED ONE FACTOR SOLUTION USING ITEM-TO- PRINCIPAL FACTOR WITH ITERATION ITEM MEAN TOTAL CORRELATION FACTOR LOADING COMMUNALITY Powl 3.89 .062 .872 .761 Pow2 3.17 .313 .329 .108 Pow3 2.35 -.554 -.752 .565 *Pow4 2.58 .222 .498 .248 *Pow5 2.54 .101 -.165 .027 *Josl 2.57 .053 .271 .074 JosZ 1.84 .256 .867 .848 Jos4 1.66 .294 .749 .637 JosS 2.96 .129 .448 .890 Pow = Powerfulness, Jos = Job-security. * = removed form the scale because of low values. Power and Powerfulness Powl = My job lets me use my skills and abilities. Pow2=I have much influence on the decisions made by my supervisor. Pow3 = An individual who wants to make his own decisions can be quickly discouraged here. Pow4 = Others make conflicting demands of me on the job. Pow5 = My job involves unpredictable situations. Job security Josl = There is much job security in my position. JosZ = There should be more job security in my job. Jos3 = Job security is very important to me. Jos4 = If my performance went down I may be demoted. JosS = If my performance went down I may be transferred. Table 3:3. 110 Item analysis of work orientations (flexibility vs rulemindedness and universalism vs favoritism) Scales (N=101) CORRECTED ONE FACTOR SOLUTION USING ITEM-TO- PRINCIPAL FACTOR WITH ITERATION I TEM MEAN TOTAL CORRELATION FACTOR LOADING COMMUNALITY *FLRl 2.88 -.006 .150 .150 *FLR2 3.16 -.136 .241 .129 *FLR3 2.10 .096 .229 .257 FLR4 1.91 .019 .341 .227 FLR5 2.94 .110 .486 .276 UNFl 2.55 .432 .663 .440 UNF2 2.30 .452 .678 .459 *UNF3 1.90 .265 .338 .114 FLR = Flexibility - Rulemindedness, UNF = Universalism — Favoritism. * Removed from Scale because of low values. Flexibility vs Rulemindedness FLRl = workers who question old and accepted ways of doing things may cause trouble. FLR2 = In any effective firm, an employee must sometimes disregard the wishes of his/her superior if he/she is to do the job properly. FLR3 = People are better off when the organization does not provide a complete set of rules. FLR4 = Even if one does not like a rule they should obey it. FLR5 = Often to be effective an employee must be flexible and not rigidly follow rules. Universalism vs Favoritism UNFl = Individuals who comply with managers should be allowed to bend rules a little. UNF2 = A3 a manager it is impossible not to tend to favor one or a few subordinates. UNF3 = Well behaved employees should not be punished as much as those who do not behave well when they do something wrong. 111 Table 3:4. Item Analysis of Work Orientation (Participation- authoritarianism) Scale (N=101) CORRECTED ONE FACTOR SOLUTION USING ITEM-TO- PRINCIPAL FACTOR WITH ITERATION ITEM MEAN TOTAL CORRELATION FACTOR LOADING COMMUNALITY Paul 2.00 .778 .872 .761 Pau2 1.93 .712 .757 .573 Pau3 1.36 .725 .791 .625 Pau4 2.55 .768 .874 .765 Pau5 2.54 .407 .432 .187 Pau = Participation Vs authoritarianism. Participation vs authoritarianism Paul= The most important thing to teach workers is obedience to managers/employers. Pau2=The world of work is complicated. The only way to know what to do is to rely on managers/experts. Pau3= Managers should not encourage subordinates to find new ways of doing things. Pau4=A manager should negotiate with subordinates to obtain results. Pau5 = A manager should be strict with subordinates in order to gain their respect. In the flexibility-ruleminded scale (Table 3:3) we dropped the first three items (FLRl, FLR2, and FLR3) because their values were so low across the board. Items FLR4 and FLR5 have substantial though relatively low factor loadings. These items do not actually meet our decision rules, however, a flexibility scale was needed and because they are 112 the only ones with high factor loadings we retained them to form the scale. This shortcoming should, however, be noted in the explanation of results using this measure. For the universalism-favoritism scale, item UNF3 has an item to total correlation of .338 which is substantial. We have decided to drop item UNF3 from the scale because it is low on communality (.14) and on item to total correlation (.265). The item to total correlation, the factor loadings, and communality scores for UNFl and UNF2 are high enough to justify the assumption that they are measuring the same thing. The last Likert-type item we have is that designed to capture the work orientation of participation vs authori- tarianism (Table 3:4). The corrected item to total correlation (a measure of reliability), the communality, and the factor loadings of items in the participation- authoritarian scale are all substantial (most being above .500). We therefore feel quite justified to state that they are measuring the same thing. i.e. they form an interrelated scale upon which different individuals' authoritarian and/or participative attitudes may be differentiated. CHAPTER 4 FINDINGS Promotional Effects We have utilized three statistical methods to test the first sets of hypotheses which deal principally with the effect of achievements on work orientations. We have used the S.P.S.S. program popularized by Nie et a1 (1975). From it we have used regression analysis, zero-order Pearsonian correlations, and S.P.S.S. graphics. HYPOTHESES SET I Hypotheses 1 and 2: "The more a given manager sees chances for promotion, the higher his/her S.I. in work," and "The more promotions a manager has had in a given organization the higher his/her S.I. in work." These in“) hypotheses relate promotion to self investment in work. In hypothesis 1 we investigated the individual's perception of promotion an attitude i.e. as Chances of being promoted. In hypothesis 2 we focus on the individual's experience of promotion, treated as the actual number of movements that an individual has experienced within and/or between hierachies in present or past organizations. The prediction according to these hypotheses is that the more promotions an individual has had, the 113 114 greater his/her self investment in work. Promotion.is the independent variable. The intercorrelations between promotion and other variables (including work orientation variables) are shown in Table 4:2. We expected promotions to be related to work orientations, and this has essentially been shown not to be the case. In Table 4:2 all the Pearsonfil correlations involving promotion are non-significant except ! for those involving global self-esteem, which is, however, negative. However, even this significant correlation is very l low, r=-.18. Table 4:1 shows utilization of promotion as afi/’ independent variable using regression analysis. We use regression in part because it enables us to estimate and compare the size of variance contributed by each factor. Regression is a statistical technique for analysing relationships between dependent and independent variables. It is a descriptive tool by which we can summarize and decompose the linear dependence of given variables on others. In terms of description, regression can be used to assess the relationship of one variable or another and to evaluate the contribution of a given variable to a set of variables. In Table 4:1 we assess the impact promotions have on self—investment in work in particular and work 115 Table 4:1. Regression Coefficients of Promotional Effects (as number of promotions) on Work Orientations WORK PROMOTIONAL Fa R2 Fb ORIENTATION CONSTANT EFFECT S. ESTEEM 11.0 -.667 (-.184) 3.4 .03 83.4 S.I. IN WORK 10.5 -.250 (-.090) .81 .008 125.0 FLEXIBILITY 5.4 -.128 (-.075) .56 .005 89.0 UNIVERSALISM 6.0 -.318 (-.149) .22 .02 70.8 PARTICIPATIVENESS 6.4 -.382 (-.142) 2.0 .02 51.5 * P < .05, ** P < .01 Fa are F's pertaining to the net effect of promotional classification. Fb are F”s pertaining to the equation as a whole. Standardized Coefficients are in parentheses. 116 Table 4.2 Zero Order Correlation Matrix for Promotion, Job Security,Power, Age and Work Orientation Variables (N = 101). 1 2 4 5 6 7 9 10 ll 1 PROMOTION 1.00 2 S.ESTEEM -.18* 1.00 3 S.I. IN WORK -.09 .55** 1.00 4 FLEXIBILITY -.07 .55** .27** 1.00 5 UNIVERSALISM -.14 .56** .44** .38** 1.00 6 PARTICIPATION -.14 .71** .321" .49** .44** 1.00 7 JOB SECURITY -.O9 .56** .27** .37** .39** .53** 1.00 a LENGTH .07 -.62** .09 —.4a** -.4sfl -.as** -.58** .00 9 POWER -.06 .27** .25** .05 .13 .2l** .33** .11 1.00 10 AGE .07 -.13 .02 -.02 -.03 -.10 -.26** .27.. -.25** 1.00 11 RANK .13 .57** .18* .47** .38** .71** .55** .70** .08 -.14 1.00 12 EDUCATION .02 .36** .07 .29** .31** .40** -.34** .41** -.23** .007 .60** 1.00 117 orientations in general. None of the equations is significant in Table 4:1. A look at R2 (i.e. the percentage of variance in the dependent variable explained by promotion) shows that at best promotions explain less than 3% of the variation in work orientation variables. The variance accounted for by promotion in the other variables is low. A similar picture emerges when we look at the relative impact of promotion on the variables listed in Table 4:1. The unstandardized betas - which represent the direct translation of change in independent variable into change in dependent variable - show that the impact of promotion on participativeness is less than one half (1/2) of a standard deviation unit. The impact of promotion on self-esteem, and S.I. in work, flexibility and universalism is also less than one half of a standard deviation unit. All these relationships are negative. According to these data then, promotions do little to Change or affect an individual manager's self investment in work. In Table 4:1 the relationship between promotion and S.I. in work is not significant. According to these data we are forced to reject hypotheses 1 and 2 in favor of the null hypotheses which posit no relationship. This acceptance of the nulf\\ hypotheses is reinforced when we lOok at the graphs 5 depicting the relationship between promotion and 8.1; in/J{ work. Figure la shows the regression lines depicting the 118 relationship between promotion and 8.1. in work. The slope which this regression line shows is not so steep, but it clearly indicates a negative relationship between the number of promotions and 8.1. in work. That is, as the number of promotions increases, among managers included in this sample, so does their level of 8.1. in work tend to decrease. This slope is essentially the regression of the mean scores which are plotted in Figure 1b. Figure 1b shows the mean S.I. in work scores for peOple with different number of promotions. The means show a peak with two promotions. Individuals with two promotions seem to have higher S.I. in work than others. It then starts to decline. Individuals who have had three or four promotions show less and less S.I. in work. The curve for peOple with five promotions falls even more sharply - i.e. they show even lower levels of S.I in work. In sum, our test on these data show a negative relationship between promotions and S.I. in work. What could be the possible explanation for this? Hypotheses 1 and 2 were formulated in regard to the writings of Kanter (1977), Jennings (1971), and Martin and Strauss (1959) on promotions. These assume that promotions ultimately mean S.I.IN WORK 4.000- 3.500? 119 FIGURE 1a EFFECTS OF PROMOTION AND RANK ON S.I. IN WORK EFFECTS OF PROMOTION 8N0 RRNK ON S.I.IN WORK PX + 2.000- 1.500' 1.000: .500; 1.000 1.500 2.600 2.500 3.300 3.500 4.800 4.500 5.300 NUMBER OF PROHOTIONS(+I LEVEL OF RHNK(R) S.I.IN NORK 0000" -SOD-r .OOGr .500“ 0000" 0500‘” 0000" 0500“ JJZC) FIGURE 1b EFFECTS OF PROMOTION AND RANK ON S.I. IN WORK EFFECTS OF PROMOTION 8N0 RRNK ON S.I.IN WORK 1.000 1.300 2.600 2.300 3.300 3.300 4.000 4.300 5.000 NUMBER OF PROMOTIONSI+I LEVEL OF RRNKIR) 121 advancement in the corporate ladder or hierarchy, and imply greater reward packages such as power, income, creativity, and autonomy. This may be the case in industrialized countries where norms and culture have conformed or come to terms with industry for a long time. In industrial countries an individual's promotion may ultimately mean breaking from a given hierarchy to an upper one. In less developed countries like Zambia this may not be the case. An individual manager who is in the lower echelons of the hierarchy may receive four or five promotions without necessarily breaking out of the lower levels of the organization. For such a manager, promotions may mean greater pay, but it is doubtful if they lead to jobs with greater power, creativity, and autonomy and may ultimately prove to be frustrating. On the other hand, an individual who begins in the upper levels of the organization may advance into even higher ones with one or two promotions. A look at what has been written in Zambia about promotions may serve to illustrate our point here. At independence on October 24th, 1964, among its 5 million or so people, Zambia had less than 1000 high school diplomas, and less than 100 had a first university degree. Post primary school was very underdeveloped (Elliot 1975) Primary education fared somewhat better. During the colonial era Zambia had a dual labor market structure, stratified on the basis of race (Mhone 1984). Europeans occupied and were 122 recruited to positions in the organizational hierachies with more rewards in terms of wages, creativity, power, and the like. Africans were in career lines involving unskilled, low autonomy, and low paying jobs.‘This dual labor market structure was justified on the basis of race during the colonial era and on the lack of education and skills after the attainment of independence. There were pressures for the promotion of Africans in industry during the colonial period. Thus nearly all the trade unions and social political movements of the time were preoccupied with the issue of "African advancement." After independence the pressure for promoting Africans intensified.'The concern was with what was called "Zambianization." Africans pushed for greater promotions in industries for themselves now that the government was in their hands, but they did not have the requisite education and skills necessary at upper levels of the existing organizations given the generally low level of education among the African populace. After independence promotions were anomic (Burawoy, 1972; Mhone, 19840. The problem of anomie, as posed by Zambianization (promotions) was defined in economic terms by Mhone (1984 p. 156) as the dysfunction of economic goals and norms with economic structure, resulting from the inability of the latter (i.e. economic structure) to be conducive to the achievement of the fermer (goals and norms) by a particular group. In other words the goals of many Zambians in industry was to advance up the corporate ladder, but given that they had 123 little education and skill (due to the limiting colonial condition) there was an absence of legitimate means for them to achieve this goal. However, with independence, the political climate was such that such promotions had to occur. In this situation many firms responded by moving quite a number of Zambians from their jobs to different ones while simultaneously increasing their wages. This was effectively interpreted as "Zambianization" or promotion in industry. These were job changes accompanied by wage increases without concomitant changes in power, creativity, and autonomy. Case studies in Zambian mines by Burawoy (1972) reveals how this was the case. The strategy of mine firms consisted in job fragmentation - a strategy in which promoted Africans could only take over a portion of the job at the higher level usually occupied by an expatriate, while at the same time creating new positions for the displaced expatriate labor. This served to increase authority levels between Zambians and their expatriate supervisors (Burawoy 1972 p. 30). The newly promoted Zambians and the expatriates displaced into the newly created positions do much the same jobs as they did before the succession occurred. All that has changed are the salaries, which have both increased, and their job titles, which both sound more important than the previous ones (Burawoy 1972 p. 30). Presently the racial element has been expunged from industry. We suspect and/or believe that job fragmentation and/or jOb modification accompanied by change in title and wages has remained the dominant response to pressures for 124 promotion. Promotions in Zambia, especially at the lower levels of the hierarchy, do not necessarily lead to an individual Changing his/her rank or the hierarchy into which one's job falls. A look at Table 4:2 shows that this is not a far fetched statement. In Table 4:2 the correlation between rank and promotion while positive is insignificant and extremely small in magnitude (r = .13). According to these data it may just not be the case that the more you are promoted the higher the level of the hierarchy in which your job falls.'Fhere is another possiblity with regard to the workings of the variable promotion. The low correlation between promotion and rank instead of being genuine may be due to measurement error. Promotion as measured in our interview schedule may have turned out not to mean movement up the organizational hierachy - a meaning which we initially intended. The problem of validity of measures is a real one in view of the sensitivity of promotions and activities surrounding them. There are differences among people that systematically affect their willinguess to give accurate responses, particularly to certain kinds of questions such as those that might affect their self-esteem (Bradburn, 1983, p. 292). We asked peOple to indicate whether every job»change they told us about was a promotion or not. In hindsight we could have done better if we had used feed-back questions such as why they regarded given job changes as promotions; did these changes entail income and power increases, did they lead 125 them to middle or upper level management, etc. One of the most persistent features of the colonial or dual labor market structure of the colonial era is the legacy of the emphasis on rank/hierarchy. As already pointed out, there were two labor markets. The European labor market had its own pyramidal shape with top, middle, and lower hierarchical levels. Below it was the African labor market - also pyramidal and differentiated hierarchically. The working conditions of the top of the African labor market did not even come close to those at the bottom of the European labor market. Zambianization was in part aimed at integrating the two labor markets. During the colonial era the rigid insistence on rank/hierarchy was based on race and access to education and skills. After independence the race element dissipated slowly. In these data rank emerges as a very strong predictor of work orientation - and to a lessor extent 8.1. in work. Table 4:3 shows the relationship between rank and work orientation variables. The regression coefficient showing the effect of rank on S.I. in work is significant and positive (r = .18). Table 4:3 shows bivariate regression equations involving rank as an independent variable. These equations do explain considerable variation in work orientation variables, ranging from 3% for SJL.in.work to 51% for the equation involving participation. The variation 126 Table 4:3. Regression Coefficients of Rank Effects on Work Orientations WORK RANK Fa R2 Fb ORIENTATION CONSTANT EFFECT S. ESTEEM 5.5 1.7 (.57)** 49.5 .33 77.1 S.I. IN WORK 8.8 .43 (.18)* 3.6 .03 232.5 FLEXIBILITY 3.7 .67 (.47)** 29.5 .22 135.3 UNIVERSALISM 3.7 .67 (.38)** 17.1 .14 77.4 PARTICIPATIVENESS 2.1 1.59 (.71)** 104.9 .51 28.5 * P < .05, ** p < .01 Fa are Fs pertaining to the net effect of Rank classification. Fb are F3 pertaining to the equation as a whole. Standardized Betas are in parentheses. due to rank in self-esteem and flexibility work orientation is fairly moderate (being 33% and 22% respectively). In part regression analysis is used to examine relationship between dependent and independent variables with the aim of evaluating the impact of one variable on another. To do this the important statistics in Table 4:3 are constant and unstandardized.betas."The unstandardised betas represent the direct translation of change in the independent variable into change in the dependent variable" (Achen, 1982, p. 69). In Table 4:3 the unstandardized coefficients represent increments in standard deviation units of self-esteem, ELI. in work, and work orientations associated with each increment in rank. In Table 4:3 the theoretically predicted score on self-esteem and participativeness is 5.5 and 2.1 127 respectively when rank equals zero. There is a 1.7 and 1.5 standard deviation unit increase respectively in self-esteem and participation whenever an individual's rank rises one level. There is close to one standard deviation change in universalism with each increase in rank. The increase in the level of S.I. in work and flexibility is approximately one half a standard deviation and that of flexibility near one standard deviation with each increase in individual manager's rank. Rank thus seems to positively affect all work orientations, but it does so more in the case of self-esteem and participativeness. A look at the graphs depicted in Figures 1a and b essentially reveal the same picture. The plotted regression line in Figure 1a indicates as mean levels of individual manager's rank rise, the level of self- investment in work also rises. The plotted means in Figure 1b also show that, in general, S.I. in work rises with rank. More specifically, individuals with second level rank show a moderate rise in the mean levels of their S.I. in work over those with jobs falling in the first hierarchical level. This rise in mean S.I. in work somewhat declines with the third and fourth levels of rank. It however rises again with executives (level five in the hierarchy). The overall picture, then is that the higher one's rank is the more they invest themselves in work. This is essentially in line with the findings by Clelland et al. (1970), Newman (1975), and 128 Davis(1966). The rewards we normally associate with promotions are for people in the higher levels of the organizational hierarchy. People at higher levels of the hierarchy are more likely to have jobs which are more creative, autonomous, and powerful than those at the bottom levels of the hierarchy. The higher the level of the manager in.the hierarchy, the more unstructured, unspecialised, and long range the job; the more complex, intertwined, and extended in time the issues handled are the less focused the work. Lower level managers are more oriented toward operating problems, and experience greater fragmentation in their work; they spend more time in the decision roles, especially in disturbance handling and negotiation (Mintzberg 1973 p. 130-1). The nature of managerial work allows for more SJL.in work at the top levels because it satisfies intrinsic needs. Related Hypothesis la: "The longer a given manager stays at one particular level, the lower their S.I. in work is associated with a number of promotionsfl' This hypothesis also relates to the supposed effects of both promotions and rank. An individual who stays long in a given job is one who is neither experiencing promotion nor moving up the organizationfls hierarchy and therefore not experiencing additions in autonomy; creativity; andlother job rewards. On this basis we expect lower S.I. in work levels for people who stay long in a given job. The zero order correlations in Table 4:2 do not reveal this to be so. 129 There is a nonsignificant correlation between S.I. in work and length of stay in present job (r=-.09). This correlation is low enough for us to conclude that the length of stay in a job, does not affect an individual managers level of 8.1. in work. Thus we reject hypothesis 1a. Does length of stay in a job affect the relationship between promotion and S.I. in work? This does not seem to be the case. As Tables 4:8 and 4:9 reveal, for those with less than 2 years in the job, promotionis not significantly correlated to S.I. in work (r = 0.17). The correlation between number of promotions and 8.1. in work also does not turn out to be significant for those with 3 to 4 years in the job (r=-.Ol). This correlation is even negative, but its magnitude is so small that it is safe to conclude there is no relationship between S.I. in work and promotion for people with 3 to 4 years in present job. This is also essentially the case for those with five or more years in their jobs. The correlation between 8.1. in work and promotion is also not significant for these managers (r=-.15). However the negative sign implies a situation whereby level of S.I. in work falls with number of promotions. Thus S.I. in work does not seem to relate to promotion for all managers regardless of how long they have been in their present jobs. Related Hypothesis 2a: "The older a manager is relative to others at the same organizational level the lower their S.I. in work is associated with number of promotions" and 130 "The older a manager is the less their 8.1. in work." According to this hypothesis the older an individual manager is in relation to others at the same organizational level the lower the relationship between promotion and S.I. in work. This hypothesis ideally can be tested by comparing the correlation of S.I. in work and promotions of individuals with different ages at one rank within the organization. We cannot have this direct test because we do not have enough cases of persons in the same organization to permit comparisons with others at the same rank. We can however test for all cases by dividing age by rank and then comparing the correlations of high grouping and lower grouping. The frequency distribution of the ratio of age to rank has been derived. The median falls at 11.7. Those cases with values below 11.7 belong to lower grouping, iJL, the group who are young by comparison to others at the same rank. Those with values above it belong to the group who are old by comparison to others at the same rank. Table D3 in Appendix D shows the Correlations of promotion and S.I. in work for the two groups. This correlation is not significant for the higher group (r = 0.22) and neither is the correlation for the lower group (r = CL08). On the basis of these correlations, we reject hypothesis 2a. Although it might be noted that the difference is in the predicted direction. 131 Table 4:4. Zero order Correlations of Achievement and Orientation Variables for Upper level Managers (N=35) PROMOTION 1.00 S.ESTEEM -.39** 1.00 S.I. IN WORK -.19 .59** 1.00 FLEXIBILITY -.35** .55** .29** 1.00 UNIVERSALISM -.26* .58** .52** .35** 1.00 PARTICIPATION -.42** .79** .33** .47** .44** 1.00 JOB SECURITY -.29* .51** .22* .32** -.39** .46** 1.00 LENGTH .28* -.55** -.01 -.41** -.36** -.65** -.49** 1.00 AGE .001 -.10 .03 -.02 .03 -.10 .30* .25* 1.00 POWER -.05 .34*‘ .27* .03 .16 .28** .37** -.12 -.29** 1.00 EDUCATION -.23* .34** -.01 .15 -.21* .20 .26’ -.17 -.22* .36** 1.00 *P<.05, **P<.01 132 Table 4:5. Zero order Correlation Matrix for Middle level Managers (Lower diagonal N=37) and for Lower level Managers (upper diagonal N=29) 1 2 3 6 7 9 10 11 1 PROMOTION .07 .18 .24 .05 -.33* .15 .31 .33‘ .07 .15 2 S.ESTEEM -.33* .29 .31 .06 .04 .16 .15 .10 .09 .36* 3 S.I. IN WORK .08 .41* -.06 .06 .01 .59** .20 .26 .24 .19 4 FLEXIBILITY .36* -.28 -.01 .05 -.09 .03 .05 .29 .08 .35* 5 UNIVERSALISM .19 -.38* -.36* .28 .04 -.31 .13 -.18 -.17 .05 6 PARTICIPATION -.02 - 10 -.03 .06 .10 .06 .31 -.07 .17 .10 7 JOB SECURITY .02 -.14 .06 -.32 .51** -.10 .08 .57** .30 .22 8 LENGTH .17 .07 .01 .05 .17 .58** .004 .21 .23 .19 9 AGE .22 - 20 -.12 .22 .32 .36** -.28 .47** .22 .43* 10 POWER .20 -.23 .01 .09 .01 .13 .09 .35* -.06 .23 11 EDUCATION .07 .08 .24 .39* .09 -.04 -.43** .12 -.08 -.11 * P < .05, ** P < .01 133 The second hypothesis says the older an individual is relative to others at the same organizational level the lower their 8.1. in work. In general, as shown in Table 4:2, age does not significantly correlate with S.I. in work (r = -.02). Does rank serve as some kind of accommodative mechanism in the relationship between age and S.I. in work? Looking at Tables 4:4 and 4:5 we find that rank seems not to have such an influence. For managers in the upper levels of the hierarchy, the correlation between age and 8.1. in work is not significant (re.03). The magnitude of this correlation is such that we can conclude there is no relationship between the age of top level management and their S.I. in work levels. For top level managers (executives) we reject hypothesis 2a in favor of the null hypothesis which predicts no relationship. The situation is similar for the middle level management. Table 4:5 reveals that for middle level managers there is also a non- significant correlation between age and S.I. in work (r: —.12). We also reject hypothesis 2a for middle level management. Table 4:5 reveals a positive but not significant relationship between the age and S.I. in work levels of lower level managers (r=.26). The hypothesis for lower level managers is that the older one is the lower their S.I.:hn work. Since this correlation is not significant we can say that we reject hypothesis 2a for lower level managers. 134 It seems then that for top level managers (executives) age does not affect the S.I. in work. We speculate that other factors may be more pertinent at this level. The perquisites of office such as status, prestige, challenge, and autonomy may be more relevant to one's S.I. in work at the upper levels of the hierarchy. An individual in the upper levels of the hierarchy can look back at what he has achieved and be satisfied with it regardless of age. lkn: middle level management, the perquisites of Office may not be as many. The older one is at middle level management the less likely that they will achieve these perquisites and moreover they do not have much to look forward to in career achievements. This may account for the negative association between age and.S.I. in work for middle management. For those at the lower levels of management, older individuals may have just given up hope for attaining certain organizational rewards. Table 4:6. Relationship between age and rank. RANK AGE Under 30 31-40 years over 41 years Lower management 7(26.2%) 15(55.4%) 5(19.5%) Middle management 10(26.3%) 21(55.3%) 7(18.4%) Upper management 4(11.1%) 17(47.2%) 15(41.7%) Totals 21 53 27 135 These explanations of relationships in hypothesis 2a are reinforced if we look at the distribution of management according to age. Table 4:6 shows the relationship between rank and age. The majority of our managers (53%) are middle aged (that is between 31 and 40 years). Middle—aged managers are more or less equally distributed among the three management levels. Thus 15 (i.e. 28%) of these are in lower management, 21 (39%) are in middle, and 17 (31%) are in upper level management. The distribution of younger (below 30 years) and‘older (above 40 years) managers in the three different management levels is more skewed. Fifty six percent, 26%, and 19% of older managers are respectively in upper, middle, and lower level management. For those under 30 years of age, 33% are in lower, 48% are in middle, and 19% are in upper level management. Comparing the age distribution of lower and upper level management, there are more young managers at the lower levels than at the upper levels. Even more, there are more older managers at the upper level than at the lower level. It seems many people at the lower management level are young enough or may be at an early enough career stage, to hope for eventual mobility which in turn may serve to sustain their self investment in work. Many older managers are at the upper level of management. This may matter less, because those at the upper levels of management may be at a level which in itself sustains S.I. in work irrespective of age. Why the negative 136 association between age and S.I. in work for middle management? One possible explanation is that middle level managers are older than managers at other levels.'This is however not the case as more older (i.e. over 41 years) managers exist at upper levels of management (i.e. 56%) than at middle levels (i.e. 26%). Another possible explanation may be that middle management may have a wider range of age than other levels. The age range at lower and upper management may be insufficient to produce significant correlation. A comparison of age distribution among the three levels lends little credence to this argument. For instance the age distribution in middle management (14% 26.3%, 55.3%, and 18.4% for younger, middle, and older individuals) is not dissimilar to age distribution in lower management (i.e. 26.2%, 55.4%, and 19.5%) for younger, middle, and older individuals). It seems to us that it is the pressures at middle management rather than age distribution which account for the fact that the direction of the association between age and S.I. in work is negative in mid-management. HYPOTHESES SET II Hypothesis 1: "Participative and flexible work orientations positively vary with number of promotions and rank" This hypothesis is aimed at gauging the effects of 137 promotions on participative and flexible work orientations. The regression coefficients in Table 4:1 relating promotion to participative and flexible work orientations turn out to be insignificant (r=.382 and r=-.128 respectively). Zero order correlations in Table 4:2 also reveal the same picture with regard to the relationship between promotion and flexibility and participativeness. The correlation between promotion and flexibility is not significant (r=-.07). The correlation between promotion and participativeness also turns out to be not significant (r=-.l4). The magnitudes of both correlations are extremely low, and we canassume no relationship between promotion and the work orientation variables of participativeness and flexibility. As a matter of fact we have a negative variation; therefore wereject hypothesis 1 in favor of the null hypothesis. As we argued when we were testing the effect of promotion on S.I. in work, the assumption that Wpromotions" inevitably lead to moving up the ranks of the organization may, at least according to these data, be ill founded. We state this on the basis of the low correlation between promotion and rank (r=.13). However, this may only be so assuming that promotion as measured here is valid. There is a possibility that it is not. On the other hand rank, at least as it represents the job rewards we supposed come with promotion, emerges to have very strong positive effects on participation and moderate positive effects on flexibility. 138 Table 4:3 shows regression coefficients of rank on both flexibility and participativeness. The coefficients are r=.47 for the equation involving rank on flexibility and r=u71 for the equation involving rank on participativeness. These magnitudes, especially the one involving participativeness, suggest that the higher one's rank, the more participative and to a lesser extent the more flexible they are. The zero order Pearson correlation coefficients of rank on flexibility'(r=a47) and of rank on participativeness (r=.38) essentially reveal the same picture. It seems that at the upper levels of the hierarchy the concentration of job rewards such as autonomy, creativity, and challenge, may serve to make high level management more flexible and participative. Hypothesis la: "The longer a manager stays at one level of the hierarchy the less participative and flexible he/she is." This hypothesis is aimed at investigating how length of stay in present job affects an individual manager's flexibility and participativeness relative to others at the same hierarchical level. In general looking at Table 4:2, length of stay is negatively related to flexibility (r=-¢48) and is negatively related to participativeness (r=-.68). Both these correlations are significant. The magnitude of the correlation between length of stay in present job and participativeness is quite high and therefore we can say 139 that the longer an individual stays in one job the less participative they are. So is the magnitude of the correlation between length of stay and flexibility. The longer one stays in present job the less flexible he/she is. Does this picture stand when we consider managers who are at the same hierarchical level? Turning to Tables 4:4 and 4:5, for upper level managers the length of stay in current job significantly but negatively correlate with flexibility (r=-.41). The longer one has stayed in present job the more unlikely he/she is to be flexible. We therefore accept hypothesis 1a for upper level managers. For middle level managers, length of stay in present job does not significantly and positively correlates to flexibility (r=.05). For middle level managers the length of stay in present position does not affect an individual manager's flexibility, and therefore we reject hypothesis 1a for middle level management. For lower level managers length of stay also does not correlate with flexibility (r=.05). Why are upper level managers different? The magnitude of the correlation between flexibility and length of stay in present job may be large and significant for upper level managers on account of their variegated exposure to many tasks and functions. Staying in one job for too long reduces "visposure" - a condition allowing for flexibility and participativeness. For lower level managers few may have 140 such variegated exposure and hence the insignificant correlation. Mid management is also characterised by lower visibility and exposure and for them, where their mobility has been arrested, negative reactions to flexibility and participation may develop. The relationship between length of stay in present job and participativeness in some important respects parallels that between length of stay in present job and flexibility for the different ranks. The magnitudes of the correlations are however higher for the correlations involving participativeness. Tables 4:4 and 4:5 show that for upper level managers participativeness strongly but negatively correlates with length of stay in present job (r=-.65). For upper level managers the longer an individual has stayed in present job the less participative he/she is likely to be. Therefore for upper level managers we accept hypothesis 1a with regard to participativeness. On the other hand, for middle level managers, participativeness strongly and positively correlates with length of stay in present job (r2u58). For middle managers the longer one stays in present job the more participative they are likely to be. Crozier (1971) talks of the middle level manager as a temporizer and/or diplomat who "must take into account the diverse interest and the contradictory attitudes of those he must direct and with whom he must collaborate. He must necessarily act as a temporizer and as a diplomat" (Crozier 141 1971 p. 130). A study by Child and Partridge (1982) also reveals that middle level managers are caught in the middle between other parties who disagree. Typical comments Child and Partridge (1982) obtained when they interviewed middle management include "I think it is part of my job to have pressure from the top and from the bottom," "You can be pulled in all directions," "You have got to keep both sides happy" (Child and Partridge 1982 p. 91). Katz and Kahn (1966) also talk of middle level management as having "a two way orientation"--that is they have to take cognizance of subordinates and superordinates in their actions. The middle level manager may learn to do this (be a temporizer and/or diplomat) the longer he/she has stayed.in his/her present position. This is inherently a participative process. A middle level manager cannot be tactful or appear to give compliance or collaboration to others who often may have diverse interests without letting them in, on what he/she does. The art of compromise takes time to learn and the longer one is in a given job the more experience he/she may have at compromise. We think this lies at the base of the high correlation between length of stay in present job and participativeness. The correlation between length of stay in present job and participativeness is not significant for lower level managers (r=.31). The positive nature of this correlation may reflect the fact that at lower levels of management, individuals also have to negotiate and 142 compromise a lot in face of diverse demands from their superiors. Hypothesis 2: "Authoritarianism, rulemindedness and favoritism positively vary with job insecurityJ' Another way of stating this hypothesis is that participativeness, flexibility, and universalism positively vary with job security. This is because, as we may recall, we conceived of participativeness, flexibility, universalism and job security as the exact opposites of authoritarianism, rulemindedness, favoritism, and job insecurity respectively. We also operationalised each of these opposites as lying on the same continuum. Table 4:2 shows zero-order Pearson correlations for job security with flexibility, universalism and participativeness. The correlation between job security and flexibility is positive and significant (r=.37) and so is the correlation between job security and universalism (r=.39). The correlation between job security and participativeness also turns out to be significant and positive (r=.53). This means that job security positively varies with flexibility, universalism and participativeness, or conversely, job insecurity positively varies with authoritarianism, rulemindedness and favoritism. The correlation, involving rulemindednes and favoritism are weak (i.e. .37 and .39) and therefore a little caution may be in order. Overall, however, Hypothesis 2 is accepted. We think that these low or weak correlation may be reflective of the 143 accomodative workings or influences of rank. A look at Tables 4:4 and 4:5 shows that job security quite strongly and positively correlates to participation, flexibility and universalism (r=.46, .32, .39 respectively) for upper level managers. The case is different for middle level manager where the correlation between job security on one hand and flexibility and universalism is not significant (r=-.32, -.10 respectively). The correlation between job security and participativeness is significant but negative (r=-.51). For middle management the greater the security the lower the participativeness. For lower level managers the correlations between job security and flexibility (r=-.o3) and between job security and universalism (r=.31) turn out to be non- significant. The correlation between job security and participativeness (r=.06) although, positive is also not significant. We think that the correlations are significant for upper level managers because security at this level adds to confidence which in turn allows one to be flexible and to involve others in what they are doing. At upper management peOple strive for greater and greater visibility and exposure. "Visiposure" is best done by involving others in work and by being flexible with regard to work tasks. Middle level management again show a unique relationship. For them job security does not lead to participativeness. It is only when middle management is threatened with dismissal that it is prepared to involve others in decisions they make. 144 Granick's (1976) observes that because French managers are not easily fired they tend not to accept the company's goals as their own and have little reason to let lower level personnel participate in the decision making process - may be more applicable to middle management than to other levels of management. This is related to Jenning's (1971) discussion of shelf sitters (people whose mobility has been arrested) who he contends are to be found mostly in middle management. Job-security, Jennings indicates, may be the compromise struck between the firm and the shelf sitter. Job security in this context becomes a reward for loyalty and past services from the company to the individual. Jennings talks in terms of the vast numbers of people at middle management who recline gracefully on corporate shelves. Many corporations have been reluctant to fire a manager whose upward mobility has become arrested and level of performance has lowered appreciably. A kind of humanitarian ethic has developed during the decades that, if a manager has given the best years of his life to the corporation, he will not be fired or have his financial security jeopardised (Jennings, 1971, p. 75). Jennings uses the epithet "burn proof" to describe this situation. Burn proof is security compromised by reduced chances for advancement. It is likely to lead to sour grapes attitude not conducive to involving others in decisions. People in this situation may also develop an alienated attitude towards work. As Jennings (1971, p. 80) recognises, "shelf-sitters who resign themselves to their fate tend to gradually decrease their investment of energy, but they hope 145 to increase their benefits and rewards" Our data does bear on the possibility that in such managers there develops an instrumental attitude. When we look at the relationship between job security and 8.1. in work in Tables 4:4 and 4:5, we see certain patterns. We assume that an instrumental attitude towards work is the opposite of S.I. in work. For upper level managers job security leads to some S.I. in work (r=.22 in Table 4:4). Job security leads to S.I. in work in greater proportion for lower level managers (r=.59 in Table 4:5L.But atiniddle management, job security has little if any relationship to 8.1. in work (r=.06 in Table 4:5) Job security, we theorize, reduces S.I. in work (there by producing instrumentalist attitudes) and therefore engenders lack of participation at the middle managerial level. Consequences of Self-Esteem and S.I. in Work So far we have dealt with hypotheses sets I and II. These two sets deal with promotional and rank effects. We now turn to hypotheses sets III and IV. These sets of hypotheses in a large measure relate self esteem, and S.I. in work (which we perceived as moderating variables in our model) to work-orientation variable. HYPOTHESES SET III Hypothesis 1: "There is a lower relationship between promotion and self-esteem for managers with a lower S.I. in work than for managers with a higher S.I. in work" 146 According to this hypothesis we expect a lower relationship between promotion and self esteem for managers with lower levels of S.I. in work. The argument is that promotions are more likely to affect an individuals self esteem when the activities which lead to promotion (in this case individuals work performance) are seen as important by the individual managers or relevant others. In line with this argument we divide our sample into those with low S.I. in work, those with moderate 85L.in.work and those with high S.I. in work. These divisions are based on scale scores from Likert items dealing with S.I. in work. On a scale of 1 to 5, the low S.I. in work group scores 2 or less, the moderate S.I. in work group scored around 3 points, and the high 8.1. in work group had a score of not less than 4. Table 4:7 below shows the effect of promotion on self esteem holding 8.1. in work constant. Before we discuss the implication of Table 4:7 for hypothesis 1 we should consider the relationship between self esteem and 8.1. in work. In our introductory chapters we argued that self esteem and S.I. in work do not necessarily vary together. Table 4:7 Regression coefficients of promotion effects on self-esteem holding for S.I. in work Promotion S.I. in Work Self Esteem Promotion Low .13 (.10)R2 = 01,F = .16 Promotion Moderate -.39(-.10)R2 = .01,F = .60 Promotion High -.60(-.13)R2 = .018,F = .07 * P < .05 ** P < .01 Standardized Betas are in parentheses 147 In these data, however, S.I. in work and self esteem have a very high and significant correlation (r=.55). When we divide our sample according to rank, this picture changes. As Tables 4:4 and 4:5 show, the high correlation between S.I. in work and self esteem seems to be more applicable to top level management. For upper level managers the correlation between S.I. in work and self esteem (r=.59) suggests that individuals with S.I. in work are more likely to have high self esteem. This correlation is also significant for middle level managers (r=.41). The correlation itself has a magnitude which is high and hence suggesting that there is quite a strong relationship between SJL.in.work and self esteem among middle level managers. The correlation between 8.1. in work and self esteem is not significant (r=.29) for lower level managers. Except for lower level, according to these data then S.I. in work and self-esteem appear to vary together even when we take rank effects into account.‘Table 4:7 shows the coefficients of the effects of promotion on self esteem for the three S.I. in work groups. As we argued earlier with regard to use of regression analysis our concern is with ascertaining the impact of given variables over others. To this end our focus is on unstandardised beta coefficients. Table 4:7 shows that the impact of promotion on the self-esteem of people with moderate and high S.I. in work is negative. Moreover the variance in self-esteem explained by the promotion variable 148 is in these equations very low ranging from 1% to 1.8%. None of the three equations turn out to be significant. We therefore accept the null hypothesis which posits no difference in the effect of promotion on self esteem for low and high S.I. in work groups and reject hypothesis 1. A possible explanation is that we have too few cases to obtain a significant relationship. Another explanation for this result is identical to the one we proffered earlier for the lack of promotional effect. It seems that promotional effects either occur infrequently or occur in certain circumstances. That is we see a need to specify types of promotions and conditions under which they have an effect. Here we have in mind such things as the meaning promotions take to the individuals who experience them, whether promotions lead individual to move from lower to middle to upper levels of the organization's hierarchy (i.e. promotions across hierachies) and the context (relative promotions occurring) in which such promotions take place. Context may help define type of promotion in the sense of distinguishing between promotions which occur in places (hierarchy) in which promotions are rare; promotions which occur where promotions are frequent and promotions which occur in hierachies where promotions are very frequent. Hypothesis 2: "Self-esteem is inversely related to authoritarianism, favoritism, and rule-mindednessfl' 149 Another way of stating hypothesis 2 is to say that, self esteem is positively related to participativeness, universalism, and flexibility, whiCh are work orientations functionally opposite to authoritarianism, favoritism and rulemindedness in our model.‘Table 4:2 shows correlations of self esteem to the work orientation variables of participativeness, universalism and flexibility. All these correlations are high, positive and significant. The correlation between self esteem and flexibility'is r=.55, between self esteem and universalism is r = .56, and between self esteem and participativeness is r =.72. Thus on the whole individual manager's flexibility, universalism, and participativeness requires a high degree of global self- esteem. Individuals with low self esteem may find it difficult to distribute their authority to others. They are less accepting of themselves and ultimately less accepting of others - and are therefore less willing to involve others in what they do. Authoritarian executives perceive a certain hostility from their subordinates (Crozier 1971 p. 122). Authoritarianism breeds what Crozier (1971) refers to as the phenomenon of "selective perception." This may be enhanced by a strategy whereby one sticks to rules (rulemindedness) and uses them to reward those who adopt a strategy'of least resistance (favoritism) towards his/her actions. On the whole the correlations between self esteem and flexibility, universalism and participativeness are such 150 that we accept hypothesis 2. HYPOTHESES SET IV This set of hypotheses relate self investment in work to work-orientations. Again our main statistical instrument here remains the zero-order Pearson correlation matrices. Hypothesis 1: "Participative, flexible and universalist work orientations are positively related to S.I. in workJ' According to this hypothesis we expect managers' participative and universalist work orientations to rise with their S.I. in work. The zero-order correlations between S.I. in work and participative and universalist work orientations shown in Table 4:2 essentially confirms that this is the case. The correlation between S.I. in work and universalism is positive and significant (r=.44) and so is the correlation between S.I. in work and participativeness (reu32). These correlations suggest that as S.I. in work rises, so do managerial participative and universalist work orientations. Thus according to these data, hypothesis 1 is confirmed. It seems as Hall and Rabinowitz (1977) describe the job involved person - he/she participates in decisions affecting him or her and by that very token is open to letting others participate in decisions affecting them. The relevant question here is - what accounts for the strong, positive covariation between S.I. in work and universalism and participativeness? A participative, just like a 151 universalist work orientation serves to improve the interpersonal climate of the working place, thereby helping in the fulfillment of social needs of both subordinates and superiors and in this way providing a climate in which individuals' 8.1. in work may occur. Does hypothesis 1 hold for managers at different hierarchical levels? A look at Table 4:4 reveals that for upper level managers S.I. in work highly and positively correlates with universalism and participativeness (r=.52, r=.32 respectively). Table 4:5 reveals that for middle level managers S.I. in work moderately but negatively correlates with universalism (r = -.36L.The correlation between S.I. 1J1 work and participativeness is, on the other hand, not significant (r=~.03) for middle level managers. For lower level managers S.I. in work, is not significantly correlated to both universalism and participation (r=.06, and r=.01 respectively). Why is S.I. in work positively and significantly correlated in) universalism and participativeness for upper level managers and not for those in the lower levels of the hierarchy? Researchers have contended that intrinsically motivated individuals should show higher job involvement than extrinsically motivated individuals when the job meets their respective salient needs (Kanungo 1982). Empirical research has generally supported this contention (Kanungo 1982, Hall and Rabinowitz 1977). Upper level managerial jobs do satisfy intrinsic needs to the extent 152 they offer greater autonomy, variety and challenge than lower level managerial jobs. This is the reason Tannenbaum (1974) argued that those who hold high level jobs in the organization show more or greater levels of job involvement. Upper level managers do invest themselves in work more because they enjoy greater autonomy, challenge, and variety in their jobs. This autonomy and variety, in turn, allows them to be more participative and universalist in orientation.10n the other hand, middle level managers exist in a situation of ambiguity which requires them to balance contradictory and diverse interests (Crozier 1971, Child et a1. 1982, Watson 1977). Not only may their jobs fail to satisfy intrinsic needs but their roles engender ambiguity. In this situation it may be difficult to apply objective standards (universalist criteria) in dealing with situations. We believe this is behind the negative correlation between S.I. in work and universalism. Lower level manager jobs may not have this "ambiguity" which many have attributed to middle management jobs, but all the same their jobs are unlikely, in relative terms, to satisfy intrinsic needs. At lower levels, managers are expected to be administrators in the sense of using existing structure, 14%, they should see to it that rewards and sanction are carried out fairly and that everyone does what they are supposed to do. Their jobs deal in extrinsic factors. This may lie behind the low correlation between S.I. in work and 153 universalist and participative work orientations for lower level managers. Jobs at the lower levels of the management hierarchy may not be too different from manual work in the sense of being organized on an instrumental basis (i.e. organized around the attainment of certain objectives). As a group, lower level managers may be a focus of control very different from other groups of managers. In so far as their autonomy may be relatively limited, so is their participation in decisions which are strategic to the company. People who do not participate in certain activities may find it difficult to involve others in what they do. Hypothesis 2: "The longer an individual stays in one job, the less their flexibility is associated with S.I. in work." This hypothesis is aimed at investigating the influence of length of stay in present job on the relationship between S.I. in work and flexibility. The hypothesis is based on the expectation that an individual who has stayed too long in one job may become so psychologically involved in it (high ELI. in work) as to find it difficult to do tasks not so related to the job and hence his/her flexibility may be impaired or negatively affected. As shown in Tables 4:8 and 4:9, we have divided our sample into three groups: those with less than 2 years in present job (N=39), those with 3 to 4 years in present job (N=l8) and those with over 5 years in their present jobs 154 (N=44). Table 4:8 shows that the correlation between S.I. in work and flexibility for managers with 2 years or less in present job is not significant (r=.17). The correlation itself is so low in magnitude we take it that there is no relationship between the variables. The correlation between S.I. in work and flexibility for managers with 3 to 4 years is shown in Table 4:9. It is very low and insignificant (r=-.13). We conclude that for individuals with 2 or less years in present job and for those between 3 and 4 years in present job, S.I. in work does not seem to relate to flexibility. On the other hand, the correlation between S.I. in work and flexibility is positive and significant (r=.31) for those with 5 or more years in present job. The magnitude of this correlation suggests that flexibility at work rises with S.I. in work for individuals with five or more years in present job. On the basis of these correlations we reject hypotheses 2 which suggested that flexibility and S.I. in work are more likely to be significantly related the shorter one has stayed in present job. How are these contrary results to be explained? The impact of job conditions on individual psychological functioning may not be an overnight process. Job conditions are not readily modified to suit the needs or capacities of the individual worker. Over a long enough time, however - certainly over a period as long as 10 years - many men either modify their jobs or move to others more consonant with their functioning" (Kohn and Schooler 1978 p. 48). The reciprocal effects of job condition and individual 155 functioning may not be contemporaneous. A stay of 5 or more years in a given job may serve to ensure that one's job skills and investments are not underutilised. Adaptability or flexibility to a situation may require familiarity with the nitty gritty of the job situation, and this may lie behind the high and significant correlation between S.I. in work and flexibility for those with five or more years in present job. This reasoning is indirectly supported in that the most significant relationships are found among the 44 people who have been in their jobs for more than five (5) years. Table 4:8 Zero order correlations of work orientation variables for managers with 2 years and less in present position (N = 39) 1 2 3 4 5 6 7 1. PROMOTION 1.00 2. S.ESTEEM -.09 1.00 3. S.I.IN WORK .17 .45** 1.00 4. FLEXIBILITY .25 .28* .17 1.00 5. UNIVERSALISM -.11 -.09 -.10 -.15 1.00 6. PARTICIPATION .14 .40** .40** .29* -.24 1.00 7. JOB SECURITY -.12 .41** .34** -.10 .01 .28* 1.00 *P < .05 ** P < .01 156 Table 4:9. Correlations of work orient. Variables for managers with 3-4 years (lower diagonal) and for managers with over 5 years in present job (upper diagonal. N = 44) 1 2 3 4 5 6 7 1. PROMOTION -.14 -.15 -.24 -.16 -.22 -.05 2. S.ESTEEM -.28 .61** .S4** .66** .80** .52** 3. S.I.IN WORK -.01 .34 .31** .59** .28** -.25** 4. FLEXIBILITY .25 .06 -.13 .47** .45" .41** 5. UNIVERSALISM .17 -.03 -.13 -.35* .49** .38** 6. PARTICIPATION .84 -.06 .26 -.15 .39* .52** 7. JOB SECURITY .02 .15 .05 -.01 -.10 .08 *P<.05 **P<.01 A note on education Education is correlated to a number of factors. The higher one's education the higher their self-esteem (r=.36), the higher their flexibility (r=.29L. the higher their universalism (r=.40), the higher their job security (r=.34) and the higher their rank (r=.60). Education significantly but negatively relates to length of stay in a given job (r = 157 -.41),and to power (r=-.23). Education does not seem to relate to S.I. in work, age and promotion. The negative relationship between power and education is baffling to us. It may relate to the issue of legitimate as distinct from nonlegitimate power. In so far as power is nonlegitimate it may have a negative relationship to education. Education does not seem to have independent effects of its own on these variables. In regression model 3, appendix B, we have controlled for achievement effects and for theleffects of other personal variables. The result is that none of the regression equations involving education are significant. The lack of independent effect may relate to the fact that most of the managers included in this sample are fairly well educated. Only 10 managers did not complete high school. Of the 33 who completed high school 11, 13, and 9 are respectively in lower, middle and upper management positions. The remaining 58 have been to college, and they are more or less evenly distributed among the managerial levels. Hypotheses Set V Hypothesis 1: "S.I. in work is inversely related to powerlessness among managersfl' We argued, in support of this hypothesis, that feelings of powerlesness on the part of the individual manager do not engender a positive climate for S.I. in work. Having power 158 or perceiving oneself as powerful on the job fosters S.I. in work. The relationship between power and.S.I. in work is partly shown in Table 4:2. The correlation between individuals' power in the organizational setting and their S.I. in work is both positive and significant (r = .25). The regression equation involving power and S.I. in work is positive and significant (r=.23 regression model 1 appendix D). This is so when we hold other achievement factors constant. Power still positively and significantly affects 8.1. in work (r=.22) when we control for the effects of personal variables (iJL education, and age in regression model 3 appendix D). This relationship is strong enough for us to conclude that the more power an individual reckons he or she has in the organization, the more S.I..Ln work they exhibit. A cross-tabulation of the relationship between the two variables is shown in Table 4:10. Table 4:10 does show that in general most of the managers we interviewed do evince relatively high levels of 8.1. in work. More specifically, the overwhelming majority of individuals who see themselves as exercising considerable power in the organization (i.e. those with a score of 4 or 5 on the power scale) also have a high S.I. in work. Thus 26 of the 32 "powerful" individuals also show a high S.I. in work. For those with a minimum score on the power scale, three out of five show low S.I. in work. The picture is more even for those in the middle. Of the 40 individuals who see 159 themselves as exercising power in moderate level (i.e. with a score of 3 on power scale) 21 have low S.I. in work while 25 have high S.I. in work. In general, it seems the more power an individual manager has the higher their S.I. in work. Both power and S.I. in work are evaluative processes. A.powerful manager in an organization is one who Table 4:10 The S.I. in work levels of managers with differing degrees of powerfulness. Power Self investment in work Total "'2 """"" 5 """" 2 """"" é """ 1 0 3 2 0 5 2 0 4 l3 1 18 3 3 18 23 2 46 4 1 5 22 2 30 5 0 0 2 0 2 ’ "" ESE; """ Z """" 36 """ 32 """"" E """"" 161" is able to negotiate resources for him/herself and for others as a matter of role obligation. Legitimate- powerfulness in itself is a convincing demonstration that an individual managers actions are appropriate to the credibility (based on successful performance of their role obligations) they have staked for themselves in the organization. Other individuals in the organization must be willing to acknowledge one's performance of role obligations 160 as successful. On the other hand, an individual manager's successful performance of role obligations and its recognition as such by relevant others is crucial to the process of S.I. in work. A high positive evaluation of an individual's performance brings forth a high level of S.I. in work for such individuals. Legitimate power in the organization is in part a high and positive evaluation of an individuaiis credibility and "ability to get things done". As such it also seems to foster one's self investment in work. Hypothesis 2: "Authoritarianism, favoritism, and rule mindedness are positively related to powerlessness among managers." The concepts of authoritarianism, favoritism, and rulemindedness are respectively functional opposites of participativeness, universalism and flexibility. The correlations, in Table 4:2, between power and flexibility, and between participativeness and universalism are all positive. The correlation between power and flexibility is not significant at (r=.05); between power and universalism is also not significant (r=.13); but the correlation between power and participativeness is significant (r=.21). According to these data there is little or no association between the power individuals have and their flexibility. Although the magnitude is small, the more power an 161 individual has the more participative he/she tends to be. In a sense, power implies an increase in participation both in decision making and over the working place. The multiple regression model in appendix D reveals that the effect of power on participation, though significant, is small (r=.13) in regression model 1. Where we control for S.I. in work and self-esteem (regression model 2) this effect disappears, suggesting that power mainly affects participation through other variables. The effect of power on participation also disappears when we control for personal variables (regression model 3 in appendix D). Thus we need to be cautious in associating power to participation. A legitimately powerful manager is one with the credibility necessary to successfully negotiate resources necessary for carrying out organizational actions. For one to "get things done" — which is, as we have argued, what power is about - they must count on the cooperation and involvement of people at various levels. Such cooperation may be best affected through consultations with various people involved in the work process at different levels, it hence is inherently participative. Power may be seen as both legitimate and non legitimate. Where it is not legitimate consultations may be less necessary in organization action. As a result people with less legitimate power do not need to be flexible in dealing with individuals who may have different work conditions which usually are the basis for different 162 subcultures within the work placeuand therefore different situational adjustments among different work groups. Securing the cooperation of such groups may not so much require flexibility as threat of punishment by those with less legitimate power. Kohn et a1. (1983) found that powerlessness is related to close supervision, routinised work, and work which has little substantive complexity. Less legitimate power may operate in similar ways. The point is that substantive complexity; nonroutine work. and freedom from close supervision are determinative of occupational self direction (i.e. application of initiative and independent judgement at work) which in turn is a precondition for flexibility. The relationship between power and favoritism may be more direct. An individual manager who tends to favor this or that individual at work as those without legitimate power may be prone to doing, may find it very difficult to secure compliance or enthusiastic participation of others in projects under his/her direction. Related hypotheses Hypothesis 1a "Powerful managers in organizations with high total control invest themselves in the job more than powerful managers in organizations with low total controlJ' Earlier we stated the rationale behind this related hypothesis that organizations with a high total amount of control are also the ones with greater integration. Greater integration translates into increased mutual influence 163 and/or consensus among members. This enables those with power, in organizations with a high total amount of control, not to be bogged down in solving labor problems, thus freeing them for greater involvement with aspects of their jobs.'Table 4:11 shows in part the correlatons between power and 85L.in.work for people in firms with low, medium, and high total amount of control. Table 4:11 Correlations of power on S.I. in work and work- orientations variables according to type of control in the firm. l I I | S I IN WORK | FLEXIBILITY | UNIVERSALISM | PARTICIPATION I I I I I I I I I | I I I I I I TYPE OF I | I I I l | I l I | I CONTROL | HIGH|MEDIUM|LOW|HIGH|MEDIUM| LOWIHIGHIMEDIUMI LOWIHIGHIMEDIUMI LOW IN FIR" I I I I l l | I I I | | I I |___| I | | I I | l | I | I I I | I I I I l I POWER |.49**| .13 |.os|-.01|-.31* |-.22| .05| .23 |-.19|.28*| .17 | -.07 I I |___| I | I | I | I | *’ P’ < .(35 **P < .01 There were 38 individuals in firms with high total control, 34 in firms with medium total control, and 29 in firms with a low total amount of control. The correlations in Table 4:11 does confirm hypothesis 1a, namely we found a high and positive correlation between powerfulness and 85L in work for individuals who work in firms with a high total 164 control (r=.49). A correlation of this magnitude suggest that the more power an individual manager has in a firm with a high total amount of control, the more they invest themselves in their work. On the other hand the correlations involving individuals in medium and low total control firms are, although positive, low and.1nsignificant (r=.13 and .05 respectively). Individual managers in firms which have medium and low total control reveal little or no relationship between power and S.I. in work. It seems then that the amount of control in a firm is consequential with regard to the extent to which those with power invest themselves in their work activities. Firms with a high total control may be more integrated. Powerful managers in these firms do not need to deploy their efforts solving labor problems and may thus have greater chance immersing themselves in their work. Hypothesis 2a: "Authoritatianism, favoritism, and rule- mindeness will be more positively associated with powerlessness among managers in firms with low total control than in firms with high total control)’ Related hypothesis 2a in part predicts that powerless managers will be less universalist (more favoritist) in orientation in firms with low total control than in firms with high total control. Table 4:11 reveals correlations involving powerfulness and universalism in firms with high, 165 medium, and low total control. Powerfulness does not seem to relate to universalism in firms with high total control (r%.06). Neither is the correlation involving power and universalism significant for individual managers in firms with medium total control (r=.23). Both these correlations between.power and universalism in firms with high and medium control vary in the same direction. They are positively associated. The correlation between power and universalism for individuals in firms with low total control is also insignificant but negative. It seems that these correlations are insignificant because in general power and universalism have little or no relationship according to these data (r=.13 and shown in Table 4:2). The two variables do not relate to each other even when we control for the effects of other achievement factors (r=.09 in regression model 1, appendix D). They also do not associate when we control for the effects of gender, education, and age (r=.05 in regression model 3, appendix D). Table 4:11 also shows the correlations of power and flexibility in the three types of firms. The correlations in two types of firms turn out to be insignificant. The correlation involving power, and flexibility is insignificant in firms with high as in firms with low total control (R=.01 and -.22 respectively). This suggests that there is little or no relationship between being powerful and being flexible in organizations which have high or low 166 total control. This may reflect that the two variables are generally not related as shown in Table 4:2 (r=.05). Regression equations in appendix D, of power and flexibility also are insignificant even when controlling for other factors. On the other hand power is significantly but negatively correlated to flexibility in firms with medium total control (r=-.31). In firms with medium total control, as a manager's power increases their flexibility deminishes. This is an unexpected finding; and we do not know how to interpret it beyond speculating that it may relate to the distinction between legitimate and nonlegitimate power. An individual with nonlegitimate power may not care much about being flexible in their dealings at work. On the whole we fail to confirm hypothesis 2a with regard to rulemindedness or flexibility. The correlations of power and participation as shown in Table 4:11 conform to our expectations in hypothesis 2a. Power and participation are correlated for individuals in firms with a high total control (r=.28) Although it is positive the correlation of power to participation is weak and insignificant in firms with medium total control (r=.17). This relationship is insignificant but negative for managers in firms with low total control (R=-.O7). Powerful managers in firms with high total control are more participatory than their counterparts in firms with lower total control. Hypothesis 2a is confirmed with regard to 167 the relationship between power and participation. The magnitude of the correlation is rather small and this calls for caution. The small magnitude may relate to the possibility that power affects participation mostly through other factors. Thus where we control for S.I. in work, self- esteem, education, gender, and age (see regressions in appendix D) the effects of power on participation deminish. A multiregressional analysis of the data For a general assessment of the working model in the introductory chapter we have used multiple regression. We have three such regression models (as shown in appendix D). The first two regression models reflect the linkages specified in the working model. There are two general linkages in the working model: First is the extent to which career experience (achievements at work) affect work orientations. Regression model 1 represents this linkage. The independent variables are the three achievement variables of power, rank, and promotion. These are regressed on the moderating variables of self-esteem and Swl. in work and on the work orientation variables of flexibility, universalism, and participation. The second general linkage in the working model is the extent to which career experiences (achievement factors) affect work orientations as mediated by self-esteem and S.I. in work. Regression model 2 represents this linkage. Here, to the independent variables of power, rank, and promotion are added S.I. in 168 work and self-esteem. Flexibility, universalism, and participation remain the dependent variables. Regression model 3 has S.I. in work and self-esteem as dependent variables together with work orientations. However to the independent variables of power, rank and promotion are added gender, education and age as variables likely to influence the specified relationships. A multivariate analysis is in part being used as a way of estimating the relative magnitude of direct as opposed to mediated linkages which.sre of interest in the working model. In regression model 1, representing the direct effect of achievement on orientation variables, promotion significantly but negatively affects self-esteem, universalism, and participation, however these coefficients are quite low (r = -.25, —.19, and -.23 respectively). The low coefficients call for caution in saying that promotion leads to lower self-esteem, universalism, and flexibility. Moreover, we are not sure about the measures of the promotion variable. In regression model 1 we also see that power positively and significantly affects S.I. in work, self—esteem, and participation. These coefficients are also low (r=.20, .23 and .13 respectively). These low coefficients makes us cautious to conclude that the more power one has the greater the SOI. in work, self-esteem, and participation. On the other hand, the regression coefficients involving rank are quite strong. Rank positively and significantly affects 169 self—esteem, flexibility, universalism, and participation (r=.59, .49, .40, .73). Coefficients of this size suggest that as rank rises so do self-esteem, flexibility, universalism, and participativeness. The regression equation involving rank and S.I. in work is not significant when we control for promotion and power (regression model 1). This relationship, although very low, was significant when rank alone was regressed on.S.I. in work (Table 4:3). This suggests that rank affects S;L.in.work indirectly through other variables. When we add S.I. in work and self-esteem to the independent variables we obtain slightly different results. This is represented in regression model 2 which deals with the second linkage in our working model. Here some achievement factors diminish in importance. The coefficient of promotion on self-esteem and the coefficient of promotion on universalism are no longer significant. The coefficient of power on participation is also not significant, neither is the coefficient of rank on universalism. These changes seem to be the result of the relationship between.S.I.:h1 work and self-esteem on one hand and power, rank, and promotion on the other. It seems self-esteem and S.I. in work reflect some variance in flexibility, universalism and participativeness not tapped by work achievement variables or alternatively promotion is related to universalism, power is related to participation, and rank is related to 170 universalism (regression model 1) for other reasons. Regression model 2 reveals that S.I. in work and especially self-esteem have many significant and positive effects on work orientation variables. 8.1. in work positively and significantly affect universalism (r=.23), but has little relation to flexibility and participation (r=.01 and -.02 respectively). Self-esteem positively and significantly affects flexibility, universalism and participativeness. The magnitude of the relationships involving self-esteem are high enough for us to say the higher one's esteem the more universalist, flexible, and participative they are. To the extent that the moderating variables of S.I. in work and especially self-esteem reflect the variance not tapped by achievement variables of rank, power and promotion, some effects (especially those of promotion on universalism, rank on universalism, and power effects) may be more due to self- esteem and S.I. in work. This view is in part supported by the changes in variance explained (R2) with the inclusion of moderating variables in the equations. Self-esteem and S.I. in work serve to increase the variance explained considerably i.e. from 24% to 35% for flexibility, from 19% to 35% for universalism and from 59% to 68% for participativeness. The significant correlations of S.I. in work and self-esteem with the achievement factors (as shown in Table 4:2) also suggest the latter may work through the former in their effects on work-orientations. Regression 171 model 3 include personal characteristics as independent variables. Rank remains a variable with strong independent affects on some work orientations. Promotions only affects participation. Power independently affects S.I. in work and self-esteem. Education and age do not seem to have significant effects. Gender on the other hand shows strong effects in model 3. This may in part reflect the fact that we only have about 10 females in our sample. CHAPTER 5 SUMMARY, DISCUSSION AND CONCLUSION Summary of problems encountered: Before we summarize what was found in the preceding analysis we should indicate the limitations of the data as we have so far analyzed them. The most important limitation faced is a possible measurement error in the variable "promotion". That is, we did not fully appreciate the difficulties of operationalizing the concept "promotion" as an achievement variable before actually doing the interviews. In our interviews we asked respondents to state the various permanent (full time) jobs they have occupied in their careers. For each job, in their career, respondents were asked to indicate whether the Job was a promotion, a demotion or some kind of lateral transfer. This mode of gauging promotion treats it as a unidimensional concept. That is we deal with a single aspect -the quantity or number of promotions. There are many additional dimensions to be considered in measuring an opportunity variable such as promotion. This unidimensional definition of promotion did not allow the researcher to go beyond individual analysis.It seems clear that promotion measurement should address both 172 173 issues of individual and job analysis. It is by considering job analysis that we can deal with issues pertaining to the "quality of Job promotion". By considering only the number of promotions, we can never know whether the respondent had been: (1) promoted to a low ceiling, low opportunity job which may not allow many individuals to advance beyond the job and thus a first step towards failure to advance through the managerial hierarchy, (2) whether somebody was promoted to their present job in compensation for losing out to competition for the better job, and (3) whether the promotion was to a position where his or her incompetency may be more exposed. It is when we deal with the quality dimension of promotions that we can know the meaning individuals attach to the promotions they receive. A focus on the number of promotions only also implies two questionable assumptions: (1) job ladders or number of steps associated with them are comparable across industries or for that matter across different departments/sections of the same firm. The division of job classifications into a large number of steps may still not mean much opportunity (or achievement) for the pe0p1e in them if the distance between bottom and top (in status, pay, working conditions) remains same" (Kanter, 1979, §L57). Thus in our case some firms may have longer job ladders than others. A promotion in a firm with a long job ladder may not mean the same thing as a promotion in a firm or department with a short job ladder. It is probable that firms with long 174 job ladders have a higher rate of promotion than firms with short job ladders. In cases with higher rates of promotion, promotion may be perceived as less important because it is commonplace. People in firms with infrequent promotions may attach greater significance to them. In our interview schedule we tried to deal with this issue of promotion rates by asking personnel managers - how many people in the firm had been promoted in the last 10 or so years. Unfortunately this is one of the questions we had to cancel because this type of information was treated as sensitive in cases where records were kept.'The second dubious assumption in treating promotions, only in terms of number is that all the Jobs an individual has had are seen as continuous or discontinuous along some meaningful line (Kohn et al., 1983 p.122). As Kohn (1983), following Wilensky (1961) argues, it is more realistic to conceptualise job changes as, in part, representing logical progressions in a meaningful sequence and in part representing shifts out of given career sequences into other career sequences. Unfortunately we know of no satisfactory way of dealing with this issue given the data we collected. There also is the possibility that promotions may not really mean advancement given the current stage of Zambianization of industry. Job-fragmentation and resultant job changes in industry may be taken to represent promotion. 175 Another broad limitation we see with regard to our data is that some of our scales were rather short. Our specific reference here is to the concept of universalism vs. favoritism and flexibility vs. rule—mindedness. After running the reliability subprogram recommended by McIver and Carmines (1981) we ended up with two items showing a strong correlation with each other. We thus ended up with two items as indicators of the concepts in question. We presume that if we had started with many items for each of the said concepts we could have ended up with Likert scales having a lot more items than we did. We feel this could have been of help with regard to some concepts where we suspect insufficient range. These limitations should be taken into account in our assessment and or discussion of implications - practical and theoretical - of the findings we have come up in this study. A summary of the findings: We have tested five sets of hypotheses linking achievement variables in industry, independently as well as indirectly through self-esteem and S.I. in work, to flexible, universalist and participative work orientations. We have mainly relied on the SPSS programs to produce zero- order correlations, regression analysis, graphics and cross tabulation. The findings in this study have varied. In some cases the hypotheses have been confirmed and in other cases not. 176 Hypothese set 1: These hypotheses linked achievement variables at work (rank, promotion, relative age at given rank, length of stay in current job) to S.I. in work. According to these data, promotions are not associated with S.I. in work. That is Pearson Correlations involving promotion and S.I. in work were small and insignificant. The graph depicting the relationship of the means of promotion to the means of S.I. in work showed a somewhat haphazard picture. Apparently, individuals with two promotions have a higher S.I. in work than the rest. This changes as people with three or four promotions show less S.I. in work. The curve falls even more sharply in the case of those with five promotions. The case was similar when we used regression analysis. The unstandardized beta coefficients representing the impact of promotion on S.I. in work are small and insignificant. The promotion variable explains less than 1% of the variance in S.I. in work. According to these data, the achievement variable of rank does relate to S.I. in work. Rank and S.I. in work do positively and significantly correlate (r=.18). Although they are not steep, both the graphs of the regression and the means of rank on S.I. in work reveal that as rank rises so does S.I. in work. We also tested the effects of length of stay in a given Job and the relative age of a manager at a given hierarchical level on S.I. in work. A manager who stays too long in a given job, like one who is older than 177 his/her colleagues at a given level of the hierarchy is not achieving much and therefore may be expected to have less S.I. in work. Our findings do indicate that this is not the case. Length of stay in a Job does not correlate to S.I. in work (r=-.O9). When we take into account the variable rank, the situation does not change. At upper levels of management there is little correlation between length of stay in a given job and S.I. in work (r=-.01). At middle and lower management levels, length of stay in a given job also do not seem to relate to S.I. in work (r=.Ol and -.20 respectively). The relative age of a manager at a given level, generally does not seem to relate to S.I. in work. This is also the case when we take into account rank. For managers in the upper levels of the organization age does not affect S.I. in work. Age is also not important for S.I. in work at lower levels of the managerial hierarchy. At middle level management, age also is not associated with S.I. in work (r=-.12). HYPOTHESES SET II: These hypotheses linked work achievement variables to work orientations. Achievement variables were represented by promotion, rank, length of stay in given job and job security. Work orientation variables included participativeness, flexibility, and universalism. The regression of the promotion variable on participative and flexible work orientation turned out to be insignificant. 178 Promotion explains less than 1% of the variation in flexibility, and about 2% of the variance 1 n participativeness. The correlations of participativeness and flexibility with promotion turns out to be insignificant (i.e. r = -.07 and -.14 respectively). With such small magnitudes it is safe to assume no relationship. On the other hand, rank does strongly and positively relate to participativeness (r = .71). It is also positively, and significantly related to flexibility (r = .47). Flexibility and participativeness do rise and fall with rank. Rank does explain about 22% of the variation in flexibility and in around 50% of the variation in participativeness. These data reveal that in general length of stay in a given job is significantly and negatively correlated to flexibility (r = -.48) and to participativeness (r = -.68). The longer one stays in a Job the less participative they are. The smaller magnitude in the case of flexibility suggests caution. Do these relationships remain when we take rank into account? Some do for upper and lower level managers and not for middle level managers. For upper level managers the longer they stay in a Job the less flexible they are. For middle level managers length of stay in a Job does not affect an individual's flexibility on the job. In upper level management, the longer the individual stays in present job the less participative they are. "The opposite is the case for middle level managers. The longer they have 179 been in their present job, the higher is their participativeness. There is no significant relationship between participativeness and length of stay in present job for lower level managers. What are the supposed effects of job security on work orientations? According to these data, the more job insecurity an individual manager experiences, the more authoritarian, rule-minded and favoritist is their orientation. The relationships between insecurity on one hand and work orientations on the other, are week. We believe this is due to the accommodative working of rank which in itself affects job insecurity. At the upper levels of management job insecurity strongly relate to rule mindedness, authoritarianism, and favoritism. These relationships are not significant for lower level managers. The situation is more varied for middle level managers, for whom Job insecurity leads neither to authoritarianism, nor to rule-mindedness; however, for middle management the more security on the job the less universalist is their orientation towards work. Hypotheses Set III These hypotheses largely dealt with the relationship between achievement, self-esteem, and work orientations. The expectation was that promotions will affect self—esteem more for people with higher levels of S.I. in work than for people with low levels of SJL.in.work. We reckoned that for an activity to affect an individual's esteem it must be 180 perceived as important by that individual. The prediction that there will be a difference in the effect promotion has on self-esteem levels of people with low S.I. in work from those with high S.I. in work was not upheld. For both groups the variance explained by the promotion variable was around 10%. None of the regression equations were significant. In these data self-esteem has a negative effect on authoritarianism, rule-mindedness and favoritism. The more self—esteem individual managers have, the less authoritarian, favoritist, and rule-minded they are. Hypotheses Set IV The hypotheses relate Sal. in work to work orientation variables. These data reveal that in general there is a strong and positive relationship between S.I. in work, universalism, and participativeness. The more S.I. in work a manager has, the more participative and universalist will be their style of management. Again, it seems hierarchical level serves as an accommodative mechanism in these relationships. For upper level managers S.I. in work strongly and positively relates to participation and universalism. For middle and lower level managers these relationships, with the exception of one, are weak and non— significant. The exception is that, for middle level managers, the higher one's S.I. in work the less universalist is their style of management. For purposes of 181 testing one hypothesis in this set we divided our sample into three groups: those who have had their present Jobs for less than two (2) years, those who have had their jobs for three to four (3-4) years and those with over five (5) years in the present job. According to these data and contrary to our prediction the relationship between S.I. in work and flexibility for those less than 4 years in present Job is very weak and insignificant. On the other hand, these data strongly indicate that Sal. in work increases as does flexibility for those in the Job for five (5) years or more . Hypotheses Set V These hypotheses deal with the effects of power (an achievement variable) on S.I. in work and on the work orientations of universalism, flexibility and participativeness. According to these data there is a significant and positive relationship between holding power in the organization and the S.I. in work of individual managers Cr==.25). The more power an individual has the more they invest themselves in work. We found that 25 (81%) of individuals with high scores on the power scale also scored high on the S;L.in work scale. We did not establish any relationship between power and flexibility (r = .05) or power and universalism (r = .13), while power positively and significantly related to participativeness (r = .21). Thus, according to these data, the more power an individual has in 182 the organization the more they are participative in their management styles. In this set of hypotheses we also tested for two related hypotheses. These basically posited the same relationships but taking into account the differential spread of control in firms. We divided firms into three types: those with low, those with medium and those with high total control. Correlations reveal differences in the way powerful managers invested themselves in work for the three different control type firms. We found that power positively and significantly correlated to S.I. in work in firms with a high total amount of control. On the other hand, in firms with medium and/or low total amount of control, correlations were insignificant. The amount of total control in the firm, seems to be consequential for the S.I. levels of those with power. With regard to»participation we fOund that the amount of total control tends to benefit those with power in firms with high total control more than those with power in firms with medium and/or low total control. In this study the amount of control in the organization did not affect the degree to which powerlessness or powerfulness affects individual managerwa universalist and flexibility work orientations. It should be noted that with regard to the relationship between power and flexibility we are baffled by the significant but negative relationship for individual 183 managers in firms with medium total control. We felt that the distinction between legitimate and non-legitimate power may affect flexibility to produce significant differences for individuals in firms with medium total control. Multivariate Analysis Summary The achievement factors have different effects on different variables. Some of these effects seem to be independent while others are through other variables. Power has independent effects on the moderating variables of S.I. in work and self-esteem, while it affects participation through other variables. Power seems to be unrelated to universalism and flexibility according to these data. Rank has independent effects on work orientation variables, while it affects moderating variables (Self— esteem and.SMI. in work) through other variables. Promotion seem to have few independent effects of its own on both moderating variables and work orientation variables. When we control for self-esteem, S.I. in work, and personal variables (i.e. age, education and gender) the effect of promotion disappear with the exception of that on participation. However, we should keep in mind the limitations we believe lies in our measurement of the promotion variable. Self-esteem has independent effects on all the work— orientation variables, while S.I. in work seems to have 184 independent effects only on universalism. Discussion This section concerns the theoretical and, to a lesser extent, the practical implications of the findings of this study. From the findings it seems to us that not all achievement variables are of equal import in individual manager's orientation toward work, self—esteem and S.I. in work. From these data we find little support for Kanter's (1977) and Rosenbaum's (1976) view that promotion chances may be a fundamental determinant of a wide variety of other attitudes and behaviors, and are perhaps more important than an individual's personal ‘ Egaits or level in the organization (Rosenbaum 1984, p. From these data, promotions have not been more important than rank (or level in the hierarchy) in the determination of individual social psychological functioning. This finding is incidentally in line with that of South et. a1. (1985) who, when they tested some of Kanter's (1977) ideas on the effect of opportunity, concluded that promotional effects either are infrequent in occurrence or occur in special circumstances. While we are dubious about the claim that promotions are determinant of a wide variety of attitudes and behaviors, we should be cautious about rejecting it out of hand. This is because our data are limited by the fact that we have only considered the promotion variable unidimensionally, i.e., as a purely 185 quantitative variable. We have been less concerned with the qualitative side which may be crucial to understanding the meaning individuals impute to promotions. We put our focus on individual analysis at the expense of job analysis--which brings in the issue of short and long Job ladders which are important in understanding the context in which promotions occur. In any case, for these data, rank or level in the organization proved to be more consequential in determining aspects of social psychological functioning in industry. This may have something to do with the nation in which the study was done. Zambia has been politically independent for only 20 years. The colonial legacy is, arguably, still strong. Colonial stratification was on the basis of race, education and organizational hierarchy. In the words of Daniel (1979, p. 99), "the gap between the leader and the led in industry was an enormous one of race,class, skill and income." The colonial emphasis on hierarchy or rank may still be operative in the post-colonial era, and, at least in part, may be accounting for some of the observed effects of rank on social psychological functioning. We also feel that the accommodative effects of rank have been amply demonstrated in the findings. Faunce (1984) has disputed the assumption that there is a universal and inherent need for self-esteem in people. The "self investment process is a selective process through which activities become differentially imbued with significance 186 for the maintenance of self-esteem" (Faunce 1984). Faunce proposes a search for conditions which generate self-esteem. The need for self-esteem is created in recurring social situations. We believe that rank may present one such social situation in whicthhe need for self-esteem may be created. More specifically at the upper level of the organization, the pressures faced by management may create in them a greater need (and a greater capacity to satisfy the need) for self-esteem. The need for self-esteem does seem to be somewhat different at upper, middle and lower management levels. Relationships which hold at one level do not necessarily do so at other levels. This study highlights a finding emphasized by some in the sociology of work and industry--one which posits a unique situation for middle management in industry. Derossi (1981) for one says, The state of affairs for those at the upper and lower managerial strata is fairly clearly defined: the former are partially inserted into the power structure and the latter, for the most part are cut off. But managers of the intermediate level find themselves in an ambiguous situation: while they share with the upper level a similar autonomy in technical decisions, they are closer to the lower levels in their lack of participation in policy-making (Derossi, 1981, p. 50). Child and Partridge (1982), Crozier (1971), and Katz and Kahn (1978) subscribe to a similar position. At the middle level, individual managers are in a situation where they are pulled from all sides (iJL, those at the lower and those at the upper end of the organizational hierarchy). Thus they 187 in part exist in the firm to balance contradictory demands from diverse parties. They may be, as Derossi says, in a situation of ambiguity. The pressure facing them may be among the reasons why for middle management, job security does not necessarily lead to such work orientations as universalism. Their ambiguous situation may not be conducive to the application of objectivity, a precondition for a universalist work orientation. We also found middle level managers to be unique in that the longer they stayed in their present jobs the more participative were their management styles. Balancing diverse interests may require much experience. Middle level managers who have stayed longer in their jobs are those with the experience to handle or balance diverse interests. 4A manager may balance diverse interests better if he/she involves all the relevant parties, and the better one is at balancing diverse interests (which may require experience), the more participative their orientation may be. Another unique finding with regard to middle management was that job security did not relate to S.I. in work. A number of theorists have commented on alienation at management levels. Granick:(1976) claims that career development mainly consist in reward packages, which firms offer to employees. Managerial behavior to a large extent is a response to increases in personal income, authority, expansion of possibilities for exercising creativity in decision making, 188 promotion, etc. People at middle management whose careers have been arrested are robbed of incentive and freed from concerns of achieving the corporationfls goals. Granick's concern is with career development because he believes it is the key to explaining differences in managerial motivations and behaviors. In his study of French managers,Granick (1976) found that access to these reward packages is determined by the type of school a manager attended. In France most people at lower and middle management levels know of their limited chances for advancement in the firm. On the other hand, the firm does make them aware that they will not be easily dismissed. Granick says that this is the reason why lower and middle level managers in France are lackluster in their performance. In other words the fact that these managers are aware of the limitations to their career development alienates them from the company's values. They are in work for what they can obtain, Job and financial security. “Alienation refers to a disjuncture between self- esteem maintenance and status assignment processes" (Faunce, 1981, p. 138). Such managers are alienated from the corporation because the instrumentalist values (of how much they can get from the firm at minimum effort) they hold are different from the criteria firms use in evaluating employee behavior and/or performance at work. Alienation may thus be seen as one response to failure in developing one's career. Jennings (1971) makes a similar statement when he talks 189 about the process of shelf-sitting (failed career development). Jennings says that corporations have been reluctant to fire shelf-sitters as some kind of reward for past contributions to the corporation. These "burn proof" individuals develop an.Lnstrumental attitude toward work. That is, they , more and more, withdraw their psychological energies, while hoping to increase their material benefits. In this situation, job security aids alienation. These data by demonstrating that job security leads to SII. in work for upper and lower level managers but not for middle level managers have corroborated Granick and Jenning's claims. Further research may be needed to systematically investigate the relationship between.arrested.career development, job security, and alienation among not only middle management but also upper and lower level managers. Another point which has emerged in this study is the importance of separating the effects of promotion from those of rank or job level in the organization. Many studies do treat the two variables as functional equivalents. The specific reference is to the studies, we reviewed earlier, linking work achievements to S.I. in work and to self- esteem. In the study by Clelland et al. (1976) the nurse's present job level was taken to be indicative of her past promotions. Thus the head nurse was taken to have had more promotions than a general duty nurse, and an administrator as having had more promotions than both the head nurse and 190 the general duty nurse. Similar is the case with Newman's (1975) study which in.part deals with an individual's job, location, a concept subsuming hierarchical level, important in his investigation of the relationship between mobility and job involvement. These studies seem to have missed the point. Promotions and Job levels may be highly correlated but they are not one and the same thing. This study by showing that the effects of promotion differ from the effects of rank has in part demonstrated that the two concepts are different. This is only true assuming that promotions can be adequately measured quantitatively only. In the case the qualitative side of promotion is seen as important we are less willing to stand by this conclusion. In these data we found that power positively influences S.I. in work and the work orientation of participativeness but not universalism and flexibility. This essentially confirms Kanter's (1977) hypothesis that power increases a supervisors security, credibility and effectiveness in the organization. This is so provided we assume S.I. in work, and, participativeness, do increase a manager's effectiveness. There is much research to support this assumption. Power insofar as it presupposes credibility is an interactional concept. lflzresults from evaluations by significant others in the cmganization. Having legitimate power is in a way a demonstration of positive evaluation of a given individual's success at getting the job done. It 191 positively affects S.I. in work because both largely share the same functional base-evaluations from relevant others. Having said this, it must be noted that the power system is not an immutable process. New power relations are created over time and these affect the organization's climate. This brings us to the issue of the distinction between the relative distribution and the absolute distribution of power in organizations. Tannenbaum and his associates have a point in emphasizing Parson%3(1960)'view that power should not be considered as a zero-sum entity in which some people have as much power as others do not. Power is an expandable and contractable resource in organizations. Researchers should not unduly emphasize relative power at the expense of total distribution of power. In this study we have attempted to capture the concept of relative distribution of power by distinguishing between powerful and powerless managers according to how they scored on a power scale. As expected, power ‘was associated with some of the work orientation variables. When we took into account total amount of power or control in the organization, the correlations between powerfulness, S.I. in work, and work orientations were modified. Powerfulness affected managerial S.I. in work and participativeness in firms having a high total amount of control but not in firms with medium and/or low total amount of control. Powerfulness did not relate to flexibility and 192 universalism in all the firms. It seems, then, that total amount of control does influence some achievement effects in industry. We are, however, dubious about Tannenbaum's claims that total amount of control always serves as a moderating influence on relationships between achievements and work orientations in the organization. This is because of our finding, with regard to these data, that the total amount of power did not enhance the effects of individual manager's power on universalist and flexible work orientations. That is, there were no differences in the universalist and flexible orientations of powerful and powerless managers in firms with low, medium and high total amount of control. It seems there are conditions under which the moderating effects of absolute amount of control are asserted and that the effects of total control are not exactly unambiguous. They seem to occur in certain circumstances and we believe there is need for further research to specify these circumstances. There are practical and theoretical implications to all this. We can expect that as some firms increase their total amount of control, some work orientations of their members will change. The powerless may be more effective in an environment in which they can influence discussion and decision, and this environment is provided by firms with high total amount of control. APPENDIX A Letters of introduction to the firms in which interviews were carried out. MICHIGAN STATE UNIVERSITY EAST LANSING . MICHIGAN . 48824-1035 AFRICAN STUDIES CENTER 0 100 Center for International Programs January 15, 1985 To Whom It May Concern: Concerning: Mr. Penpelani Mufune This letter is to introduce and commend to you Mr. Pempelani Mufune, a post-graduate research student at Michigan State University in East Lansing, Michigan. Mr. Mufune holds the B.A. degree from the University of Zambia and an M.A. degree in Sociology from Michigan State University, where he has completed his post-graduate studies with excellence. He now has completed all course and seminar work and other requirements for the Ph.D. degree (Doctor of Philosophy in Sociology), and is conducting research in Zambia toward the completion of his dissertation for that Ph.D. degree. His research concerns styles of management and careers in large organizations in Zambia. His research is for academic purposes, and he is pledged to the highest of ethical standards to protect the iden- tity and ensure anonymity of all persons interviewed by him. Please provide him with all possible assistance and cooperation in completing this project and his research. It will be a major contribu- tion to the development of a new faculty member for the University of Zambia and to the completion of his doctoral degree. If you wish further reference on this excellent student, please contact me directly. Sincerely yours, David Wiley Associate Professor of ciology Director, African Studie Center Michigan State Universit 193 MSU a an Alflnnduc Action/Equal Opportunity Imam Telephone: (517) 353-1700 Cable MSUINTPRO ELSG Telex 11: 810-251-0737 THE ~ UNIVERSITY OF ZAMBIA SCHOOL OF HUMANITIES ADDSOCIAL SCIENCES P.O. Box 32379 LUSAKA . ZAMBIA Telephone: 213221 Ext. 314, 316 Telegrams: UNZA LUSAKA Telex: 44370 V Our Reference: 5th "BrCh' 1985 TO WHOM IT NA? CONCERN This letter is to introduce you to the bearer, Mr. P. Mufune. He is a staff Development Fellow under the auspices Of the University Of Zambia. At this time he is carrying out a research project in the area of Industrial Sociology. Any help that you can give him will be deeply appreciated. Thank you. .Yours sincerely, ~ Elisabeth E. Brooks. PhD Head DEPT, OF SOCIAL DEVELOPMENT STUDIES. DEB/cum 194 ZAMBIA INDUSTRIAL AND MINING CORPORATION LIMITED 6 March 1985 TO WHOM IT MAY CCNCERN ZIIHCO I‘qusc Cairo Road P.O. Box 30090, Lusaka Republic of Zambia Telephone 212487I215008I212802 Telegrams “Zimco” Telex ZA 40790 This letter is to introduce to you Mr. Pempelani Mufune, a post graduate research student at Michigan State University who is currently attached to the University _ of Zambia as a Staff Development Fellow. Mr Mufune, who holds a Michigan State University M.A. degree in sociology and the University of Zambia B.A. degree is conducting research in Zanbia towards the completion of his dissertation for a Doctor of Philosophy degree in Sociology. As part of his dissertation will include research in Industrial Sociology which will entail him calling on many of the industrial concerns within the Zimco group of companies, any help that you may accord him will be deeply appreciated by Mr Muftme, the University of Zambia and indeed by Zimco . Sincerely yours 0 J ya P IC REIATICNS OFFICER v 195 GM Mulenga 5M Chuunga CL.30 15 March 1985 RE: ? MUFUNE Mr Mufune is a staff Development Fellow under the auspices of the University Of Zambia and is currently carrying out a research project in the area of Industrial Sociology. The purpose of this project is to enable him to gather information necessary for the completion of his dissertation for the PhD degree. 1 should be grateful if you would avail him the oppOrtunity to interview and hold discussions with some of your staff who are in supervisory positions and any others from whom he thinks he can get useful information in réspect Of his project. e‘eeeeeeeeoeeeeeeoeee Same Memo sent to: CH Kaunda_- Interchem GM Simpungwe - Meridian Motors ER Syder - Meridian Bank , , C 196 Lonrho CAIRO ROAD P.O.Box 32568 LUSAKA ZAMBIA zaMUla . I TELEPHONE 217271 TELEX ZA 42230 CABLES LONRHO Ilmlted YOUR REF: INfiggrfh/IIRBAITED OUR REF: A 9th April 1985 Mr D H Atkinson General Manager Star Commercial Limited LUSAKA Dear Mr Atkinson This letter serves to introduce you to the bearer Mr P Mufune who is a staff development .fellow under thetauspices .ofthe University‘of Zambia. Mrimfme is carrying out aresearch'project in Industrial Sociology. He is doing this research in order to help himwith the programme of studyfor .PhD degree. We should be grateful if you would assist him as much . as possible. Yours sincerely I R LUHANGA "SECRETARY _ .197 APPENDIX B a) List of firms in the Zimco directory from which parastatal firms were selected. b) A typical example of organizational chart in a parastatal firm a) List of firms in the Zinco directory from which parastatal firms were selected. Zambia Industrial and Mining Corporation Limited (ZIMCO) (Parent Company Mineral Exploration Department (MINEX) Zimco Institute of Management (ZIM) Agriculture The Agricultural Finance Company Limited Kawambwa Tea Company Limited Lakes Fisheries of Zambia Limited Zambia Agricultural Development Limited Zambia Forestry'and.Forest Industries Corporation Limites (ZAFFICO) Zambia PorK Products Limited Zambia Seed Company Limited Communications Posts and Telecommunications Corporation (PTC) Energy Lublen Limited Tazama Pipelines Limited Zambia Electricity Supply Corporation Limited (ZESCO) Zambia National Energy Limited (ZNEL) Agip (Zambia) Limited 198 199 BP Zambia Limited Finance Zambia National Commercial Bank Limited (ZNCB) Zambia National Insurance Brokers Limited (ZNIB) The Zambia State Insurance Corporation Limited (ZSIC) Auto Care Limited Avondale Housing Project Limited City Radio & Refrigeration Supplies (1975) Limited Zambia International Insurance Services Limited Hotels National Hotels Development Corporation Limited (NHDC) Eagle Travel Limited (ETL) Zambia Hotel Properties Limited Industry Industrial Development Corporation Limited (INDECO) AFE Limited Chilanga Cement Limited Consolidate Tyre Services Limited (CTS) Choma Milling Company Limited Crushed Stone Sales Limited General Pharmaceuticals Limited Indeco Estate Development Company Limited Kabwe Industrial Fabrics Limited Kafironda Limited 200 Kafue Textiles of Zambia Limited (KTZ) Kapiri Glass Products Limited (KGP) Livingstone Motor Assemblers Limited (LMA) Luangwa Industries Limited Lusaka Engineering Company Limited (LENCO) Mansa Batteries Limited Metal Fabricators of Zambia Limited (ZAMEFA) Zamefa Sales Limited Monarch Zambia Limited National Breweries Limited National Milling Company Limited Nitrogen Chemicals of Zambia Limited (NCZ) Norgroup Plastics Limited ROP (1975) Limited Rucom Industries Limited Supa Baking Company Limited United Milling Company Limited Zambia Breweries Limited Zambia Ceramics Limited Zambia Coffee Company Limited Zambia Oxygen Limited Zambia Steel & Building Supplies Limited (ZSBS) Zambia Sugar Company Limited (ZSC) Nakambala Estate Limited Zambia Clay Industries Limited Duncan, Gilbey & Matheson Limited 201 Dunlop Zambia Limited Nkwazi Manufacturing Company Limited Scaw Limited Mining Maamba Collieries Limited Metal Marketing Corporation Limited (MEMACO) Memaco Services Limited Reserved Minerals Corporation Limited (RMC) Mindeco Lumwana Limited Mindeco Noranda Limited Mokambo Development Company Limited Mindeco Small Mines Limited Zambia Consolidated Copper Mines Limited (ZCCM) ZCCM Subsidiary Companies Circuit Construction Limited Circuit Engineering & Tooling Limited Circuit Holdings Limited Circuit Safaris Limited Circuit Saw Milling & Joinery Copper Industry Services Bureau Limited Copperbelt Power Company Limited Mpelembe Drilling Company Limited Mpelembe Properties Limited Nchanga Farms Limited 202 Ndola Lime Company Limited Redirection Placement Limited RST Management Services Limited Technical Management Services of Zambia Limited ZAL Holdings Limited Zambia Appointments Limited Zambia Engineering Services Limited Zamcargo Limited Construction & Real Estate Zambia Engineering & Contracting Company Limited (ZECCO) Zimco Properties Limited Zimco Services Limited Trading National Import and Export Corporation Limited (NIEC) Consumer Buying Corporation of Zambia Limited (ZCBC) Mwaiseni Stores Limited National Drug Company Limited NIEC Agencies Limited NIEC Overseas Services Zambia Limited NIEC Stores Limited Zambia Horticultural Products Limited Zambia National WholesaLe & Marketing Company Limited (ZNWMC) 203 Transport Contract Haulage Limited (CH) United Bus Company of Zambia Limited (UBZ) Zambia Airways Corporation National Air Charters Zambia Limited Zambia Railways Zambia Concrete Limited Zambia National Clearing and Forwarding Company Limited (ZNCFC) Zambia Tanzania Road Services Limited (ZTRS) _ W 023‘ avian-Sr g g as: g m 5523 annual 552: m .322. 933 3 8.6.3! 3.3.9. _ 3856.. 30H”; B. 3..— pence: .In!- 833 28.. .83.. enemas. "mg.“ .HWNEB , gun 25:9. 32...... 9...! as... .35. nuflwnmfiv: “Nannie: era was we» mafia H2839 ooaoidozl a 2 camp my APPENDIX C Interview Schedule 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 205 MANAGERIAL JOBS INTERVIEW SCHEDULE During what years have you been in this firm? What is your present job title? What are your most important tasks/duties in this job? During what years have you been in your present position 2) Production 3) Personnel/training 6) Other, specify In what section do you work? 1) Sales 4) Administration 5) Research and development Have you ever changed jobs in this firm? 1) Yes 2) No If no, go to question 20. If yes, ; . what was your previous job? In which section was your previous job? Was your previous job a promotion? 1) Yes 2) No 3) Demotion In what years were you in your previous job? What was your job before your previous job? Was the job before your previous job a promotion? 1) Yes 2) No '3) Demotion In what years were you in the job before your previous job? In what section was the job before your previous job? What other jobs have you had in this firm? In which section were these jobs? Did these jobs constitute 1) Promotion 2) Moving down from previous job 3) About same as previous job During what years were you in these other jobs? What were your most important dutiesin these jobs? How do you see your chances for promotion in comparison to: (1) (2) (3) (4) (5) Very Very 9.029 929. £233 Bad £511. Managers in other firms _____ Managers in this firm ____ ____ Most of your friends ____ Most OF your relatives _____. .___ .___ .____ ____ 24. 25. 26. 27. 28. 29. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. What do you have to do to get promoted in this firm? 206 In your experience how hard has this firm? Would you say it is 3) Somewhat hard to do 4) Easy Here are a number of statements whether you 1) Strongly agree; 5) Strongly disagree There is much job security in my position. it been for yOu to gEt promoted in 1) Very hard to do concerning job security. 2) Hard to do Indicate 2) Agree; 3) Undecided; 4) Disagree; (1) There should be more security (1) in my job. Job security is very important (1) 130 me. If performance went down I may (1) be demoted. If performance went down I may (1) be transferred. Have you ever worked in a job outside this firm? If no, proceed to question 44. last job outside this firm? (2) (3) <2) <3) <2) <3) <2) . <3) <2) <3) 1) Yes (4) (4) (4) (4) (4) 2) No (5) (5) (5) (5) (5) If yes to question 31, what was your In which section was this job? In which company was this job? In what years were you in this job? Was this job a promotion? 1) Yes 2) No 3) Demotion What were your most important activities in this job? What other jobs have you had in firms outside this company? What were your most important duties in these jobs? In what section were these jobs? In which company were these jobs? In what years did you work in these jobs? Which of these jobs were promotions? Which of these jobs were about the same as the previous one? Which of these jobs constituted a step or so down from previous one? 46. 47. 48. 49. 50. 51. 52. 53. 54 55. 56. 57. 58. 59. 207 In which level does your position fall in this firm? 1) Level 1: Person is in charge of the whole unit of the organization e.g. CEO: G.M. 2) Level 2: Is in charge of a whole segment in the organization, e.g. upper level manager in charge of sales in the unit of the organization 3) Level 3: Is middle level manager in charge of a small segment of the organization and directly falls below upper level manager. 4) Level 4: Is a junior manager who may sometimes help supervise workers. 5) Level 5: Consist of foreman/technical staff who supervise workers and in many cases may be recruited from worker's rank. Are you older, younger or about the same age as most people in positions at your level in the firm 1) Older 2) Younger c) About same age Firms seem to differ as to the amount of influence different persons exercise on decision making. In your opinion how much influence do the following persons actuallyhave on what happens in this firm? (I) (2) (3) (4) (5} Very Quite A Great Little Little Some Alot of Deal of Control Control Control Control Control ———~—_——-“—_ Chief Executive Officer —————.——————.—-——-———-—————-—— Upper Level Managers ——-———— -—.————— -_——-——- —————— ._———_ Middle Level Managers Lower Level Managers mmflmm Supervisors/Foremen wwwmm Employees/Workers How much influence should the following-persons actually have on what happens in this unit of the organization (1) (2) (3) (4) (5) Very Quite A Great Little Little Some Alot of Deal of Control Control Control Control Control .Chief Executive Officer -Upper Level Managers -Middle Level Managers .Lower Level Managers Supervisors/Foremen Employees/Workers 208 Here are a number of statements about your influence on the job. Indicate whether you 1) strongly agree; 2) agree; 3) Undecided; 4) disagree; 5) strongly disagree. 60. My job lets me use my skills and (5) (4) (3) (2) (l) abilities. 61- . I have much influence on the deci- (5) (4) (3) (2) (l) sions made by my supervisor 62. An individual who wants to make his (5) (4) (3) (2) (I) own decisions can be quickly dis- couraged here. 63- It is very important for me to move (5) (4) (3) (2) (1) up (get promoted). 64. My job involves unpredictable (5) (4) (3) (2) (1) situations. 65. ‘ I feel uncertain about being (5) (4) (3) (2) (l) promoted. 66. '" Others make conflicting demands (5) (4) (3) (2) (1) of me on the job. Eknxaare a number of statements concerning work in general. Indicate whether you 1) strongly disagree; 2) disagree; 3) undecided; 4) agree; 5) strongly agree 675 The major satisfactions in my life (1) (2) (3) (4) (5) come from work. 68. I do what my job description requires, (1) (2) (3) (4) (5) no firm has a right to expect more. 69. The most important thing to teach workers (1) (2) (3) (4) (5) is obedience to managers/employers. 70- ‘Workers who question old and accepted ways (1) (2) (3) (4) (5) of doing things may cause trouble. 71. In any effective firm, an employee must (1) (2) (3) (4) (5) sometimes disregard the wishes of his superior if he is to do his job properly. 72- People are better off when the organ- (1) (2) (3) (4) (5) ization does not provide a complete set of rules. 73. I feel I do not have much to be proud of. (1) (2) (3) (4) (5) 74. I take a positive attitude towards myself. (I) (2) (3) (4) (5) 75. Individuals who comply with managers should (1) (2) (3) (4) (5) be allowed to bend rules a little. 76. Most things in my life are more important (1) (2) (3) (4) (5) than work. 77. 78. 79. '80. 81. 82. 83. 84. 85. ‘86. 82. 88. '89. 90. 91. 209 To me, my work is only a small part of (1) who I am. The world of work is complicated. The (1) only way to know what to do is to rely on managers/experts. Managers should not encourage subordinates to (I) find new ways of doing things. Even if one does not like a rule they (1) should obey it. Often to be effective an employee must (I) be flexible and not rigidly follow rules. I feel that I have a number of good (1) qualities. All in all, I am inclined to feel that (l) I am a failure. As a manager it is impossible not to tend (1) to favor one or a few subordinates. I feel I am a failure as a person when I (I) fail at something to do with work. A manager should bargain with subordinate (l) to obtain results. A manager should be strict with sub- (1) ordinates in order to gain their respect. I feel I am a successful person to the (l) extent I do my work well. I certainly feel useless at times. (1) I feel I do have much to be proud of. (1) Well behaved employees should not be (1) punished as much as those who don't behave well when they do something wrong. (2) (2) (2) (2) (2) (2) (2) (2) (2) (2) (2) (2) (2) (2) (2) (3) (3) (3) (3) (3) (3) (3) (3) (3) (3) (3) (3) (3) (3) (3) (4) (4) (4) (4) (4) (4) <4) (4) (4) (4) (4) (4) (4) (4) (4) (5) (5) (5) (5) (5) (5) (5) (5) (5) (5) (5) (5) (5) (5) (5) 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 105. 106. 107. 108. 109. 210 Name of firm you work in Gender (sex) 1) Female 2) Male Highest education completed: 1) Grade 7/8 certificate 2) Form 2/3 certificate 3) Form 5/6 certificate 4) BA/BS degree 5) Masters degree 6) Other, specify On the job training (indicate number of years): Marital status: 1) Married 2) Single 3) Widow(er) 4) Other, specify Number of children Age Nationality: 1) Zambian 2) Non Zambian (specify country) Mother's education (i.e. number of school years completed) Father's education (i.e. number of school years completed) Father's occupation when you were 18 years (was this a supervisory position?) Was your mother employed when you were 18 years (if yes, what was her occupation?) Personal ' Income bracket: 1) K6,000 and below 2) K8,000 and below 3) K11,000 and below 4) K15,000 and below 5) Above K15.000 Religious denomination In Zambia what is this firm? 1) Principal unit 2) Subsidiary (with legal identity) 3) Head branch (with HQ on same location 4) Branch How long has this unit of the organization been in operation? How many people are employed in this unit of the organization? What are the main products and services of this firm in Zambia? 211 110. In the last ten years how many people have been promoted in this unit of the organization? 111. In the same period how many have been demoted? Thank you very much. We believe we have all the information we need, but just in case we missed something, can we have your name, and telephone number, please? APPENDIX D ADDITIONAL STATISTICS 212 Table D:1 Distributim of respa'zses (11 five key scales. 1 Sizrngly 2 Disagree 3 Don't 4 Agree 5 Strongly Total Concept Disagree Know Agree N S . I . in work 0 4 3O 62 5 101 Self-esteem 2 28 39 28 4 101 Participa- tiveness 1 32 50 18 O 101 Universalism O 31 49 19 2 101 Flexibility 3 15 64 15 4 101 Powerfulness 5 18 46 3O 2 101 213 Table D:2 Multiple regression models of achievement and work orientation variables. Model 1: Regression of work achievement variables on measures of work orientation, S.I. in work and Self-esteem N Pmrotim Power Rank R2 SELF-ESTEEM 101 -.91(-.25)** .46(.20)**1.7(.59)** .44 S.I. INIADRK 101 -.27(-.09) .39(.23)* .41(.18) .10 FLEXIBILITY 101 -.24(-.14) -.85(-.O3) .70(.49)** .24 UNIVERSALISM 101 -.42(-.19)* .12(.09) .7l(.40)** .19 PARTICIPATION 101 -.62(-.23)** .23(.l3)* 1.6 (.73)** .59 Standardized Beta coefficients are in parentheses. Model 2: Regression of Achievement Variables. S.I. in Work, and Self—Esteem on Measures of Work Orientation Variables Prunortion Power Rank S.I. in Self Work Estean R2 FLEXIBILITY -.05(-.O3) -.O9(-.O9) .34(.24)* .009(.Ol) .19( .42)** .35 UNIVERSALISM -.20(-.02) -.20(-.09) .30(.l7) .18(.23)* .18(.31)** .35 PARTICIPATION -.35(-.l3)* .09(.05) 1.09(.49)** -.02(-.02) .30(.41)** .58 *P<.OS ** P > .05 Model 3: Regression of Achievement Variables. of Work Orientation Variables. S. Esteem. 214 Gender. Education. and Age on Measures and S.I. in Work. Promotion Power Rank Gender Education Age R2 FLEXIBILITY -.12(-.O7) -.Ol(-.01) .40(.28)* .83(.23)* -.01(-.OO7) .Ol(.lO) .29 UNIVERSALISM -.17(-.06) .05(.05) -.OO3(-.001) 1.5(.51)** .19(.09) .03(.13) .32 PARTICIPATION -.32(-.11)* .17(.10) .9(.42)** 2.1(.53)** -.25(-.O9) .03(.10) .72 S.I. IN WORK -.O4(-.01) .39(.22)* -.10(-.O4) 1.8(.43)** -.27(-.09) .05(.17) .20 SELF-ESTEEM -.38(-.10) .33(.15)* .51(.17) 3.5(.65)** -.19(-.05) .03(.08) .62 Standardized regressions are in parenthesis * P < .05 ** P < .01 Table D:3: Oorrelatim of Pranotion and S.I. in Work According to Group of Age at Given Rank Median Pearscn Correlation of S.I. in Work and Prurotim Lower (N = 32) -.O88 Higher (N = 69) 0.229 * P< .05 ** P > .05 215 BIBLIOGRAPHY Achens, C. H.: 1982 Interpreting and Usigg Regression, Sage, Beverly Hills, CA. Anderson, A., C. Barsilevsky, and P.Hum.: 1983. "Measurement Theory and Techniques," in Anderson, A; Rossi, P. and Wright, J.: A Handbook 9: Sggygy Research, Academic Press, N. York. Anderson, A, P. Rossi and J. Wright: 1983. A Handbook 9: Survey Research, Academic Press, N. York. Aronoff, J. and L. Meese.: 1971. "Motivational Determinants of Small Group Structure," Journal 9; Personality 29;; Sggigl Psychology, 17:319-29. Aronoff, J.: 1970. "Psychological Need - A Determinant in the Formation of Economic Structures," Human Relations, 20:123-128. Asher, H. B.: 1983. Causal Modeling Sage, Beverly Hills, CA. Bachman J. and P. O'Mally: 1977. "Self-Esteem in Young men: A Longitudinal Analysis of the Impact of Educational and Occupational Attainment," ggggggl 9; Personality and Social Psychology, 35:365-80. Bernard, C.: 1938. The Functions 9f the Executive. NLI.T. Press, Cambridge, Mass. Bartoke, K, W. Escheweiler, D. Flechsenger, and A. Tannenbaum: 19EL."Workers Participation and Distribution of Control in Ten German Companies," Administrative Science Quarterly 27. pp. 380-97. 216 217 Bhagavan, M: 1978. Zambia: Impac tof Industrial Strategy on Regional Imbalance and Social Ineguality, Scandinavian Institute of African Studies. Boshier, R.: 1969. "A Study of the Relationship Between Self Concept and Conservatism," £93592; 9: Social Psychology, 77:139-140. Bradburn, N.: 1983. "Response Effects" in Rossi, et al. (eds.): 5 Handbook 9: Survey Research, Academic Press, N. York. Bray, D., R. Campbell, and D. Grant: 1974. Formative Years ig Business: a Long- -Term 5;.3; and T;_Study g: Managerial Lives, Wiley, N'. York. Burawoy M.: 1971. The Colour of Class of the Zambian Copperbelt, Manchester University_ Press, Manchester. Child J. and J. Patridge: 1981. Lost Managers; Supervisors in Industry and Society, Cambridge University Press, Cambridge. Clelland, V.; A. Bass; W. McHugh and J. Montano: 1976. "Social and Psychological Influences on Employment of Married Nurses,"iNursing Research, 25:90- 7. Coopersmith, S.: 1967. Antecedents g; Self-esteem, W.IL Freeman, San Francisco. Crozier, M.: 1971. The World 9: the Office Worker, Schocken books, N. York. Daniel Philip: 1979. Africanization, Nationalization and Inequality, Cambridge University Press, Cambridge. Derossi, F.: 1981. "Decision Making and Policymaking in Italian Industrial Firms," International Journal 9: Sociology, Vol. 11 No. 3 Fall. Davis, J. W.: 1966. "Work Involvements of Executives," Personnel Administration, 29:6-12. 218 Dubin, R.: 1956. "Industrial Workers Worlds: A Study of the 'Central Life Interests' of Industrial Workers," Social Problems, Vol. 3, pp. 131-142. Dubin, R. and J. Champoux: 1974. "Worker's Central Life Interests and Job Performance," Sociology 9: Work and Occupation, Vol. 1, No.3. Dunne, E.; M. Stahl and J. Melhart: 1978. "Influence Sources of Project and Functional Managers in Matrixorganizations,” Academy g Management Journal, 21:135-40. Elliot, J.: 1975. "Education in Zambia," South African Journal 9;: African Affairs, 5:3-9. Farris, G. and D. Butterfield: 1972. "Control Theory in Brazillian Organizations," Administrative Science Quarterly, Vol. 17, No. 4, Dec” pp. 574—85. Faunce, W.: 1984. "School Achievement, Social Status, and Self- esteem," Social ngchology Quarterly, Vol. 7, No. 1, 3- 14. Ghisselli, E. and D. Johnson: 1970. "Need Satisfaction, Managerial Success and Organizational Structure," Personnel Psychology, 23:569-78. Granick, D.: 1976. ”Why Managers Perform Differently in Different Countries: France, Britain, U.S.A., U.S.S.R. and East Germany," in Boddewyn, J. (eds) European Industrial Managers _-_ West and East. International Arts and Sciences Press Inc., White Plains, N. York Halaby, C. and M. Sobel: 1979. "Mobility Effects in the Work Place," figggiggg £93553; 9: §ggiglggy, Vol. 85. No. 2:385-416. Hall, D. T. and F. Hall: 1976. "Career Development: How Organizations Put Their Finger on People," in Dyer, L. (ed.) Qgggggg £3 Qgganizations, Cornell University,rL York School of Industrial Relations, pp. 1-15. N. York. 219 Hamilton, G. and N. Biggart: 1985. ”Why People Obey: Theoretical Observations on Power and Obedience in Complex Organizationsfl' Sociological Perspectives, Vol.28 No. 1 January,3-28. Hope K.: 1975. "Model of Status Inconsistency and Social Mobility Effect," American Sociological Reviem, 40(June):322-43. Jackson, E. and R. Curtis: 1972. "The Effects of Vertical Mobility and Status Inconsistency: A Body of Negative Evidence," American Sociological Review, V01. 37(Dec), 701-13. Jacques, J. and K. J. T. Chason: 1977} "Self-Esteem and Low Status Groups: A Changing Scene," The Sociological Quartergy, 18, pp. 399-412. Jennings, E.: 1971. The Mobile Manager, McGraw Hill, N. York. Jones, A., L. James, J. Bruni: 1975. "Perceived leadership Behaviour and Employee Confidence in Leader as Moderated by Job Involvement," Journal 9: Applied Psychology, 60:146-9. Kanter, R. M.: 1977. Men and Women 9: the Corporation, Basic Books, N. York. Kanter R.M.: 1979. "Differential Access to Oppurtunity and Power," in Alvarez and Lutterman (ed.) Discrimination £2 Organizations, Jossey-Bass San Francisco, CA. Kanungo, R.: 1982. Work Alienation Praeger Publishers, bLYork. Katz, D. and R. Kahn: 1978. The Social Psychology 9: gmganizations John Wiley,rm York. Kessin, K.: 1971. ”Social and Psychological Consequences of Intergenerational Occupational Mobility," American £99332; 9: §ggiglggy, Vol. 77 No. 1 July pp. 1-18. Kim, J. and C. Mueller: 1978. Factor Analysis, Sage Publications Beverly Hills, CA. 220 Kipnis, D. and W. Lane: 1962. ”Self Confidence and Leadership," QQEEEEl 9f Applied Psychology, 46, pp. 291-5. Kohn, M. and C. Schooler: 1973. "Occupational Experience and Psychological Functioning: An Assessment of Reciprocal Effectsfl' American Sociolggical Review, 38(February) pp. 97-118. 1983. Work and Personality: Am Inquiry Into the Impact of Social Stratification, Ablex Publishing Corporation, N. Jersey. Lodahl, T. and M. Kejner: 1965 "The Definition and Measurement of Job Involvement," 292E221 9: Applied Psychology, 49, pp. 24-33. Lorence, J. and J. Mortimer: 1981. "Work Experience and Work Involvement," Sociology 3f Work and Occupations, Vol.8 No.3 (Aug) pp. 297-326. Luhman, N.: 1973 "The Attribution of Promotions Among Civil Servants," Zeitschrift fur Soziologmg, 2(4) Oct :326- 351. Mannheim, B.: 1975. "A Comparative Study of WOrk Centrality, Job Rewards and Satisfaction," §pgi_ology 9: Work and Occupations, 2, pp. 79-102. Martin, N. and A. Strauss: 1959. "Patterns of Mobility Within Industrial Organizations," in Warner and Martin (eds.) Industrial mgm, Harper and Brothers, N. York. Maurer, J.: 1969. Work Role Involvement 9; Industrial Supervisors, MSU Business Studies, East Lansing, Michigan. McIver, J. and E. Carmines: 1981. Unidimensional Scalling, sage publications, Beverly Hills, CA. Mendenhall, W.: 1971. Introduction :9 Statistics 229 Probability Duxburg Press, Belmont, Mass. 221 Meyns, P.: 1984. "The Political Economy of Zambia," in Woldring and Chibaye: Beyond political independence. Mouton Publishers, Berlin pp. 7-22. Mhone, G. C.: 1982. The Political Economy 9; _a_ Dual Labor Market i9. Africa the Copper Industry and Dependency iE.§§EE£§L 1929-1969, Farrleigh Dickinson University Press, Rutherford, NY. Mintzberg, H: 1973. The Nature 9: Managerial Work, Harper and Row, New Ybrk. Mitchell, V., V. Baba, and T. Epps: 1975. "On the Relatioship Between Job Involvement and Central life interest," Relations industrielles 39: 166-180. Mortimer, Jeylan: 1979. Changing Attitudes Toward Work, Work in America Institute Scarsdale, N. York. Nie, N., C. Hull, J. Jenkins, K. Steinbrenner and D. Bent: 1975. S.P.S.S., McGraw Hill Book Company, N. York. Newman, J.: 1975. "Understanding the Organizational Structure - Job Attitude Relationship Through Perceptions of the Work Environment," Organizational Behavior and Ammgm Performance, 14:371-97. Parsons, T.: 1960. Structure and Process $2 Modern Society, Free Press, Glencoe, New York. Pedhazur, E. J.: 1982. Multiple Regression £2 Behavioral Research, Holt, Rinehart and Winston, N. York. Pelz, D.: 1951. "Leadership within a Hierachical Organization" Pfeffer J.: 1981. Power in Organizations, Pitman Publications, Marshfield Mass. 222 Porter, L. W.: 1962. "Job Attitudes in Management 1: Perceived Deficiencies in Need Fulfilment as a Function of Job Level," Journal pf Applied Psychology, 46: 375-84. Rabinowitz, S. and D. Hall: 1977. "Organizational Research on Job Involvement," Psmhological Bulletin, V. 84 No.2 pp. 265-88. Rosenbaum, J.: 1979. "Career Paths and Advancement Opportunities," in Alvaretz and Lutterman (eds) Discrimination $2 gmganizations, Jossey-Bass, San Francisco pp. 69-85. 1984. Career Mobility Am 2 Corporate Hierachy, Academic Press Incorporation, New York. Rosenberg, M.: 1979. Conceiving the Self, Basic Books, New York Schuman, H: 1966. "The Random Probe: a Technique for Evaluating the Validity of Closed Questions,".American Sociological Review. 31,2:218-22 April. Seeman, M.: 19773 "Some Real and Imagined Consequences of Social Mobility: ta French-American comparison," American Siegel, A. and R. Ruh: 1973. "Job Involvement, Participation in Decision Making, Personal Background and Job Behavior," Organizatigmgl Behavior and Human Performance, 9:318-27. Simon, H.: 1976. Administrative Behavior, Free Press N. York. Sofer, C.: 1970. Men Am Midcareer, Cambridge University Press, Cambridge. South, J., Corder, J. and W. Markham: 1982. "Sex and Power in the Federal Bureaucracy," Work and Occupations, Vol. 9 No. 2 May pp. 233—54. 1985. "Gender and Opportunity in the Federal bureaucracyy".American Journal 9£.§99121992r Vol. 91 No.1:129-150 July. 223 Spenner, K., L. Otto, A. Call and R. Vaughn: 1982. Career Lines and Careers, Lexington Books. 1981. DesAgm For a Study of Entry into Careers, Lexington Books, Lexington, Mass. Spilerman, S: 1977. "Careers, Labor Market Structure and Social Economic Achievement," American Journal o_f Sociology, 83:551-93. Suarez, P: 1981. "Alternative Coding Procedures and the Factorial Structure of Attitude and Belief Systems" in Jackson, D. and 3- Borgotta (ed), EEEEQE 522l1§i§ £29 Measurement Am Sociological Research, Sage, Beverly Hills. Tannenbaum, A. (ed.): 1968. Control Am Organizations, McGraw Hill, N. York. Tannenbaum, A” EL Kavicic, M. Rosner, M. Vimello and A. Wieser (eds): 1974. Hierarchy Am Organizations, Jossey- Bass, London Tharenou, P.: 1979. "Employee Self— esteem: A Review of the Literature, "gpmmmmA 9A Vocational Behavior, 15, 316- 46. Tausky, C. and R. Dubin: 1965. "Career Anchorage: Managerial Mobility Aspirations, " American Sociological Review, 30 (October), 725- 35. Thomas, A. B.: 1983. "Managerial Careers and the Problem of Controlfl' Social Science Information, 22(1) pp. 1-25. Waldstein, N.: 1957. The StruggAg for African Advancement within Copper Industry of Northern thesia: A Case Study, Cambridge Center for International _Studies, MIT Cambridge, Mass. Watson, T: 1977. The Personnel Managers: tu dy in the Sociology of Work and Employment, Routledge and Kegan Paul, London. 224 Wilensky, H.: 1961. "Orderly Careers and Social Participation: the Impact of Work History on Social Integration in the Middle Mass", American Sociological Review, 26 pp. 521- 39. Woldring, Klass and Chibwe, Chibaye (eds): 1984. Beyond Independence: Zambia's Development Predicament Am the 19803, Mouton Publishers, Berlin. Young, A.: 1973. Industrial Diversification Am Zambia, Praeger, N. York IC VAT 1 ER {WI/NW{ill/111017111!!! 3 1293 1071 111/1'1 9888 / 111/7,1! 111