IIIIIIIIIIIIIIIIIIIIIIIIIII 3 1293 )0607 7419 )V153I.J RETURNING MATERIALS: PIace in book drop to LIBRARJES remove this checkout from -:——. your record. FINES will be charged if book is returned after the date stamped below. If: a (a. {:3 ‘0 \j’ i;-"‘.,"!"~~-‘ .3 Juui?53¢N7§3 ’Jfltill t)/Q§ @5 8231/ A MALE-FEMALE COMPARISON OF PARTICIPATION IN A LOCAL LABOUR UNION BY Steven Lattimore McShane A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Social Science 1983 I“ /‘ 317 "r 5 ABSTRACT A MALE-FEMALE COMPARISON OF PARTICIPATION IN A LOCAL LABOUR UNION By Steven L. McShane The three objectives of this research were to: 1) test the multidimensionality of union participation, 2) empirically evaluate a motivation-opportunity model of union participation, and 3) com— pare the patterns of union participation between men and women union members as well as determine the factors contributing to any differ- ences in union activity levels. Data were gathered on 275 members of a public sector union local situated in southwestern Ontario through mailed questionnaires as well as personnel and union files. The response rate was over 60 percent and the demographic features of the sample were quite similar to those of the pOpulation. Information on the union involvement activities of the respondents was delineated into three types based upon theor- etical grounds and findings from previous research: administrative participation, meeting attendance, and voting participation, respec- tively. The results supported the hypothesis that union participation is multi-dimensional. Intercorrelations among the three dependent measures were quite low relative to the average intercorrelation within each scale. Moreover, most independent variables were dif- ‘ferentially predictive of the three union participation measures. Steven L. McShane Support for the union participation.model was also found. Variables such as education, seniority, employment status, and dis- tance to the union hall which were hypothesized via the model to constrain Specific forms of union activity did, in fact, signifi- cantly covary with the correSponding dependent measures. Variables associated with different personal needs -— including extrinsic sat- isfaction, friends in the union, and job involvement -- also corre- lated with the specific types of union involvement which were hy- pothesized to fulfill those needs. And, as predicted, the importance of union attitudes increased with the centrality of the union activ- ity to the operation of the union. That is, both the general atti- tude towards unions and the specific interest in union business variables correlated strongest with administrative participation, followed by meeting attendance and voting participation, respectively. Finally, this study reported only marginal differences between men and women with respect to the three types of union participation. Women were slightly less involved than men in the administration of the union, just as likely to attend union meetings, and slightly more likely to attend the union votes. None of the predictors accounted ' for these differences. Although women reported nearly twice the number of hours of housework per week than did men, this variable did not significantly associate with any of the three forms of union activity. None of the other predictors were significantly differ- ent between men and women. ACKNOWLEDGEMENTS No project of this magnitude can be successfully completed without the full support and assistance of many people. For their talents, ideas, and cooperation, I am deeply indebted. Without the permission of the union local, with its professional leadership, this study could not have even begun. Thanks are particue larly extended to Hugh Anderson, Monika Ballance, Andy Breschuk, Eleanor Deroy, Carol Ann Dunlap, Mike Ferrigan, Vince Marina, Sharon Murray, Gayle Weston, and Pam Young, who were more patient and coop- erative than I had ever hoped for. My gratitude is also extended to the union stewards who effectively communicated to the membership the importance and value of this research. The project could not have pro- ceeded nearly as smoothly without the full cooperation of the public sector organization that all of these people are employees of. Fi- nally, to the members who completed the lZ-page booklet: Thank You! Of course, research costs money. To the Research Branch of the Ontario Ministry of Labour, which paid for most of the direct costs of this study, I am very grateful. I would particularly like to ac- knowledge Maggie Smiley. As well as being an adept researcher in her own work, Maggie provided valuable liaison assistance between the gov- ernment agency and myself. Gratitude is also extended to Michigan State University which provided the necessary computer time and to 11 the Social Sciences and Humanities Research Council of Canada which paid the other bills in my life through a doctoral fellowship. Brainpower is a precious commodity which is an essential in- gredient in a project such as this. While I may have made a small contribution in this regard, much of the credit goes to my colleagues with whom I thrashed out ideas and sought solutions to theoretical and logistics problems. For their assistance in this respect as well as their friendship, I would like to thank: Richard Block, Leslie Corbitt, William Faunce, Lawrence Foster, Robert Heneman, Daniel Kruger, Michael Moore, David Pincus, Steven Premack, and Paul Reagan. Even with the help of all of these people, this project would not have been nearly as enjoyable, as rewarding, or as successful without the unfaltering support and encouragement provided by Donna McClement. iii Chapter TABLE OF CONTENTS Ammmms O O C O O O O O O C O O O O O O O O 0 LIST OF TABLES . . . . . . . . . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . . . . . . . . . I INTRODucrION . O O O O O O O O O O O O O 0 O O 0 General Introduction . . . . . . . . . . . . . Purpose of the Study . . . . . . . . . . . . . Outline of This Study . . . . . . . . . . . . II A MODEL OF UNION PARTICIPATION . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . A Reconceptualization of Union Participation . A Model of Union Participation . . . . . . . . Personal Needs . . . . . . . . . . . . . . General Union Attitude . . . . . . . . . . Opportunity to Participate . . . . . . . . Background Characteristics and Experiences Summary . . . . . . . . . . . . . . . . . . . III A LITERATURE REVIEW AND DISCUSSION OF HYPOTHESES OF THE VARIABLES IN THE STUDY . . . . . . . . . . Introduction . . . . . . . ',- . . . . . . . . Male-Female Differences in Union Participation Education . . . . . . . . . . . . . . . . . . Marital Status and Hours of Housework . . . . Seniority . . . . . . . . . . . . . . . . . . Employment Status . . . . . . . . . . . . . . Distance from Home to the Union Hall . . . . . Friends in the Union . . . . . . . . . Value of unions and Interest in Union Business Extrinsic Job Satisfaction . . . . . . . . . . Job Involvement . . . . . . . . . . . . . . . Intent to Quit . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . iv Page ii vi vii \IUIF‘ h‘ 10 19 21 25 29 3O 32 35 35 35 39 42 44 45 46 48 49 52 S4 54 Chapter IV METHODOLOGY . . . . . . . . . . . . . . Introduction . . . . . . . . . Selection of the Union Local . . . . . . . . . . Characteristics of the Union Local . . . . . . . Pretest Study . . . . . . . . . . . . . . . . . Sample for the Present Study . . . . . . . . . . The Questionnaire . . . . . . . . . . . . . . . Measurement of the Dependent Variables . . . . . . Administrative Participation . . . . . . . . . Union Meeting Attendance . . . . . . . , . . . Voting Participation . . . . . . . . . . . . . Independent Variables . . . . . . . . . . . . . . Value of Unions and Job Involvement . . . . . . Extrinsic Job Satisfaction . . . . . . . . . . Other Union/Job Attitudes . . . . . . . . . . . Union/Job Characteristics . . . . . . . . . . . Individual Characteristics . . . . . . . . . . Method of Analysis . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . V MALYSIS 0F DMA O O O O O O O O O O O O O O O O O 0 Introduction . . . . . . . . . . . . . . . . . Analysis of the Union Participation Model . . . A Male-Female Comparison of Union Participation . Summary . . . . . . . . . . . . . . . . . . . . VI DISCUSSION AND CONCLUSIONS . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . Limitations . . . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . Mhltidimensionality of Union Participation . . Validity of the Union Participation Model . . A Male-Female Comparison of Union Participation Conclusions . . . . . . . . . . . . . . . . . . . APPENDIX A: The Questionnaire . . . . . . . . . . . . . . BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . Page 60 60 60 63 65 66 67 69 69 71 71 72 73 78 78 79 80 81 82 84 84 84 . 95 103 105 105 105 108 108 110 114 116 122 134 10. 11. 12. 13. 14. LIST OF TABLES Page Female Union Membership in Canada . . . . . . . . . . . 3 ' Measures of Union Participation . . . . . . . . . . . . 12 Union Participation Types and Their Instrumentality in Personal Need Fulfillment O I O O O O 0 O O O O I O O O 26 Hypothesized Associations Among the Independent Var- iables and Three Forms of Union Participation . . . . . 55 Comparison of Specific Characteristics for the Sample and Population of the Union Local . . . . . . . . . . . 68 Item Analysis and Reliability of Value of Unions Scale 74 Item Analysis and Reliability of Job Involvement Scale 76 Factor Analysis of Value of Unions and Job Involvement 77 Union Participation for Men and Women . . . . . . . . . 36 Zero-Order Correlations and Standardized Multiple Re- gression Equations for Three Forms of Union Participa- tion: F1111 sample (N . 275) a s o e e e s e o e o o o 87 Means, Standard Deviations, and t-statistics of Inde- pendent Variables for Men and Women . . . . . . . . . . 96 Zero-Order Correlations and Standardized Multiple Re- gression Equations for Administrative Participation: Male and Female Samples . . . . . . . . . . . . . . . . 98 Zero-Order Correlations and Standardized Multiple Re- gression Equations for Meeting Attendance: Male and lee samples 0 O O O O O O O O O I O O I C O I O O O 99 Zero-Order Correlations and Standardized Multiple Re— gression Equations for Voting Participation: Male and FMle samples 0 I O O O O O O O O O O O O O I O O I O 100 vi LIST OF FIGURES Figure Page 1 A Model of Union Participation . . . . . . . . . . . . 20 vii CHAPTER I INTRODUCTION General Introduction Few issues in the field of industrial relations have received such keen interest as participation in labour union locals. Liter- ally dozens of books and articles have been written on the subject and no fewer than four published reviews have attempted to aggregate the mass of findings (Perline & Lorenz, 1970; Spinrad, 1960; Swanson, 1981; Tannenbaum, 1965). According to one source (Huszczo, 1975), _over 60 variables have been hypothesized as correlates of union par- ticipation. Why, then, should another study on union participation be con- ducted? To answer this question, the reader should be informed that most of the existing research was conducted prior to the mid-1960's - a period which one veteran has labelled the 'Golden Age' of union gov- ernment research (Strauss,'l977). This represents a valuable founda- tion for present-day investigation, but there are also numerous drawb backs. One problem is that the pioneering research consisted either of qualitative case studies or, at best, bivariate analyses of union par- ticipation (Nicholson et a1., 1981). Since other factors were not partialled out, some of these early findings could be spurious. Also, many of the conclusions have questionable validity because the instruments used to collect the data do not have the methodological rigour demanded in current research. As one writer has indicated, the quality of union participation research would be significantly enhanced by better scale development of the variables (Huszczo, 1975: 81). Another problem with the earlier research is that it was pre- occupied with the characteristics of union 'activists' rather than the correlates of union participation (Nicholson, 1978). For ex- ample, union members were arbitrarily bifurcated into 'active' and 'inactive' camps and then were compared for differences in back- ground characteristics and experiences. Aside from the loss of valuable information that this methodological routine causes, the delineation of union members into actives and inactives assumes (in- correctly, in our opinion) that union participation is a dichotomous rather than a continuous function (Huszczo, 1975). Methodological issues aside, new research into the dynamics of union participation is necessary because the characteristics of the labour movement have altered since the Golden Age of union partici- pation research. One difference is that public sector unionism has dramatically increased and today represents a large proportion of the labour movement in Canada. Yet few studies have been conducted on union participation in the public sector unions (Swanson, 1981). There has also been a dramatic increase in the proportion of the labour movement that is comprised of women. As Table 1 details, only 16.4 percent of Canadian union members were women in 1962. By 1980, however, the number of organized women in this country was Table 1 Female Union Membership in Canada: 1962 - 1980 Percentage Change Number 0f Percent from Previous Year Women of all Year Members Members Men Women 1962 248,884 16.4 .. .. 1963 260,567 16.6 +3.2 +4.7 1964 276,246 16.7 +5.3 +6.0 1965 292,056 16.6 +6.5 +5.7 1966 322,980 17.0 +7.5 +1o.6 1967 407,181 19.8 +4.6 +26.1 1968 438,543 20.4 +3.8 +7.7 1969 469,235 21.2 +1.9 +7.0 1970 513,208 22.6 +0.6 +9.4' 1971 558,138 23.5 +3.6 +8.8 1972 575,584 24.2 -o.8 +3.1 1973 635,861 24.6 +7.9 +1o.5 1974 676,939 25.2 +3.2 +6.5 1975 711,102 26.0 +1.0 +5.0 1976 750,637 27.0 +0.1 +5.6 1977 782,282 27.7 +0.6 +4.2 1978 835,263 28.7 +1.6 +6.8 1979 890,365 29.3 +3.5 +6.6 1980 932,883 30.2 +0.7 +4.8 Source: Statistics Canada (1982). approaching one million and represented 30.2 percent of the Canadian labour movement (Statistics Canada, 1982). There are several reasons for this change (McShane, 1981) including the increasing number of women entering the labour force, the rapid growth of public sector unionism, the shift toward formal collective bargaining of several female—dominated employee associations (e.g. nursing, teaching), and the increasing efforts of the Canadian Labour Congress and its affiliates to organize financial services and other industries where women are typically employed. Unfortunately, few studies have at- tempted to directly compare male and female union members in terms of their involvement in the operation of the union local and the factors which contribute to or constrain this union participation. Two other reasons for conducting another study of union partici- pation should be briefly mentioned. A small group of researchers have quite recently started to question the premise that union participa- tion is a unified concept that can be measured with a single global index (Portwood et a1., 1981). It is useful.to refine this construct so that our understanding of the subject can be enhanced. The other reason for continued research is to develop and validate a theoret- ical model of labour union participation. Two research teams have begun this task (Nicholson et a1., 1981; Portwood et a1., 1981), but there is still much to be desired. Future endeavours would enable us to clarify relationships among variables and formulate a coherent theory of union participation. It is with the above-mentioned reasons in mind that we have de- cided to contribute to the subject of local labour union participa- tion. Purpose of the Study The original objective of the present investigation was to compare male and female union members in terms of their relative levels of union participation and the factors which contribute to any differences in union participation. This remains the major (ultimate is perhaps a better word) purpose of our research as the title of this manuscript suggests. But our initial attempts to understand the meaning and operational interpretation of union par- ticipation as well as to search for a theoretical model within which to house the research were less than successful. Rather, this writer found himself questioning the basic assumptions behind earlier (in- cluding most of the recent) union participation literature. Does it make sense to group different types of union participation be- haviours into a global index? Should not our theory consist of more than a list of independent variables? Why are some predictors (e.g. job satisfaction) significant in some studies but not in others? Is there a problem with representing union participation by different union behaviours in different studies? These and other queries forced us to redefine our research ob- jectives to include the testing of a reconceptualization of union participation and a model of union participation. We say 'forced' because without a clear and concise understanding and operational- ization of the dependent variable and without an appropriate model within which to conduct our research, we would not be satisfied with the investigation of male-female differences in union participation. Accordingly, there are three major objectives to this research. First, we hope to refine the meaning of union participation by illus- trating that it is more appropriately viewed as a multidimensional construct. That is, union participation should be thought of not as a single domain of homogeneous behaviours, but as a set of dis- tinct clusters of behaviour which should be studied independently rather than together in a global index. This reconceptualization is described more fully in Chapter II. Our second objective is to empirically test the validity of a model of union participation which we have developed for this (and presumably future) research on the subject. Some parts of the model are tested more directly than others. For instance, although per- sonal needs comprise a significant part of the theoretical framework, these needs are only inferred from other variables because of prob- lems with directly operationalizing personal needs in a questionnaire research design. A full description of the union participation model is presented in Chapter II and the variables which are analysed are discussed in the context of the model in Chapter III. With a sound understanding of union participation and a model within which to study the dynamics of local labour union involvement, our third and final objective is to compare male and female union mem- bers in terms of their relative level of union participation and the factors which contribute to any sex differences in union participa- tion. The majority of hypotheses in this regard focus on male-female differences in the magnitude of the variables associated with union participation. However, we also explore any differences between male and female union members in the structures of the regression equations to determine if any of the predictors have a differential effect on men and women. The hypotheses specifically referring to male-female differences in union participation are discussed in Chapter III. Outline of This Study In this chapter we have described several reasons why more re- search on union participation is needed and have stated the three major objectives of the present investigation. Chapter II begins with a reconceptualization of union participation and is followed by a detailed description of a model of union participation. A sel- ective literature review is included in these respective discussions. Chapter III includes a review of the literature as it applies to the independent variables in the analysis. Specific hypotheses are advanced with respect to a male-female comparison of union par- ticipation and to the validation of the union participation model. The procedures employed to select the union local, operation- alize the independent and dependent variables, and analyse the data are described in Chapter IV. Special attention is paid to the repre- sentativeness of the sample, the appropriateness of the multidimen- sional conceptualization of union participation, and the reliability and construct validity of the variables. Chapter V presents the results of the data analysis on our sample and is divided into two parts. First, we compare men and women in terms of their levels of union participation and analyse the multiple regression results using the full sample (i.e. men and women combined). The latter half of the chapter compares the male and female samples on the independent variables and explores the structure of multiple regression equations for the male and female sub-samples. The final chapter in this study, Chapter VI, discusses the re- sults in terms of the broader objectives of this research and pro- poses several implications with respect to our understanding of the subject of union participation and the directions for future research. The limitations of the present investigation are also outlined. on the independent variables and explores the structure of multiple regression equations for the male and female sub-samples. The final chapter in this study, Chapter VI, discusses the re- sults in terms of the broader objectives of this research and pro- poses several implications with respect to our understanding of the subject of union participation and the directions for future research. The limitations of the present investigation are also outlined. CHAPTER II A MODEL OF UNION PARTICIPATION Introduction The earlier research on union participation has concentrated its efforts on the identification of variables which are associated with participative behaviour. This work has resulted in a list of correlates which have been classified into manageable groupings such as demographic characteristics, job characteristics, union charac- teristics, and so on (Perline & Lorenz, 1970; Spinrad, 1960; Swanson, 1981). However, there remains a large gap in our theoretical under- standing of these relationships since most explanations have been admittedly post hoc. That is, it is still relatively unclear why these covariances exist. Another concern with our current understanding of the subject is that the union participation construct is not unambiguously de- fined. Consequently, the behaviours included in measures of union participation vary considerably from one study to another. Activi- ties which many of us would consider to be critical to the Operation of the union local are excluded from some investigations, while other behaviours with marginal relevance to the union participation domain are regularly included. Furthermore, assumptions about the unidimensionality of union participation are left untested, result- ing in conflicting or inconclusive results. 9 10 Before we can compare the participative behaviours of male and female union members, we must first resolve the issues regarding the conceptualization of union participation and pull together the ex- isting research evidence into a comprehensive model at the individ- ual level of analysis. These are the objectives of this chapter. We begin by observing how union participation has been conceptual- ized by studying the operational interpretations of the construct in the literature. Specific issues are raised, especially regarding the multidimensionality of union participation, and a solution is advanced to correct these problems. The second half of the chapter describes a model of union par- ticipation and includes a selective review of the literature per- taining to the various components of the model. Specifically, we discuss the relevance of personal needs, general union attitude, the opportunity to participate, the willingness to participate, and background characteristics and experiences to our understanding of the dynamics of the individual union member's participation in the union local. A Reconceptualization of Union Participation Union participation has been defined as the behavioural involve- ment of union members in the Operation of their local labour organ- ization (Tannenbaum & Kahn, 1958: 50). This conceptualization is. quite similar to other definitions of union participation (see for example: Ramaswamy, 1977; Sayles & Strauss, 1953; Tannenbaum, 1965), but it is nevertheless quite vague. Although one writer views the definition of union participation as "necessarily imprecise 11 (Ramaswamy, 1977: 470), we believe that it has resulted in conflicting results and generally fragmented research. Therefore, before we can even begin to approach the literature on the determinants of union participation and understand how men and women differ in this regard, it is first necessary to understand how the construct has been de- picted and to resolve the conflicts that are present. These are the objectives of this section. ‘. Our analysis begins with a review of the operational interpre- tations of union participation. Drawing upon the measurement of the construct to understand how it has been conceptualized in the liter- ature is both necessary, since few researchers have bothered to de- fine their terms, and desirable because the measurement of a vari- able is the ultimate and clearest indicator of how it is perceived. In his review of the early union participation literature, Spinrad (1960: 238) observed a variety of union participation in- dexes with only a few activities common to most indexes (e.g. meet- ing attendance, holding office, voting in elections). Nevertheless, he concluded that the research findings were generally unaffected by the indexes used. Our review of the research to date also finds a variety of indexes, as Table 2 illustrates. However, only meeting attendance is employed by all 17 studies listed and union election voting is the only other form of union involvement measured in the majority of studies. The diversity of Operational interpretations of union participa- tion is further indicated in Table 2 by the fact that over 20 types of union activity (including seven behaviours in the "Other" category) have been included in one index or another. Mbreover, the number 12 Avoncuucoo o~nmuv moquw>wuom nonuo mo soduoemmaxo uOu axon some mooswcandus no consumes woman o~amaam> aeovcoeoeeu so on vo>O~asm6 mowvsum us human: M m Q \0 \O h CD H H vs 9-4 AssaAO useuouaaau «6 ssauuo>aes insane cases: a soaaoeuuuz “ensue snag a asaaeoceae isno~v Haouuas Aaaaae .aa 88 sooasuoe Rename .H6 66 cooaoeuaz “cease cause a comma: flamed. seconds“ Aoaaae ensues AnsoAV ouuuaae Aoamav commons: Aaemae Assess spasms: Aomaav .m6 86 :ovuoo axeAAAV .a4 so sumac Aanodv .aa uo oumou Asmaae came Aoaaav =66u6u=< N xxxxx. xxxxxxxxxxxxxxxxx x x o>H x x x momma uuo> mandamus um mouuuc oouuuo sound-loo ooso>ouuu msowuoouu monorOOuu< some: oxauum secuusosa sous: sows: sous: a mafia sous: magnum: comm scouu< xm<\xmoam now an: ego: so o>uom an ouo> noaumafioauumm sous: mo mounmmoz N manna 13 mouuw>auoo nosuo uo souuoeoHaxo uOu has» some mmoswewudwa no nauseous roman o~aa«ue> ueovcoeoesq so no vo>o~asma n mowvsum mo uoasnz accede useuomaaao so ssaou6>aea insane :66H6z a 66au6e666: x Aenoav seas a asoae6ecse Rename HH666=A x “Heady .H6 66 vooasuom Aaoaae .H6 66 souaoeuaz Amaaae 63666 a =66~6z AammAV =6eoaasx immune caeoox x finesse 66666:: iaaaae soaautaa Assaae escape spasms: Access .au 66 666666 sausage .H6 66 suede impede .au 66 6666s assess :66: Aaa6~V =66666:< omo«u«>«uo< huuuoasom cows: onu saws: assoc \auamuonaoz :« macaqum coed: o>m= mandamus on access: veooom\oxsx muco>m memumoum Howoom :Ouumosou Ouo> noesauamasue casuu< sows: casuu< uoauusoo assuu< Ae6=aaoeouv N 6Hn6e 14 of items included in a given index of union participation varies substantially. Five investigations rely solely on meeting atten- dance data tO estimate the degree of each union member's involve- ment. Among the 12 other studies, only two include exactly the same set of participation activities, and this similarity is found only because one is a replication of the other (Geare et a1., 1979; Wertheimer & Nelson, 1975). The diversity of union participation instruments would not be a problem if all items were equally representative of the same con- struct. There is reason to believe, unfortunately, that this is not the case and that the lack of consensus on the behavioural set of union activities is the source of different effect sizes for some predictors and mixed results for others. This statement is based upon two prepositions which we will discuss in some detail below: 1) union behaviours vary in the degree to which they are central to the operation of the union and 2) union behaviours vary to the extent that they are a function of different influences. The proposition that union behaviours vary in the degree to which they are central to the Operation of the union can be iden- tified throughout much of the union participation literature. Even in the earliest research there has been an awareness that some union activities require more time and energy than others and have more of an impact on the union's Operation (Sayles & Strauss, 1953; Spinrad, 1960; Tannenbaum.& Kahn, 1958; vall, 1970). In par- ticular, several statements have been made to the effect that hold— ing union office requires a "higher activity level" (Sayles & Strauss, 15 1953) or "more intense involvement" (Spinrad, 1960) than other forms of participation. Nicholson (1978) has taken this prOposition one step further by introducing a continuum of union involvement within which are ranked a variety of union activities. At the upper end of the continuum is direct involvement in the administration of the union local. This would include holding office (elected or appointed), running for union Office, and possibly actively engaging in some union meeting functions (e.g. presenting motions, answering questions, etc.). At the lower end of the participation continuum are the extremely "pas- sive" forms Of union behaviour including reading the union newspaper and "the act Of membership" (Nicholson, 1978: 37). However, as Nicholson points out, some of the lower activities might be so pas— sive as to cease to merit inclusion as a form of participation under certain conditions. Returning to Table 2, we find that the 17 studies vary in the degree to which the indexed activities are central to the Operation of the union. For example, only six research teams included holding union office in their index of participation although it is the most obvious and central form of union participation according to the def- inition. In contrast, information on grievance filing and attending union votes was collected in a number of surveys. A few investiga- tions even measured such behaviour as reading the union paper, being a union member, and having friends in the union, yet ignored other activities more central to the operation of the union. Based upon these observations, we suspect that some research findings will vary I 16 from study to study because the measures of union participation vary in their degree of centrality to the Operation of the union. The second proposition states that union behaviours vary to the extent that they are a function of different influences. In other words, the things which influence a union member to participate in one type of union activity differ from those which affect involve- ment in another type of union activity. This prOposition has been recently confirmed by Nicholson and his colleagues (1981) who dis- covered that the predictive power Of different needs and attitudes vary with the type of union activity. For example, individuals with a highly favourable union ideology are more likely to attend union rallies but are not significantly more likely to vote in union elec- tions than members with less positive union attitudes. Also, union members with strong affiliation (i.e. social) needs are more likely than Others to get involved with group activities such as union meetings and rallies (Hagburg, 1966; Nicholson et a1., 1981). The implication of this proposition is that union participation is not really a unidimensional concept and, therefore, should not be represented by a single global index which combines different types of union activities. For instance, Portwood and Others (1981) have called for a refinement of the union participation concept by delin- eating union meeting attendance from union voting behaviour. This recommendation is drawn out of their findings that voting behaviour primarily fulfills instrumental needs (i.e. improving job-related concerns) while union meeting attendance is more apprOpriate for the fulfillment Of expressive needs (i.e. the desire to be involved in the decision—making process). Using factor analysis, McShane (1982) l7 discovered that union voting behaviours grouped well together but were not strongly associated with administrative behaviours such as holding union office. Together, these analyses suggest that our understand- ing of union participation would be greatly enhanced if participation in union administration, union meeting attendance, union voting be- haviours, and possibly other clusters of union activity were studied separately rather than as part of a global measure. The reader will recall that most of the studies listed in Table 2 have combined the union participation data into a single global index. Considering what we have just said about the multidimensionality of union participation, several problems can be identified with the emr ployment of a global index. First, assuming a good representation of items across the behavioural spectrum, a global index will have a relatively low internal consistency. Few studies have stated the re— liabilities of their indexes, but the values that have been reported are only in the 0.5 to 0.6 range (Anderson, 1979; Nicholson et a1., 1981). Since the measured association between two variables is partly a function of the reliabilities of those variables, effect sizes will be diminished in studies using global measures of union participation. In particular, the relevance of predictors which correlate with some forms of union participation but not with Others will be diluted. A final concern is that since no two global indexes of union participa- tion appear to be equally representative of all clusters of activities, effect sizes for many predictors probably vary considerably from one study to another. For instance, dissatisfaction with the job con- text (e.g. pay, promotions, and supervision) would be highly corre- lated with a global measure which was weighted toward activities l8 fulfilling this need (probably such as voting in the strike and contract ratification votes), but not with another purported meas- ure of union participation weighted more toward union administra- tion behaviour. To summarize our discussion thus far, the broad and imprecise definition typically given to union participation is responsible for a multitude of Operational interpretations of the construct. This diversity of measures is likely to have created inconsistencies in previous research because union behaviours differ in the degree to which they are central to the operation of the union and to the ex- tent that they are influenced by different factors. One solution to this problem would be to study several homogeneous clusters of union activities such as union administration, union meeting at- tendance, and voting in special union decisions. This would not only solve the problem of analysing a less meaningful global index (which is basically putting "apples and oranges" together), but would also reduce the diversity of measurement scales employed and thereby increase the consistency of results from one study to another (e.g. union meeting attendance is less ambiguous than union partici- pation). With respect to a review of the literature, the conceptu- alization of union participation that we have put forth also suggests that we look more closely at the operational interpretation of the construct to understand apparent contradictions and ambiguities in the research findings. 19 A MOdel of Union Participation Although the major problems with the conceptualization of union participation have been resolved, we still face a fragmented array of predictors scattered across the pages of academic journals. One writer who thoroughly reviewed the literature identified over 60 var- iables which have been tested at one time or another as correlates of local labour union participation (Huszczo, 1975: Table 2 to Table 8). The situation is not unlike the plight of industrial relations research that Dunlap (1958) described over 20 years ago. Mbuntains of facts (on union participation) are awaiting an integrating theoret- ical framework to pull everything together. The objective of this section is to describe a model of union participation which will pro- vide a framework for our research. A selective review of the liter- ature is also included in our description of the model, although a more detailed review is presented in Chapter III. Figure 1 outlines the model of union participation that forms the theoretical foundation of our research. The diagram illustrates that involvement in the union is a function of background character- istics and experiences, personal needs, general union attitude, the willingness to participate, and the Opportunity to participate. Con- sistent with our discussion in the previous section, union participa- tion is conceptualized as a multidimensional construct consisting of three types of union behaviour: administration of the union local, union meeting;attendance, and union voting_behaviour. These three groupings are not intended to exhaustively cover the range of union activities under the rubric of union participation. However, we believe that they clearly fall within the definition of union coauma«oauumm does: «0 HooOz_< "a ouswam 20 633»: no“; TIIII Hmuomoo mooeoahomxm mcaao>o msdvo>o ooosocoav< msauoozo ooeooooa¢< wouaoozo all 615866222. > 6>385355<6 no :ofiasafid 63.5 384386.“ 3 666.952: 6 528a. assoc? 6636236820 Begonia? _ memo: Hosomnom oesoumxoom 666363.48 oa .A finest—Ego 21 participation and, as mentioned earlier, represent different clusters of behaviour. In the remaining pages of this chapter, we will outline in some detail the various components of our model of union participa- tion. Personal Needs. Running throughout the union participation lit- erature is the idea that union members have different needs which motivate them to become involved in those union activities which are instrumental to the fulfillment of these needs (Lipset et a1., 1956; Perline & Lorenz, 1970; Strauss, 1977; Tannenbaum & Kahn, 1958; Vall, 1970). Unfortunately, as Nicholson and his cowriters (1981: 55) point out, this perspective somehow is lost in the main theme of most re- search efforts and is under-emphasized in literature reviews on the subject: Certainly the literature indicates that member's attitudes and needs are potentially compelling forces for participatory behaviours, but ones that are surprisingly neglected in empirical re- search (perhaps because of their very obviousness) in favour of a focus on more remote correlates. The model in Figure 1 attempts to bring this neglect to an end by in- cluding personal needs as a major set of determinants of union partic- ipation. Although dozens of personal needs have been identified in the psychological research, we have grouped these into three categor- ies - job-related, social, and growth needs - which have been high- lighted in much of the union participation literature. For the majority of organized employees in North America, involve- ment in the local labour union is an occasional activity which is mo- tivated primarily by job-related needs, namely, to maintain or improve working conditions (Tannenbaum, 1965). For others, the job that they currently hold is not such an important part of their lives that they find it worth their while to improve conditions through union partici- pation. For example, unionized temporary employees would probably be less involved in the union as a result of their limited term in the job and with the organization (Seidman et a1., 1958). As a few writers have explained, the greater the employee's 'stake' in the job, the more likely they are to participate in the affairs of the union local (Strauss, 1977; Tannenbaum & Kahn, 1958). Not all forms of union participation serve job-related needs equally well. Involvement in the administration of the union might ensure that the employee's job conditions improve, however the task requires a high degree of time and energy. Attending union meetings might serve job-related needs, but this function is also time con- suming and is directed more to the internal operation of the union than directly improving job conditions (Sayles & Strauss, 1953). These propositions have not only appeared indirectly throughout the early case studies, but have been directly confirmed in the only study to measure economic (i.e. job-related) needs (Click et a1., 1977). Specifically, the authors reported near-zero correlations between economic needs and the willingness to either represent the union or attend regular meetings. For union members who view the union local as a vehicle for job improvement, participation is primarily contemplated in the form of attending the strike and contract ratification votes (Portwood et a1., 1981). These activities require minimal time and energy, yet have a rather noticeable effect on job-related conditions. Not surprisingly, then, the proportion of the membership involved in these voting behaviours is much greater than for most other forms of union participation (Sayles & Strauss, 1953; Seidman et a1., 1958). Social needs represent the second set of personal needs in our model in Figure l and refer to the desire for interpersonal accep- tance, contact and communication. As with job-related needs, social needs are more important to some union members than others. Further- more, social needs are more likely to be fulfilled by certain forms of union participation than by others. While voting in the strike and contract ratification decisions often occur in social settings, they are typically of limited duration and offer little opportunity for social interaction. Participation in the administration of the union local has been viewed as a social experience (Sayles & Strauss, 1953), but this writer's Observation is that these positions often require working alone (e.g. working out union finances) or in highly task-oriented forms of interaction. This leaves union meeting attendance as the type of union ac- tivity most likely to attract members with high social needs. This proposition has been substantiated by Nicholson and his coauthors (1981) who recently reported that the need for affiliation is a strong predictor of attendance at union meetings and rallies, but correlates more modestly with voting in union elections, raising issues or grievances, and speaking at meetings. Notice, also, that these results support a long recognized observation that meeting attendance does not necessarily imply formal involvement in the meeting's business (Tannenbaum & Kahn, 1958). Prepotent social needs might motivate the union member to attend the monthly 24 meeting, but they will have little or no effect on speaking out, raising issues, or other formal involvement at the meeting. Growth needs, the third and final set of needs in our model, include the desire for challenge, responsibility, power, decision- making, and other 'higher-order' needs. As with the other need groups, growth needs are more prepotent among some union members than others and are more likely to be satisfied through participa- tion in certain union activities than others. Direct involvement in the administration of the union appears to be the major form of union participation to satisfy growth needs. FOr example, Seidman and his associates (1958) discovered that union members who are active in the union have a greater desire for power and recognition than those who were not involved in union administra- tion. Hagburg (1966) reported that active members feel a greater sense of accomplishment than inactive members and Nicholson and others (1981) computed a statistically significant correlation between union participation and the desire to participate in management decisions. As Spinrad (1960: 242) concluded his re- view of the earlier literature: union involvement can make the job situation much more pleasant and meaningful by providing an area for creativity and interpersonal influence and an avenue for status not found in the job itself. To a lesser extent, participation (not just attendance) at union meetings appears to serve growth needs as well. Click and his cowriters (1977) found that respondents with strong decision- making needs are more willing to represent the union as well as at- tend union meetings. And Portwood et al. (1981) concluded that 25 participation at meetings is instrumental to the fulfillment of the need to be involved in the union process whereas voting in union deci- sions fulfills more job-related needs. Table 3 summarizes the above discussion on personal needs and their predicted (and in many instances, already demonstrated) asso- ciation with the three clusters of union participation behaviours. Turning first to participation in the direct administration of the union, we see that these forms of union activity are highly instru- mental to the fulfillment of growth needs, moderately instrumental for social needs, and represent a relatively poor means (consider- ing the time and energy required) of improving the participant's own job-related situation. Union meeting attendance is somewhat broad- based in terms of the personal needs it serves. Although attendance tends to best fulfill social needs, the union member might also re- ceive growth and job-related satisfaction if he or she actually par- ticipates in the business of the meeting. Finally, voting in union decisions is typically very instrumental to the fulfillment of job- related needs and, to a lesser extent, can be a means of providing social interaction where the vote involves the assembly of the meme bership. It is unlikely, though, that union voting meets growth needs given the brevity of the event and the relative insignifi- cance of the individual's efforts. General Union Attitude. Personal needs motivate union members to become involved in union activities which are associated with those needs, but there is a substantial body of literature which suggests that union participation - at least.some forms - is also a function of the union member's attitude towards unions in general. 26 Table 3 Union Participation Types and Their Instrumentality in Personal Need Fulfillment Administrative Meeting Voting Personal Needs Participation Attendance Participation Job-Related Needs Low Medium High Social Needs Medium High Medium Growth Needs High Medium Low 27 Several research efforts have identified the respondent's attitude toward unionism as either the best or one of the best predictors of union participation (Huszczo, 1975; Sayles & Strauss, 1953; Seidman et a1., 1958; Tooredman, 1981) and the willingness to work for the union (Gordon et a1., 1980). The extent to which general union attitude is predictive of each of the three forms of union participation has not yet been clearly established, but several studies suggest that the associa- tion is stronger for union activities which are centrally located in the operation of the union local. For example, Flanders (1970: 44) has commented that "every union must have at least a core of active members who feel some deeper loyalty." Others have ob- served that the most active members, and in particular those who hold union Office, have the strongest ideological beliefs in unionism (Purcell, 1954; Sayles & Strauss, 1953; Seidman et a1., 1958). The reader will recall from our discussion on the conceptual- ization of union participation that involvement in the administra- tion of the union is probably most central to the operation of the union local (Nicholson, 1978). Thus, it is not surprising that the literature finds the incumbents of these positions to have values which are highly favourable to unionism. But what about the associations between attitude toward unionism and the other two forms of union activity? Uphoff and Dunnette (1956) reported not only that union leaders have more positive union attitudes than nonleaders, but that general union attitude increases with the per- centage of meetings attended. Glick and Others (1977) computed positive correlations for Opinions of unionism with both the willingness to represent the union and the willingness to attend regular union meet- ings, but only the former association was statistically significant. Finally, Nicholson and his colleagues (1981) correlated the respon- dent's union ideology with several specific participation behaviours. Generally, the coefficients were largest for attendance at union ral- lies, smaller for meeting attendance, and near zero for voting in union elections. From these findings, we are confident in stating the proposition that the association between general union attitude and union partici- pation increases with the degree to which the union activity is central to the operation of the union. We would expect, therefore, to find a strong relationship between general union attitude and direct union administration activities, a more moderate association for attendance at union meetings, and only a small relationship for union voting be— haviours. The explanation of this phenomenon is beyond the scope of this investigation, but is probably tied into the idea of 'person-role' conflict (Miles & Perreault, 1976). That is, union roles closest to the core of the union organization probably require the incumbent's values to be more closely aligned with those of the labour movement whereas this is less important for those participating in more periph- eral activities. Where the person's values are incompatible with those of unionism, person-role conflict will occur, causing psycho- logical hardship for the incumbent. Alternatively, the union member- ship would screen out these people through the union election and appointment processes. 29 Opportunity to Participate. While personal needs and general union attitude affect the willingness to participate, as illustrated in Figure 1, actual participative behaviour is also contingent upon the opportunity to engage in the desired activities (Tooredman, 1981). Put differently, situational contingencies intervene in the linkage between the willingness to participate and the actual participation event (for a discussion of situational contingencies, see: Brayfield & Crocket, 1955; Herman, 1973). Consistent with the perspective we have taken throughout this chapter, the Opportunity to participate in a particular type of union activity is likely to be determined by a different set of situational contingencies than involvement in another type of union activity. Thus, one would expect that the variables preventing or facilitating involvement in, say the administration of the union local are dif- ferent from those which affect voting in union decisions. There are several types of Opportunity variables. One set of factors act as gate keeping criteria in the passage into certain formal roles. For example, evidence suggests that most union mem- bers usually prefer to vote for more senior and educated union elec- tion candidates in higher status jobs (Geare et a1., 1979; Sayles & Strauss, 1953; Vall, 1970; Wertheimer & Nelson, 1975). It is par- ticularly noteworthy that although employees with higher education, more seniority, and higher job status are more likely to hold union office, they are no more willing than others to participate in union administration (Glick et a1., 1977; Gordon et a1., 1980). In short, union members with these characteristics are more likely to get voted into union office. 30 Other situational contingencies mediating between the willingness to participate and actual participative behaviour include legal and time-space variables. In some union locals (such as the one we have studied), temporary employees are required to pay union dues and are permitted to attend meetings and hold a few appointed posts, but they do not have the right to vote in union decisions. Others are impeded from attending union meetings either because their hours of work con- flict with the time the meeting is held (Geare et a1., 1979; Seidman et a1., 1958; Wertheimer & Nelson, 1975) or because the union hall is a considerable distance from their home (Miller & Young, 1955; Nelson & Grams, 1978; Rosen & Rosen, 1955). Finally, many union members who would like to attend meetings or engage in other time-consuming union events are unable to because obligations elsewhere "crowd out" these activities from the union member's itinerary (Harrison, 1979). ‘ngkground Characteristics and Experiences. The experience and characteristics which make up the profile of the individual union mem- ber pervade our model of union participation. As Figure 1 clearly shows, personal need, union attitude, and opportunity variables are all a function of the individual's background. Below, we briefly out- line some of these characteristics and experiences and their associa- tion with other components of the model. Our discussion of situational contingencies in the previous sec- tion is largely based upon the impact of different personal background variables (e.g. seniority, education, hours of work, distance from home to the union hall) on the opportunity to become involved in var- ious forms of union participation. In this respect, the Opportunity 31 for the union member to become active in the union is prevented or facilitated by his or her personal attributes. The individual's attitude toward unionism is also determined by personal experiences and associated characteristics. Much of the research on the determinants of union attitude focuses on the individual's socialization experiences, particularly the union at- titudes of significant others. Several investigators have found that the general union attitude of one's parents (Nicholson et a1., 1981), friends and coworkers (Gordon et a1., 1980; Maxey & Mohrman, 1980), and spouse (Tannenbaum & Kahn, 1958) influence the individual's own attitude toward unionism. Carrying this analysis one step further, McShane (1983) discovered that the strength of the relationship de- pends upon how recent and intimate the significant other is in the person's life. Thus, the union member's own attitude toward unionism correlated the strongest with his or her spouse's union attitude, fol- lowed by the general union attitude of friends and parents, respec- tively. Union members with higher education have consistently (except for Kochan, 1978) been found to have more negative union attitudes than members with less education (DeCotiis & Le Louarn, 1981; Gordon et a1., 1980; Huszczo, 1975; Maxey & MOhrman, 1980; McShane, 1983; Uphoff & Dunnette, 1956). One hypothesis for this is that exper- iences in higher levels of formal education cause individuals to shift their attitudes and beliefs away from a pro-union position (Stern & Murphy, 1980). Other background variables (e.g. age, sex) have brought forth mixed results with general union attitude in the literature. 32 For the purposes of this study, background characteristics and experiences are most relevant to the personal needs component of our union participation model, particularly because most research has inferred personal needs through these variables rather than directly measuring these needs. Thus, prepotent job-related needs have been identified with older, long tenure, full-time, and married employees since they have a greater 'stake' in the job (Perline & Lorenz, 1970; Spinrad, 1960; Strauss, 1977; Tannenbaum, 1965). Social needs have been identified with the level of coworker interaction off the job and the proportion of friends in the union (Anderson, 1979; Dean, 1954; Hagburg, 1966; Nelson & Grams, 1978; Nicholson et a1., 1981). And growth needs have been associated with the activities that the union member engages in off the job (Hagburg, 1966). The use of background characteristics and experiences as prox- ies for personal needs is slowly being replaced by more direct meas- ures (for example, see: Glick et a1., 1977; Nicholson et a1., 1981). Nevertheless, the analysis of most types of personal needs and union participation is likely to be relatively indirect until more reliable and valid measures of these personal needs can be developed (see: Dreher & Mai-Dalton, 1983). Moreover, the background features of the union member are the most salient variables and will probably remain the primary interest to most students of the subject. Summagy The purpose of this chapter has been to clarify the meaning of union participation and to present a model which will provide a theoretical framework for our research. A review of the measures 33 of union participation revealed that there is little consistency in how researchers have Operationalized (and by inference conceptu— alized) the construct. This diversity is probably responsible for some confusion in the research findings since union participation behaviours vary both in the degree to which they are central to the operation of the union local and to the extent that they are a func- tion of different influences. That is, some union activities require more time and energy and are closer than others to the heart of the union operation. Also, the factors influencing participation in one type of activity might be unimportant to another form of union in- volvement. The solution to this problem is to view union participa- tion as a multidimensional concept. Thus, it is necessary to sep- arately analyse the correlates of each homogeneous cluster of union activities, such as administration of the union, attendance at union meetings, and voting in union decisions. The second part of the chapter outlined a model of union partic- ipation and included a partial review of the literature pertaining to the various components of the model. Personal needs (job-related, social, and growth needs) and attitude toward unionism are described by the model as the central influences on the willingness to partic- ipate in the union local. Consistent with our conceptualization of union participation, however, different needs and attitudes influence the willingness to participate in different forms of union activity. Voting participation is most instrumental to the fulfillment of job- related needs, followed by social and growth needs, respectively. Attendance at union meetings can potentially satisfy social needs, but job-related and growth needs can, to a lesser extent, be 34 fulfilled if the individual also becomes active in the business of the meeting. Participation in the administration of the union is most instrumental to the fulfillment of growth needs, followed by social and job-related needs, respectively. Finally, union adminis- tration and, to a lesser extent, meeting attendance behaviours re- quire the role incumbent to have a pro-union attitude, while this is not likely to influence the union member's attendance at strike or contract ratification votes. The model of union participation presented in this chapter also illustrates the intervention of Opportunity (also called situational contingency) variables between the willingness to participate and actual participative behaviour. These factors include attributes considered by voters to be relevant in the selection of peOple for union office as well as legal and time-space barriers and facili- ' tators to various types of union activity. Finally, the model rec- ognizes the myriad of background characteristics and experiences which are associated with the personal needs, union attitude, and opportunity variables. CHAPTER III A LITERATURE REVIEW AND DISCUSSION OF HYPOTHESES OF THE VARIABLES IN THE STUDY Introduction The purpose of this chapter is to apply the previously described model of union participation to a comparison between the involvement of men and women in their local labour organizations. In the pages that follow, the relevant literature is reviewed and hypotheses are advanced for each of the 13 variables selected for analysis. We begin by looking at the male-female differences in union participa- tion cited in the literature. Six of the 12 other variables - edu- cation, marital status, hours of housework, seniority, employment status, and intent to quit -- have been included not only because of their importance in the published research, but also because they are each expected to partially explain why women have generally been less involved than men in the operation of their unions. The six remaining variables - distance to the union hall, friends in the union, value of unions, interest in union business, extrinsic job satisfaction, and job involvement - have been incorporated into this study to represent the basic model of union participation. Mele-Female Differences in Union Participation It is now well documented by research from several countries that women are generally less active than men in the Operation of their 35 36 local labour organizations. In Canada, three investigations have reported that men are more likely to hold union office (Geoffrey & Saints-Marie, 1971; Marchak, 1973) or score higher on a general index of union participation (Anderson, 1979). It is noteworthy, however, that the difference between men and women was not statistically sig- nificant in Anderson's (1979) survey and that there was no sex dif- ference in the propensity to be active in union decision-making. Yet the individual's perceived influence over union decisions was lower, on average, for women compared with men. Quite possibly, the Eypg§_ of decisions that women participate in are further removed from the core of the union operation than those which men participate in. Researchers in the United States have, for the most part, con- firmed the idea that women are generally less involved than men in their labour unions. This difference is particularly evident in the earlier case studies (Purcell, 1954; Sayles & Strauss, 1953) where women were virtually non-involved. But more recent investigations have also found that women are still proportionately less likely to be found in leadership roles (Abicht, 1976; Divers, 1981; Tooredman, 1981; Wertheimer & Nelson, 1975). Other research has reported that women members are significantly less willing to work for the union (Gordon et a1., 1980) and, more specifically, less willing to attend union meetings and represent the union (Glick et a1., 1977). Not all of the American research has found a large imbalance in union participation between men and women. A recently published multi- union study discovered that men typically had much better attendance records at some types of union meetings and voting sessions, but women were somewhat (although not significantly) more likely to attend others 37 (Portwood et a1., 1981). Two other analyses computed zero-order correlations and standardized regression coefficients between sex and general indexes of union participation which hovered around zero (Huszczo, 1975; Kochan, 1978). In another study, women were actually more active (although not significantly) than men (Sulkin & Pranis, 1967). Research beyond the borders of Canada and the United States par- allels the findings reported above. Abicht (1976) found women to be less active than men in both Belgian and American union locals. In a recent survey of municipal public employees in Great Britain, Nicholson and his coauthors (1981) observed that women scored lower than men, on average, on their index of union participation. Howb ever, the magnitude of the coefficient was quite small ( B - -.10; ;p < .01). Among technical workers in the same country, another study (Harrison, 1979) reported that 37 percent of male union members had attended at least one union meeting in the previous year compared with only 25 percent of female union members. And in a multi-union study in New Zealand (Geare et a1., 1979), women were significantly less active than men in all union events. In particular, 49.7 per- cent of the men, compared with only 18.1 percent of the women voted in a union election; 12.9 percent and 1.7 percent of the men and women, respectively, ran for union office; and 17.6 percent of the men, against 2.7 percent of the women in the sample were elected as a delegate, steward or representative. Together, the research reviewed above suggests that women union members are generally less likely to be involved than their male co- members in the operation of the union local. It also appears that 38 the male-female difference has declined in recent years. However, the literature also indicates that the difference between men and women in union involvement varies with the specific type of activity. Mere specifically, men are considerably overrepresented in admini- strative forms of union participation such as holding union office, but the male-female difference tends to be smaller when general in- dexes of participation are employed. Quite possibly, women are not as underrepresented in union meetings or votes as has been assumed in the past. Our first hypotheses, then, is that women union mem— bers are less involved in the union local and that the magnitude of this difference is largest for administrative participation. Education MOst of the union participation research has found that the most active members of the union local tend to have somewhat higher levels of formal education than the less active members (Hagburg & Blaine, 1967; Huszczo, 1975; Kochan, 1978; Nicholson et a1., 1981; Sulkin & Pranis, 1967; Tannenbaum & Kahn, 1958). Two other studies discovered that employees with higher levels of formal education are more likely to be involved in some forms of union activity but not others (Portwood et a1., 1981; Vall, 1970). For example, Vall (1970) learned that highly educated white-collar workers were overrepresented at all levels of union administration, particularly in the highest posts, but there was no noticeable difference in education between attenders and nonattenders of union meetings. Three other reports documented either negative or marginally positive associations between 39 education and union participation (Abicht, 1976; Anderson, 1979; University of California, 1966). In terms of the union participation model, education is a back- ground characteristic which has been associated by a few writers with the opportunity to participate in certain activities (Cook, 1968; Tooredman, 1981; Vall, 1970). Specifically, educational attainment is apparently considered by many union members as one of several cri- teria used in selecting candidates for union office. 0n the other hand, education is not positively associated with the willingness to participate (Glick et a1., 1977; Gordon et a1., 1980). Our pre- diction, then, is that education is positively associated with admin- istrative participation but probably has no effect on the propensity to either attend union meetings or vote in union decisions. Educa- tion would also partially explain why women are generally less in- volved than their male coworkers in the operation of the union local if women in our sample have lower levels of formal education. This has been mentioned by Cook (1968) and was cited by Tooredman (1981) as one reason why the male respondents in her study were more likely than women to take on leadership roles. Marital Status and Hours of Housework An often-cited explanation for the disparity in union participa- tion between women and men is that women hold two jobs - one at the office or factory and the other at home (Andiappan & Chaison, 1983; Cook, 1968; Glassberg et a1., 1980; Koziara & Pierson, 1980; Wertheimer & Nelson, 1975). For example, in a recent survey of female officers of national unions and labour federations across 40 Canada, the greater responsibility that women have in the home compared with men was identified as the greatest barrier to the involvement of women in labour unions (Andiappan & Chaison, 1983). One method of operationalizing this issue is to consider the number of hours of housework (per week) that the union member is responsible for. Given that home responsibilities "crowd out" the person's opportunity to participate in time consuming union activi- ties (Harrison, 1979), we would associate the hours of housework 'with the opportunity component of our model. More specifically, we predict that the number of hours of housework that the respondent completes each week is negatively associated with participation in the administration of the union and union meeting attendance. Both of these are time-consuming types of involvement which cannot easily be fitted into the schedules of people with heavy housework Obliga- tions. On the other hand, the amount of housework should have little or no effect on union voting behaviour which is not usually a time- consuming event. Housework has a particularly detrimental effect on the union participation of women because they perform much more of the house- hold chores than do men. Recent studies have estimated that women perform between 75 and 85 percent of the housework in families (Berheide et a1., 1976; Vanek, 1980) and that this proportion is only slightly lower when both husband and wife are in the labour force (Meissner et a1., 1975). Thus, to the extent that housework is a barrier to participation in the administration of the union and attending union meetings, this variable is expected to partially explain why women are less involved than men in these activities. 41 Judging from the union participation research findings, marital status is a background variable which associates both with the will- ingness to participate and, for women only, the opportunity to par- ticipate. Tannenbaum and Kahn (1958) observed that married men are more involved in their union local than are unmarried men, but they warned that this relationship did not hold true for women. Mere recently, Harrison (1979) compared union meeting attendance among married and unmarried men and women in a British technical union. The data from her study show that married men have much better at- tendance records than single men, while the difference between mar- ried and single women is very slight. Another investigation, looking only at the participation of women union members in leadership activ- ities, concluded that married women are much less likely than un- married women to hold union office (Abicht, 1976). Other studies which did not look at the data separately for men and women reported either nonsignificant or hmrginally significant correlations between marital status and union participation (Anderson, 1979; Divers, 1981; Nicholson et a1., 1981). One explanation for these results is that married union members have a greater stake in their jobs which indicates strong job-related interest in union participation, but that this motivation is offset for women who tend to receive more household responsibilities from marriage. In terms of specific hypotheses, a positive association is expected for men between marital status and voting participation (because it is the most job-related form of union activity), followed by meeting attendance and administrative participation, respectively. For women, we anticipate that married union members are less active 42 than unmarried union members in administrative participation and meeting attendance because marriage tends to reduce the opportunity that wOmen have available to become involved in these time consuming activities. Nonsignificant coefficients are expected for the anal- ysis of men and women combined. Seniority The length of time that the employee has been a member of the union local has been discussed in the union participation literature both as a factor influencing the willingness to become involved in certain activities and the opportunity to participate in others. Seniority affects the willingness to become involved in the union to the extent that more senior employees have a greater stake in the job (Strauss, 1977; Tannenbaum, 1965). Also, being more familiar with their working environment and having fewer options for inter- organizational mobility (MOwday et a1., 1982), more senior employees have a greater interest in exercising some control over the condi- tions within which they work (Dean, 1954). From the perspective of our union participation model, therefore, seniority is associ- ated with the prepotence of the union member's job-related needs. Thus, we predict that seniority is positively associated with voting in union decisions and, to a lesser degree, attendance at regular union meetings. Seniority is also connected with the opportunity component of the model to the extent that this variable is considered by the mem- bership as a criterion in the selection of union leaders. Two studies have indicated that more senior employees are given more encouragement 43 to run for union office and to hold the more important jobs in the union (Geare et a1., 1979; Wertheimer & Nelson, 1975). This support is further manifested by the proportion of votes given to more senior members in union elections (Sayles & Strauss, 1953). From these ob- servations we would expect seniority to be positively associated with administrative participation. In general, the research published to date supports the above hypotheses (Dean, 1954; Kyllonen, 1951; Nicholson et a1., 1981; Portwood et al., 1981; Strauss & Sayles, 1953; Tannenbaum & Kahn, 1958; University of California, 1966). However, two other studies found only a modest positive relationship (Anderson, 1979; Sulkin & Pranis, 1967) and in a few other investigations the effect was either near zero or negative (Glick et a1., 1977; Gordon et a1., 1980; Huszczo, 1975). The conflicting evidence notwithstanding, senior- ity is hypothesized to positively covary with all forms of union participation: union voting behaviour and meeting attendance be- cause they serve job-related needs and administrative participation because more senior union members have a better opportunity to hold union office. Seniority might also be identified as an explanation why women are generally less active than men in the union local. Generally speaking, women tend to have less seniority than men (Cook, 1968; wertheimer & Nelson, 1975) which possibly causes them to be less willing to participate in certain union events (e.g. union votes) and less able to be involved in other forms of union participation (e.g. holding union office). 44 Employment Status Following the same line of thought as for seniority, having a temporary or permanent employment status with the company can be viewed as a factor influencing both the willingness and opportunity to participate in certain union functions and roles. With respect to the willingness to participate, Seidman et a1. (1958) interviewed several union members who were inactive because they viewed their stay in the company or industry as temporary. Similarly, Gesre and others (1979) found full-time employees to be more involved than their part-time counterparts in all forms of union activity. Cook (1968) has suggested that the sex differential in labour union par- ticipation is partly due to the fact that women (especially with young children) are less likely to become permanent workers "and hence do not identify themselves as regular members either of the workforce or of the union" (pg. 131). Indeed, women in the Canadian labour force are much more likely than men to be part-time (Statistics Canada, 1981). In terms of our model, temporary employment would reduce the 'payoff' of participation to fulfill job-related needs because the employee might not be around to reap the benefits of participation and because the job presumably represents (as Cook implies) a smaller part of his or her life interest. Thus, we would expect that perma- nent employees are more active than temporary employees in all types of union activity. And to the extent that more women than men are temporary employees in our sample, emplOyment status should partially explain the male-female difference in union involvement. 45 Temporary employees are also prevented from participating in some union activities and are more likely to be screened out of others. In the union that we have studied, for example, temporary employees must join the union. They are allowed to participate in some union activities, but the union consitution prevents them from voting in union decisions and holding elected office. We would pre- dict, therefore, to find a significant correlation between employment status and these forms of union involvement. Temporary employment status might also be a handicap to those employees who desire (in spite of the forces against this) to hold appointed union jobs. In the appointment of individuals, permanent employment status might be another implicit criterion in the selection process. Distance from Home to the Union Hall The opportunity to be involved in certain forms of union partici- pation is limited for some employees by the distance they must travel from home to the union hall where many of these activities are held. This is evident in several investigations which found a negative relar tionship between the distance to the union hall and various types of union involvement (Anderson, 1979; Miller & Young, 1955; Kyllonen, 1951; Nelson & Grams, 1978; Purcell, 1954; Rosen & Rosen, 1955; Seidman et a1., 1958; Vall, 1970). For instance, the Rosens (1955) calculated that the average union steward or officer in their multi- union survey lived 25 percent closer to the union hall than the aver- age rank-and-file member. In Kyllonen's (1951) study, only 5 percent of the out-of-town members attended union meetings regularly, com- pared with 44 percent of the local members. And in Vall's (1970) 46 investigation, 70 percent of those who commuted, against 46 percent of the noncommuters, were apathetic about attending local branch meetings. The above-cited literature generally points to union meeting attendance and participation in the administration of the union local to be affected by distance as a situational contigency. 0n the other hand, distance to the union hall should not be a barrier to voting participation which, at least in the union studied, is an occasional event taking place immediately after the work day has ended for most of the union membership. Regular union meetings, in contrast, are held monthly and typically begin at 7:00 p.m., after most people have headed home for the evening. Similarly, meetings involving the union executive, committee members, and union stewards are held in the evening after most people have first travelled home. Thus, we would expect the distance to the union hall to be negatively correlated with involvement in the union administration and meeting attendance, but not with voting in union decisions. This variable will explain the male-female difference in participation only if women union members are more likely to live further away than men from the union hall. Friends in the Union Much of the union participation research has observed that active union members are more socially outgoing than employees who are not as involved in the union local (Spinrad, 1960; Strauss, 1977; Tannenbaum, 1965). Others have noted that integration into the social life of the organization and union local provides an 47 important foundation for union participation (Dean, 1954; Hagburg, 1966; Perline & Lorenz, 1970). Furthermore, two studies have found that union meeting attendance would probably be higher if some of the nonattenders personally knew the other people attending (Geare et a1., 1979; Wertheimer & Nelson, 1975). These results can be interpreted from the perspective of social needs in our union participation model. Employees who have prepo- tent social needs are more willing to be involved in the union local, particularly in attending union meetings and other group functions. Indeed, Nicholson and others (1981) were surprised to discover how significant the need for affiliation is in predicting attendance at union meetings and rallies. One indicator of the importance union members place on social relations and the extent to which they have been integrated into the social life of the union is the proportion of their friends who are in the union. Several researchers have already indicated that union members who are more involved in their local labour organization tend to have more friends in the union and engage in higher levels of co- worker interaction off-the-job (Anderson, 1979; Dean, 1954; Hagburg, 1966; Lipset et a1., 1956; Nelson & Grams, 1978; Nicholson et a1., 1981; Tannenbaum & Kahn, 1958). Our hypothesis is more specific: the proportion of friends in the union will positively covary with the respondent's union meeting attendance record and, to a lesser extent, with voting participation and involvement in the administra- tion of the union. Although the direction of causality might be questioned - meeting attendance might cause union friendship rather than vice versa - the literature cited above suggests that union 48 meeting attendance is a result of friendships developed on the job and that the development of these friendships is a function of pre- potent social needs (see especially, Geare et a1., 1979; Hagburg, 1966; Spinrad, 1960). This variable is not expected to be a factor in the male-female difference in union participation. Value of Unions and Interest in Union Business In our earlier description of the union participation model (Chapter II), we explained that the willingness to participate in certain union events is partly determined by the degree to which ‘the union member is favourably disposed toward labour unions in gen- eral. Specifically, the association between general union attitude (which we operationalize with our "Value of Unions" scale) and union participation increases with the degree to which the union activity is central to the operation of the union. we therefore predict that the strongest relationship between value of unions and union involve- ment will be found for administrative participation, followed by meeting attendance and voting participation, respectively. The co- variance between the individual's attitude toward unionism and var- ious forms of union involvement is not expected to be a factor in the male-female union participation differential because women are no more or less likely than men to hold opinions favourable to union- ism (Gordon et a1., 1980; Huszczo, 1975; Kochan, 1978; McShane, 1983). A less general approach is to consider the union member's stated degree of interest (or lack of interest) in union business. There is much evidence that this attitude (also referred to as union apathy) significantly affects the propensity of the union member to become 49 active in the union. For example, nearly one-half of the men and women in Geare et a1.'s (1979) sample agreed that their lack of in- terest in union affairs was a barrier to their greater participa— tion. Nicholson and his cowriters (1981) suggested that apathy was not the norm among respondents in their sample, but they later pre- sented evidence (pg. 93) that over 60 percent of them did not like to attend meetings because "I'm not interested in the union gener- ally.” While this included the majority of 'reluctant unionists' and 'passive dues-payers', 41 percent of the 'selective activists' and 11 percent of the 'apolitical stalwarts' also admitted their lack of interest in union business. The associations between interest in union business and the three forms of union participation should parallel our predictions with respect to value of unions: the largest effect will be on par- ticipation in the administration of the union while the effect on meeting attendance will be smaller and the effect on voting partici- pation should approach zero. As with our measure of general union attitude, interest in union business is not believed to be a factor in our male-female comparison of union activity. Extrinsic Job Satisfaction ' Job satisfaction should be a major determinant of union partici- 'pation since dissatisfied employees have more prepotent job-related needs. That is, they are more likely than satisfied employees to want to improve their working environment. However, the literature is re- plete with mixed results on the relevance of overall job satisfaction to union participation. To this writer's knowledge, only one study 50 has found a significant negative association (Divers, 1981) while one other reported a nonsignificant negative coefficient (Huszczo, 1975). In two other investigations, the relationship between over- all job satisfaction and union involvement (or willingness to par- ticipate) was close to zero (Anderson, 1979; Glick et a1., 1977). And in several of the earlier studies, overall job satisfaction. actually appeared to be positively associated with union participa- tion (Dean, 1954; Form & Dansereau, 1957; Seidman et a1., 1958; Tannenbaum & Kahn, 1958; Vall, 1970). One possible explanation for these confusing results is that not all facets of job satisfaction should be expected to affect in- volvement in the union. In our model of union participation, job- related needs motivate union involvement only to the extent that union participation is instrumental to the fulfillment of these needs. But for the majority of union members, labour unions are not intended to improve all aspects of the job. Rather, they are designed to protect the membership against a deterioration of the job context (Tannenbaum, 1965) and do not (some say should not) act as a catalyst for the creation of more challenging work (Kochan et a1., 1974). Consistent with this view, one researcher recently learned that union member perceptions of local union effectiveness are strongly correlated with extrinsic job satisfaction but are un- correlated with intrinsic job satisfaction (Kowalczyk, 1982). Sims ilarly, dissatisfaction with the job context appears to be a better predictor than intrinsic job dissatisfaction of pro-union voting intentions and behaviour in union certification elections (Allen & Keaveny, 1981; Bigoness, 1978; Getman et a1., 1976; Schriesheim, 1978). 51 There is, in fact, some evidence to support the idea that extrinsic job satisfaction is a better predictor of union partici- pation than is intrinsic job satisfaction. In his review of the earlier research, Spinrad (1960: 242) concluded that union activists are more protesting "about particular aspects of the job situation, not about the job itself." This is consistent with Rose's (1952) observation that the union leaders in his study tended to be more critical than the ran-and-file of specific aspects of the job sit- uation. More recently, Nicholson and others (1981) identified job security concerns, job stress, departmental problems, and intragroup conflict as significant correlates with their index of union involve- ment, although satisfaction with pay and promotion had no effect. Finally, Gordon and his coauthors (1980) discovered that willing- ness to work for the union was significantly and negatively asso- ciated with extrinsic job satisfaction but did not covary with in- trinsic job satisfaction. Another possible reason for the confusing results is that not all forms of union involvement are equally instrumental in the ful- fillment of the individual's job-related needs. Rather, as we ex- plained in Chapter II, union members who are dissatisfied with.their job context are more likely to attend union votes than attend regu- lar meetings or become involved in the local union administration. This does not necessarily mean that extrinsic job satisfaction will be uncorrelated with these latter forms of involvement. Rather ex- trinsic job satisfaction should correlate most strongly with voting participation because this activity is most instrumental to the ful- fillment of job-related needs. 52 Thus, we hypothesize that extrinsic job satisfaction will be inversely correlated with union participation and the strength of this association is greatest for voting participation, followed by meeting attendance and union administration involvement, respectively. Extrinsic job satisfaction is included in our study to represent the basic union participation model and is not expected to be a factor in our understanding of why women tend to participate less than men in local union events. Job Involvement Lodahl and Kejner (1965: 24) define job involvement as "the degree to which a person is identified psychologically with his work, or the importance of work in his total self-image." One might rea- sonably associate people with high job involvement with those who have a greater stake in their jobs. Assuming this is true, union mem- bers with high job involvement are likely to be more involved than low job involvement members in their union local. 0n the other hand, Spinrad (1960: 242) observed from his review of the early literature that some union activities, such as holding union office, can replace the challenge and responsibility lacking in the job. In the other words, he predicted that employees who are less involved in their jobs might be more active, rather than less active, in the union local. Borrowing Wilensky's (1960) terminology on the "spillover" and "compensatory" hypotheses for the work-nonwork nexus, Nicholson et a1. (1981: 37) identified the two contrasting hypotheses outlined above: Applied to union involvement the former hypothesis predicts that high levels of work involvement and identification will stimulate union activism. The latter says that the unions provide alternative channels for the frustrated energies of workers. 53 The obvious question, of course, is which hypothesis stands up to the empirical test? Only Nicholson et a1. (1981) directly assessed the relationship between job involvement and union participation, although Huszczo (1975) correlated participation with a more ill-conceived measure called "job satisfaction-involvement." In both studies the effect was positive and significant, but not very large (i.e. B 8 .07 and .14, respectively). Considering the above-mentioned hypotheses in light of our multi- dimensional conceptualization of union participation, these predic- tions might not actually be in cOnflict with each other. On one hand, not all forms of union participation will compensate for the lack of challenge and responsibility in the job. Based upon our discussion in Chapter II of growth needs and union participation, we predict that administrative participation offers the best activities for the union member to feel a sense of 'involvement' because of the challenge and responsibility available in these functions. Thus, employees with low job involvement will be more likely to be active in the admini- stration of the union local. 0n the other hand, those with high job involvement will have a greater stake in the job which will primarily affect their involvement in union votes and, to a lesser extent, union meeting attendance. In short, job involvement is hypothesized to be negatively associated with administrative participation and positively associated with voting participation and, perhaps, union meeting at- tendance. Finally, job involvement is not considered to be a major factor in the male-female difference in union participation. Although women might be somewhat less psychologically attached to the labour market 54 as some have suggested (Cook, 1968; Harrison, 1979), the specific measure of job involvement does not appear to vary by sex (Rabinowitz & Hall, 1977). Intent to Quit Employees who do not intend to stay with the company much longer would proably have less interest than others in what happens to work- ing conditions associated with their current job. Phrased in terms of the union participation model, we might say that their job-related needs (at least, regarding their current job) are lower than for em- ployees who plan to stay on. Consequently, participation in the union would not yield a personal 'payoff' (Strauss, 1977), particularly with respect to voting in union decisions or attending union meetings. Thus, we expect the respondent's intent to quit to be inversely cor- related with voting participation and union meeting attendance. . Neither a positive nor negative association is predicted between ad- ministrative participation and the intent to quit. Historically, women have had higher rates of turnover than men (Economic Council of Canada, 1976; Ontario, 1980). To the extent that this is true for our sample, we would expect this variable to partially explain why women are less involved than men in some forms of union participation. Summagz Table 4 summarizes the hypotheses outlined in this chapter. Al- though these predictions are rather complex, they are based on the straightforward concepts incorporated into the model of union partic— ipation. That is, union-related behaviour is a function of the will- ingness to engage in that behaviour and the opportunity to do so. The 55 Table 4 Hypothesized Associations Among the Independent Variables and Three Forms of Union Participation Variable Administrative Meeting Voting Participation Attendance Participation Individual Characteristics Sex (Male - l) -- - - Education ++ 0 0 Marital Status Men 0 + ++ WOmen -- - 0 Hours of Housework -- - 0 Union/Job Characteristics Seniority H- -H- -H- Employment Status ++ ++ ++ Distance to Union Hall - -- 0 Friends in the Union + ++ + UnionlJob Attitudes Value of Unions ++ + 0 Interest in Union Business ++ + o Extrinsic Job Satisfaction 0 - -- Job Involvement - + ++ Intent to Quit O - -- 56 willingness to participate in a particular type of union activity is determined by the prepotence of certain personal needs and the indi- vidual's attitude towards unionism. Education has been identified as a situational contingency which affects the union member's opportunity to participate in many union administration functions. However, this variable is not expected to affect meeting attendance or voting administration. Education is believed to partially explain why women are less involved than men in the operation of the union local. For the male sample, marital status is hypothesized to be pos- itively associated with voting participation and, to a lesser extent, meeting attendance. For women, marital status should be inversely associated with union administration and meeting attendance. The reason for these predictions is that married employees have a much greater stake in their jobs (i.e. higher job-related needs), thereby motivating attendance at union votes and general union meetings. But for women marriage also reduces the time available to participate in time-consuming union events such as union administration and union meetings. The number of hours of housework per week the individual is re- sponsible for restricts the opportunity to participate in time-con- suming union activities. Thus, negative correlations are predicted between this variable and the two time-consuming forms of union par- ticipation. However, as is shown in Table 4, hours of housework will probably not affect voting participation. The difference in union participation between men and women is expected to be partly explained by the fact that women are responsible for most of the family housework. 57 We anticipate that seniority covaries with all three forms of union participation. The more years that employees have with the company and union, the greater is the stake in their jobs, which translates into a higher propensity to participate in voting deci- sions and attend union meetings. Seniority has also been identified as a variable affecting the opportunity to become involved in union administration, resulting in an expected positive coefficient for this type of involvement. The same phenomenon is predicted for emr ployment status except that opportunity as well as willingness deters temporary employees from voting in union decisions. Both seniority and employment status are relevant to our comparison of male-female union participation. The distance that the union member must travel between home and the union hall is hypothesized to be a barrier to participation in union administration and regular union meetings. As illustrated in Table 4, distance to the union hall is not expected to be a deter- minant of voting participation because these events take place ime mediately after work. The proportion of friends in the union is predicted to positively covary with the respondentis attendance at union meetings and, to a lesser extent, the other two forms of union participation. This is because union friendships indicate that the union member is socially integrated into the company and local labour organization and, there- fore, is more willing to engage in union activities which fulfill this need for interpersonal contact. Table 4 also indicates that a strongly positive correlation is hypothesized between general union attitude and administrative 58 participation, with a smaller coefficient for meeting attendance, and a zero relationship for voting participation. The same predictions are made between interest in union business and the three forms of union involvement. The rationale behind these anticipated findings is the proposition discussed in Chapter II that the importance of having attitudes favourable to unionism increases with the degree to which the union activity is central to the operation of the union local. Employees who are satisfied with their work environment are pre- dicted to be much less likely to vote in union decisions and somewhat less likely to attend union meetings. Neither a positive nor nega- tive association is expected between extrinsic job satisfaction and participation in the administration of the union local. Job involvement - the degree to which the employee identifies psychologically with his or her work -- is hypothesized to be neg- atively correlated with administrative participation based upon the growth need proposition that this form of union activity provides an alternate channel for the frustrated energies of workers. In con- trast, strongly and moderately positive associations are expected for voting participation and union meeting attendance, respectively, because employees who are highly involved in their jobs have a greater stake than others in maintaining the quality of their work environment. Lastly, union members who do not intend to stay with the company much longer have a lower stake in the job. Consequently, negative as- sociations between intent to quit and the two forms of participation instrumental to job-related interests (i.e. voting participation and 59 union meeting attendance) are hypothesized. Intent to quit is expected to partly explain why women are less involved than men in the operation of the union local. CHAPTER IV METHODOLOGY Introduction In this chapter we describe the procedures used to select the union local, operationalize the variables discussed in Chapter III, and analyse the data collected. Special attention has been given to the selection of the union local in order to minimize the problem of generalization of these results. A large part of the chapter is also taken up with the methods used to measure administrative participa- tion, meeting attendance, and voting participation as well as two other custom-made instruments: the Value of Unions scale and Job Involvement scale. Not only is the development of these measures described, but evidence of reliability and construct validity is presented. Selection of the Union Local Each year, under the Corporations and Labour Unions Return Act (CALURA), the Canadian Government, through Statistics Canada, col- lects data from all union locals in Canada whether or not they are affiliated with a national labour organization or the Canadian Labour Congress. A copy of the nonrestricted portion of the CALURA form for each union local is made available to the public at the offices of Labour Canada in Hull, Quebec. The information on the 60 61 form includes the name of the local, its address, female and male membership, names and titles of its officers, number of collective agreements it administers, and the name, industry, and address of the employer for each collective agreement. In September 1981, the author reviewed the latest CALURA files available at the Labour Canada offices for the purpose of selecting a union local with which to conduct this study. It was decided to conduct research on just one union local for two reasons. First, previous investigations have determined that the organizational characteristics of the union local do not affect the level of union participation nearly as much as the characteristics and attitudes of the individual union members (Anderson, 1978; Tooredman, 1981). Even if organizational factors were important, financial constraints would have prohibited the collection and analysis of data from a sufficiently large sample of union locals. Consequently, this investigation is based upon information from members of one union local. Several criteria were employed in the selection of the union local from the CALURA files. First, financial constraints limited our search to union locals with between 500 to 700 members which are located in southwestern Ontario (i.e. near Michigan State Uni- versity). Larger union locals were not considered because pref- erence was given to the accuracy and representativeness of census methods rather than drawing the data from a random sample of members. Another factor considered was the industry that the union serves. we restricted our search to union locals in the public, quasi-public, and service sectors of the economy because: 1) nearly 400 percent 62 of all union members in Canada are employed in these sectors, 2) the majority (over 65 percent) of all organized women are found here, 3) many of these union locals represent employees with a broad range of relevant characteristics (e.g. education, marital status, senior- ity, etc.), and 4) the union participation literature is generally underrepresented by studies in these sectors (Swanson, 1981). A final criterion used in the selection of the union local is the percentage of women in the union. Women comprised over 30 per- cent of the labour movement in Canada in 1980 (Statistics Canada, 1982). But since we are generalizing our findings to unions with women members, the exclusion of members from the over 4,000 all- male union locals would increase this percentage somewhat. Using an educated guess, we searched for union locals with between 40 to 60 percent female membership. Based upon these criteria, several union locals were selected from the CALURA files and the Optimal local was approached in October 1981. Following a meeting with the executive board, permission was given to conduct the investigation. Further meetings were held with a task force consisting of two executive board members and with the union stewards. The project was outlined to the membership in at- tendance at the monthly meeting in November 1981 and mention of the survey was made to others through the union local's newsletter. The study was also explained to the top nonelected official in the cor- porate administration and permission was received to use the inter- departmental mailing service. 63 Characteristics of the Union Local The union local from which the data were gathered represents about 600 public sector employees and is located in southwestern Ontario. For purposes of anonymity, the exact location and type of public sector union cannot be disclosed. Although the 1979 CALURA report for the union local indicated that 55 percent of the member- ship was female, this had risen to over 60 percent by the time the study was conducted. As expected, education, occupation, and seniority character- istics were widely dispersed among the membership. Although over one-third of the members have less than a Grade 12 education, over 10 percent have completed graduate work. Occupations vary both in the degree of skill required and in the type of work. FOr example, the union membership includes caretakers, financial analysts, day care teachers, and building inspectors, all of whom are required to join the union within three months of their first day of employment (i.e. union shop). In all, the union local represents employees in over 50 job categories. With respect to seniority, nearly one-third of the employees have been with the organization for 10 years or longer, but nearly a majority have less than five years of service. Several historical and institutional characteristics are also worth noting about the union local. The local was first certified in 1952, establishing it as one of the Oldest in the national union. Indeed, it actually predates the establishment of the national or- ganization which was formed during the early 1960's. Initially, the local was chartered directly with the Canadian Labour Congress (then the Canadian Congress of Labour). 64 Only two strikes have been staged by the local during its 30-year history. The first took place during the 1950's soon after the local was certified. The second job action occurred just two years prior to the beginning of this research and some of the residual labour- management animosity was still evident. All members have the right to strike with the exception of approximately 60 workers in a retire- ment home operated by the organization. Union local administration is governed by both the national union constitution and local union bylaws (which must be approved by the national president). For instance, the national constitution stipulates that the election of the local executive board must take place at least once every three years and at most every year. The local union bylaw has set such elections for every two years. The national union constitution describes the mandatory executive board positions: president, vice-president, recording secretary, secretary- treasurer, and three trustees. However, the local union bylaws es- tablish three other executive board positions: conductor, warden, and chief steward. According to the national constitution, the ex- ecutive board must meet at least once each month. The 15 union stewards (one or two from each department, depending upon the size) are elected annually by the members of each department and meet every month except during the summer. Other positions -— such as newsletter editor, regional labour council representatives, and convention representatives - are filled through appointment by the executive board. Members are also appointed to the union local's seven committees (i.e. negotiation and grievance, election and bylaws, 65 social, welfare, public relations, health and safety, and affirmative action). Most people on these committees are past and present execu- tive board members. At the time of the present study, the economy was in a major re- cession. The local economy is largely dependent on the automobile industry and was particularly hard hit with unemployment rates ap- proaching 15 percent. Layoffs were commonplace in the private sector and there were a few instances of permanent plant closings. Only a few layoffs of temporary staff took place within the organization being studied, but there was concern by the executive board through- out the investigation that management would either be requesting major concessions in wages or would impose major layoffs throughout the organization's workforce. Pretest Study In late November 1981 a predecessor to the questionnaire used in the present investigation was distributed for pretesting on a sample of 158 members of the union local. A total of 131 forms was completed and returned soon after. Tests Of reliability and nomological valid- ity (Cronbach & Meehl, 1955; Peter, 1981) were conducted on the mess- ures deveIOped for the research and the overall feasibility of the questionnaire as an instrument for data collection was assessed. The analysis was completed in January 1982 and changes were subsequently made to improve several measures. Full details of the pretest anal- ysis are available in McShane (1982). 66 Sample for the Pretest Study, Although the union local included over 600 members, only 482 were considered within the relevant population. Since union partic- ipation requires a time dimension (up to one year for the dependent variables we are employing), union members with less than one year of service with the company were excluded. In addition, we restricted our analysis to full-time (including temporary) employees, thereby ex- cluding part-time employees, mainly students who worked for the come pany. This decision was made because few union locals include part- time workers. Also, most of the part-time workers had been employed for less than one year. Finally, the size of the relevant pOpulation was reduced by turnover during the period of data collection. During the late spring of 1982, a lZ-page questionnaire (see Appendix A) was delivered through interdepartmental mail to each of the 482 union members in the designated population. Because dem- ographic and union participation data were also being gathered from company and union records, a confidential rather than anonymous pro- cedure was used. Each form was numbered and only this writer had access to the corresponding list. The respondent was clearly in- formed in the cover letter that the individual information was not anonymous but would be seen only by the researcher. A second inter- departmental mailing took place four weeks later and a final mailing through the regular postal service occurred approximately four weeks after that. Questionnaires with over 75 percent completion were returned by 308 members, representing 64 percent of the designated population. However, 11 questionnaires were discarded because the serial numbers 6] were removed. Missing data specifically required for the present study further reduced the number of usable questionnaires to 275. Table 5 compares five demographic and three union—related char- acteristics of the sample with the population. There is a very close correspondence between the sample and the population with respect to sex, marital status, the percent filing grievances in the past 18 months, and the percent making or seconding motions at meetings in the past 12 months. Employees in the sample have less seniority than those in the population and this is further indicated by the fact that the sample is slightly younger and receives a slightly lower salary than is the case for the pOpulation. The number of meetings attended by the respondents is higher than the pOpulation proportion, but the difference, again, is very small. In fact, we are rather surprised that the differences in all three union-related behaviours are not larger. Based upon our understanding of previous union research, nonrespondents tend to be much less involved than respondents in the local labour organization. Overall, the compar- ative data suggest that our sample is quite representative of the pOpulation. The Questionnaire The questionnaire used to collect the data for the present study consists of a 12-page booklet (see Appendix A) which is the culmina- tion of several earlier drafts including the pretest questionnaire. The format is based upon the procedures recommended by Dillman (1978) and was reported by the union executive board task force as well as 68 Table 5 Comparison of Specified Characteristics for the Sample and Population* of the Union Local Sample Population Variable (N . 275) (N - 482) Sex Male 352 372 Female 65 63 Marital Status Married 64! 671 Not Married 36 33 Bi-weekly Salary Less than $600 ' 241 202 $600 - $799 53 55 $800 - $999 19 20 $1000 or more 4 5 Age Under 30 392 332 30 - 39 30 29 4O - 49 14 18 50 - 65 17 20 Seniority Less than 5 years 56: 45: 5 - 9 years 21 24 10 - 14 years l3 17 15 - 19 years 5 8 20 or more years 5 6 Number of Meetings Attended** None 612 651 One 19 20 Two or more 20 15 Filed Grievance (within 18 months) Yes 122 122 No 88 88 Made or Seconded Motions at Meetings (in past year) Yes 82 62 No 92 94 *Members with less than 12 months seniority omitted; members with 24 months re- maining before retirement omitted. **Due to problems in reading handwriting in the meeting registers, the percentage of all members in our population attending at least one meeting is closer to 452. However, this error in data collection is probably uncorrelated with whether or not the person completed a questionnaire. 69 some respondents (in the comments section) to be very readable and understandable. Not all of the information gathered by the instrument is in- cluded in the present investigation, but the entire booklet is re- produced in Appendix A for convenience. The specific items used in this study are stated in the following sections on the measure- ment of the independent and dependent variables. Measurement of the Dependent Variables In congruence with our multidimensional perspective of union participation, our analysis in this section focuses on the three types of union involvement identified in Chapter II and discussed in the context of the independent variables in Chapter III. In this section we describe how these three dependent variables -- admin- istrative participation, union meeting attendance, and voting par- ticipation - have been Operationalized. Administrative Participation. Involvement in the direct admin- istration of the union local is measured by an index of eight union activities. Five of these activities —- holding elected union office, being a committee member, attending a union convention, running for union office, and holding any other appointed office (e.g. editor of the newsletter) -— are clearly within our understanding of participa- ‘tion in the administrative functions of the union local and, in fact, clustered together in the pretest analysis. However, the pretest results also pointed out that active involvement in regular union meetings strongly associates with other administrative activities. This is consistent with what Tannenbaum and Kahn (1958) have argued, 7O namely, that attendance at union meetings does not imply active participation in the business of the union meeting. Consequently, we have included three union meeting activities within the measure- ment of administrative participation: 1) presenting/seconding mo- tions at meetings, 2) asking questions/stating views at meetings, and 3) providing information at meetings. With the exception of presenting/seconding motions at meetings, the self—report data collected from the questionnaires were used for the administrative participation index. This decision was based on our observation (confirmed by some of the union executive members) that some of the data from the union files were not always reliable. For example, delegates to local or secondary conventions were not always reported in the union meeting minutes and, of course, it was not possible to record in the minutes every person at the regular meetings who asked a question or made a comment. 0n the other hand, the names of persons who presented and seconded motions at the union meeting were carefully recorded in the minutes. Consequently, only the information on presenting/seconding motions was taken from the union records. 'The index was developed by dichotomizing each of the eight items and adding the values (similar to the procedure used by Nicholson et a1., 1981). As in previous research using more global measures (e.g. Huszczo, 1975; Nicholson et a1., 1981), the resulting distri- bution is skewed. Thus, we caution the reader that the multiple re- gressions results will probably be somewhat biased downward (i.e. attenuated) by this deviation from normality. Internal consistency 71 reliability for this index of administrative participation is quite high ( o 8 .92). Union MeetingTAttendance. Ten general membership meetings are held annually in the union local under study - one meeting each month except in July and August. Information for each respondent on the number of union meetings attended during the previous 12 months was gathered both from the questionnaire and more directly from.the list of names on the meeting register. Strauss (1977) has recommended that researchers of union participation place less reliance on self-reported attendance at union meetings. Although we were also concerned that many respondents would report an exag- gerated number of meetings attended - Dean (1958) calls these people "dissemblers" -— the correlation between the self-report and officially recorded data was quite high ( r - .79), indicating that there was not much exaggeration. 0n the contrary, the author has less faith in the official data because some of the signatures were not legible (and therefore not coded) and because members who arrived late occasionally neglected to sign their names. Conse- quently, we decided to measure union meeting attendance using the self-report data from the questionnaire. VbtinggParticipation. The union members were involved in two special voting decisions during the year preceding the study: the strike vote and contract ratification vote. (Union elections and other votes were scattered throughout the regular general member- ship meetings). Well over 60 percent of the population designated for this study turned out for each of these events, compared with an average of about 10 percent for the general membership meetings. 72 The voting participation index was developed by adding the two dichotomized variables indicating attendance or nonattendance at each of the two votes. On_average, 1.22 of the two voting events was attended. A high internal consistency reliability was computed for this index ( a - .89). To test our proposition that labour union participation is a multidimensional construct, we intercorrelated administrative par- ticipation, union meeting attendance, and voting participation. If union participation is a multidimensional construct, then these three sets of activities should have only moderate associations with each other. This is exactly what we find. For voting participation, the zero-order correlations with administrative participation and meeting attendance are only .21 and .22, respectively. The correla- tion between administrative participation and meeting attendance is somewhat larger ( r - .44), but this is still much lower than the average intercorrelation of the items representing administrative participation ( f - .62). Overall, these results support the argu- ment that union participation is a multidimensional construct and that the three types of union behaviour delineated for the present study are sufficiently independent to permit separate analysis. Independent Variables The 13 variables described in the previous chapter are included in each of the regression equations. (The sex variable will, of course, be removed when we run separate regressions for men and women, leaving 12 predictors for these analyses.) Below, we de- scribe how each of these variables has been Operationalized. 73 .yglue of Unions and Job Involvement. The instruments measuring these two variables were develOped for this study and, therefore, are discussed first. The Value of Unions scale is a central concept in our model which indicates the respondent's general attitude toward unionism. In the pretest questionnaire, an adaptation of the "Union- ism in General" subscale from the IRC Union Attitude Questionnaire (Uphoff & Dunnette, 1956) was tested. Changes to the Uphoff and Dunnette subscale were required because some of the original items were not relevant to labour unions in Ontario, some lacked content validity (from discussions with union representatives), several re- quired changes to create more gender-neutral statements, and the meaning of others were reversed to minimize response set bias. In addition, two new items were added. The item and factor analyses from the pretest data revealed that the items were not as homogeneous as had been previously assumed (McShane, 1982). Rather, several statements in the original scale clearly were more relevant to the concept of attitudinal union mil- itancy (see: Dolan, 1979; Hellriegel et a1., 1970; Shirom, 1977) than general union attitude (e.g. "We need more laws to limit the powers of labour unions."). Several items were subsequently de- leted and others were added to the Value of Unions instrument based upon the pretest results and discussions with union representatives. These changes resulted in an eight-item scale using a seven-point Likert-type response format with two reverse-scored items. The item analysis of the revised Value of Unions instrument using the final sample data is displayed in Table 6. The eight items measure a very homogeneous construct ( a - .85). A high score on the Value 74 Table 6 Item Analysis and Reliability of Value of Unions Scale Corrected Question ItemrTotal Number Item Correlation R2 12a Unions are a positive force in Canada .53 .34 12d Workers who don't organize will be worse 62 40 off in the long run ' ° 13a Unions can be an effective means of pro- 57 37 tecting employees ' ' 13d I am proud to be a member of the labour .72 .58 movement in Canada 131 The growth of unionism has made our dem— 66 47 ocracy stronger ' ' 13m Employees have better wages and working 49 28 conditions when they are unionized ' ' *13o If I had to choose, I probably would not 69 56 be a member of any labour union ° ' *l3r My wages and benefits would probably be just as good if there were no unions in .58 .40 this country Average Inter-Item Correlation - .44 Coefficient Alpha (Unstandardized) - .85 N of Cases - 27S * Scoring has been reversed for negative items. 75 of Unions scale indicates that the respondent has a favourable opinion of unions in general. Job involvement refers to the extent to which the individual psychologically identifies with his or her job (Rabinowitz & Hall, 1977) and is often measured with a Likert-type response format using items developed by Lodahl and Kejner (1965). The researcher is given the Option of a 20-item scale or the shorter 6-item version, but most researchers have tended to pick any combination of statements from the 20-item scale. Furthermore, there is some concern about the or- iginal conceptualization of job involvement in the Lodahl and Kejner study (Siegal, 1971) and about the factorial unity of the measure (Saleh & Hosek, 1976), particularly when compared with items tapping intrinsic job satisfaction (Lawler & Hall, 1970). Based upon the results of several previous studies which used various combinations of statements in their job involvement scale, we selected seven itemstx>operationalize the construct (see Question 19 in Appendix A). The item analysis, reproduced in Table 7, indi- cates that the job involvement scale has a high level of internal consistency ( a - .79). To test the construct validity of the Value of Unions and job involvement scales, a factor analysis was conducted on the statements from both instruments together. As Table 8 reveals, all items within each construct load onto the same factor, thereby providing some evi- dence of convergent validity. Discriminant validity is also evident by the low factor loadings of the items on the other factor. Together these observations provide initial support for the construct validity of both the Value of Unions and job involvement instruments. 76 Table 7 Item Analysis and Reliability of Job Involvement Scale Corrected Question Item-Total Number Item Correlation R2 19a My job is one of the major satisfactions .61 .49 in my life *l9b There are many things that are more imp 51 36 portant to me than my job ' ' *l9c My job is only a small part of who I am .51 .37 19d Overall, I am very much involved in my .51 .44 job *l9e I would not want my job to be a central 52 32 part of my life ' ’ 19f I live, eat, and breathe my job .49 .33 19g The most important things that happen to 46 33 me involve my job Average Inter-Item Correlation - .35 Coefficient Alpha (Unstandardized) - .79 N of Cases - 275 * Scoring has been reversed for negative items. 77 Table 8 Factor Analysis of Value of Unions and Job Involvement Question 0 Number Item Factor 1 Factor 2° Communality ‘zglue of Unions 12a Unions are a positive force in Canada *.63 .02 .40 12d Workers who don't organize will be worse off in the long run *'63 '04 ~36 13a Unions can be an effective means of * 58 05 33 protecting aflploygc' . , . 13d I am proud to be a member of the la- * 80 01 62 hour movement in Canada 13k Too many grievances are filed by * 7O 00 45 unions these days ° . . 13m Employees have better wages and work- * 5‘ 14 31 ing conditions when they are unionized ° ' . 130 If I had to choose, I probably would not be a member of any labour union l3r My wages and benefits would probably be just as good if there were no *.61 -.03 .41 unions in this country *.76 .06 .60 Job Involvement 19a My job is one of the major satisfac- * tions in my life .07 .65 .45 19b There are many things that are more 02 * 56 36 important to me than my job ' ° ' 19c My job is only a small part of who _ 07 * 57 40 I u 0 O 0 19d Overall, I am very much involved in 04 * 61 42 w Job 0 . O 0 19a I would not want my job to be a cen- - 06 * 57 30 tral part of my life ' ° ' 19f I live, eat, and breathe my job .13 *.54 .33 19g The most important things that happen 12 * S4 34 to me involve my job ' ° ° Eigenvalue ‘ 4.16 2.93 Percentage of Variance 27.7 19.6 oWith Varimax rotation *Indicates factor that variable loads highest on 78 Extrinsic Job Satisfaction. Satisfaction with the job context was measured by the combination of three subscales from the Job De— scriptive Index (JDI): pay, promotions, supervision. The two other subscales from the JDI instrument - work and coworkers —— were in- cluded in the questionnaire but were excluded from the present anal- ysis because they represent aspects of the job over which labour unions are usually not perceived as being instrumental (see our dis- cussion in Chapter III). The JDI was selected because of its care- ful development and validation (Smith et a1., 1969). Internal con- sistency reliability of the combined pay-promotion-supervision subscales is .87. Reliability estimates for the three subscales separately ( o - .82, .86, and .89, respectively) are quite simr ilar to those reported in earlier research (for a summary, see: Cook et a1., 1981). Other union/Job Attitudes. As a measure of Intent to Quit, respondents were asked to estimate the probability that they would quit their job for whatever reason with the organization within the next two years. Answers ranged from "1002; I am absolutely certain that I will be quitting" to "0%; I am absolutely certain that I will not be quitting" (Question 20). The format of this question closely parallels that used by Anderson and Milkovich (1980). Interest in Union Business is measured by a single item: "I'm usually not interested in what goes on" (Question 5d). This is quite similar to an item used by Nicholson and his colleagues (1981: 93). Although the statement refers specifically to reasons for not attend- ing general union meetings, it is assumed that the answers generalize 79 to union business in any context. Coded values for this statement range from O to 3 with a high score indicating that the respondent has a strong interest in union business. Union/Job Characteristics. Seniority information was taken fromeuestion l in the questionnaire and has been coded as the nuns ber of months. The self-report data were quite consistent with company records except in a few cases where the respondent has left the organization and then returned. Since the official rec- ords would not have included seniority from earlier employment with the organization, the self-report data were considered to be more accurate for our purposes. The distance to the union hall was measured by Question 4 in the questionnaire. Initial analysis of the scatter diagrams be- tween distance and union meeting attendance indicated that no trend appears among respondents within a short distance of the union hall. It was therefore decided to dichotomize the variable so that respon- dents living less than 10 kilometres (about 6.5 miles) away from the union hall were coded '0' and those living 10 kilometres or further away were coded '1'. In Question 10, respondents were asked how many of their friends are members of a labour union. Subsequent conversations with people at two union meetings revealed that almost all of their union friends were members of the union local under study. Therefore, this item was interpreted as an indication of the number of friends that the individual had in this union. Possible answers ranged from "almost all of them" to "none of them." 80 Information on the respondent's employment status was collected directly from company records. Only full-time employees are included in this study. However, approximatley one-fourth of the respondents are temporary while the rest are permanent. This distinction is im- portant in terms of union privileges (e.g. right to vote and hold elected office) and employment conditions (including having a lime ited versus indefinite contract). Individual Characteristics. Lastly, our statistical analysis includes four nonwork-related personal characteristics and exper- iences: sex, education, marital status, and hours of housework. Information on the respondent's sex (Female - 0; Male - l) and mar- ital status (Not Married - 0; Married - l) was collected from Ques- tions 24 and 29, respectively. Both variables correlate above .80 with company records. -Level of education is a six-point ordinal variable (Question 25) ranging from "Grade eight or less" to "Grad- uate courses at university." However, it is assumed that this dis- tribution adequately parallels an interval scale for our purposes. For the fourth individual characteristic, the respondent was asked to estimate the number of hours per week that he/she usually spends doing housework (Question 28). In order to reduce the potential unreliability of this variable, the respondent was first asked to indicate the percentage of several household chores that he/she typically is responsible for (Question 27). It is hoped that this helped to structure the respondent's estimation of the actual amount of time spent doing housework each week. However, the list of house- hold tasks is not perfectly representative of the tasks considered 81 in major studies on housework (Berheide et a1., 1976). Therefore, to the extent that the list in Question 27 structures the respon- dent's perspective of the housework construct, the answer to Ques- tion 28 will not perfectly represent the number of hours per week that the respondent performs housework. Method of Analysis The principal method of analysis employed in this study is standardized multiple regression. This procedure has been used for several reasons. First, as other researchers have acknowledged (Nicholson et a1., 1981; Strauss, 1977), it is time to progress from bivariate analyses to multivariate approaches with respect to the relationship between union participation and its predictors. In this way, variables of key importance can more readily be iden- tified. A second reason is that the variables included in the present analysis are sufficiently at the interval scale level to permit multiple regression. The assumption of normality is not satisfied for all variables, but this violation is relatively minor since the procedure is highly robust in this regard. At worst, the esti- mated coefficients are somewhat attenuated. Finally, standardized coefficients have been computed because unstandardized values would not be intrinsically meaningful (espec- ially for attitude measures). Moreover, we want this analysis to be comparable with other multivariate research on union participa- tion which, without exception, has reported standardized coefficients 82 (Anderson, 1979; Huszczo, 1975; Kochan, 1978; Nicholson et a1., 1981; Portwood et a1., 1981). The analysis reported in the next chapter is in two stages. First, multiple regression equations for each of the three types' of union participation are explored using the full sample. Most of the hypotheses are relevant to this part of the data analysis. Then, separate equations for the male and female subsamples are compared to identify structural differences. With the exception of the marital status variable, these latter analyses will be largely exploratory. Summary This study is based upon data gathered through a pretested questionnaire as well as company and union records on members of a public sector union local situated in southwestern Ontario. The union local was selected from CALURA files at Labour Canada on the basis of four criteria: 1) the number of members in the union local, 2) the industry that the members of the union local are as- sociated with, 3) the percentage of women in the union level, and 4) the location of the union local. Permission was granted from the union local executive board as well as the t0p administrator (nonelected) of the corporation. A 12-page questionnaire (see Appendix A) was sent to all 482 meme bers designated as the pOpulation for this study. After three mailings, 308 questionnaires were returned. Missing data specif- ically required for the present analysis reduced the usuable number 83 to 275. A comparison of the sample and pOpulation with respect to several variables indicated very little sampling bias. In this chapter we also described the Operationalization of the three types of union participation and the 13 predictors. This included the presentation of reliability and validity information for the multi-item instruments developed specifically for this study. Lastly, the chapter briefly explained the rationale behind the adop- tion of standardized multiple regression procedures for the present analysis. 83 to 275. A comparison of the sample and pOpulation with respect to several variables indicated very little sampling bias. In this chapter we also described the Operationalization of the three types of union participation and the 13 predictors. This included the presentation of reliability and validity information for the multi—item instruments developed specifically for this study. Lastly, the chapter briefly explained the rationale behind the adop- tion of standardized multiple regression procedures for the present analysis. CHAPTER V ANALYSIS OF DATA Introduction This chapter presents the results of the data analysis on the 275 respondents from the public sector union local selected for in- vestigation. The discussion begins with a comparative analysis of the extent to which men and women in the sample participate in the three types of union activity delineated earlier in this manuscript. This is followed by a survey of the multiple regression results using the full sample and their implications for the hypotheses described in Chapter III. In the latter half of this chapter we compare the differences in means between men and women for each of the 12 in- dependent variables and conduct a largely exploratory analysis of the separate multiple regression equations for the male and female subsamples. Analysis of the Union Participation Model The discussion in this section focuses on the correlations among the independent and dependent variables using the full sample. In particular, the zero-order correlations and standardized multiple re- gression coefficients (beta weights) are presented and compared with the hypotheses pertaining to the union participation model. we begin, however, by comparing the male and female respondents in terms of 84 85 their average level of participation in union administration, attendance at general membership meetings, and participation in union.votes. As we see in Table 9, the only statistically significant dif- ference between men and women with respect to union participation is in the extent to which they are active in the administration of the union local. Out of the eight activities comprising the admin- istrative participation index, the average male union member is in- volved in 1.21 activities compared with only .80 for the average female. There is almost no difference between the sexes in terms of the average number of meetings attended during the year. Howh ever, women are somewhat more likely than men to have attended either the strike or contract ratification vote in the union local. These results confirm our earlier suspicions that the differ— ence in union participation between men and women varies with the type of union,activity and is not as pronounced as was observed in the pioneering research on this subject. However, we had not antic- ipated the possibility that men might be less involved than women in voting participation and that the difference with respect to meeting attendance is negligible. Table 10 presents the zero-order correlations and the standard- ized multiple regression equations for the three forms of union par- ticipation using the full sample. For comparative purposes, the variables are grouped in a manner consistent with earlier research. However, remember that each variable also corresponds to various components of the union participation model. 86 Table 9 Union Participation for Men and Women Participation Men Women Variable (N - 100) (N - 175) t-statistic Administrative 1.21"c .80 1.668° Participation (2.10) (1.69) Meeting 1.43 1.34 0.368 Attendance (2.08) (1.71) Voting 1.11 1.29 -1.535 Participation ( .95) ( .91) One-tailed test of significance ° p < .05 9 mean score ‘ standard deviation 87 Anoscwucoo oHnmuV one. see. so. so. use. go“. one. .sm. so. and. #0.- aka. mo. one.. one.. mo.- one. .ou. .HN.- .-.n one. and. no. men. mo.- no.. so. no. mo. mo.- «0.- No. no. and. cows: new :« mosoaum so.l oHH.I Ham: down: on economic Ha. mma. mnumum newshoanam uma. umN. hufiuoaeom mowumwuouomuoeu non\:ou:= 90.: Ho.| xuozomsom uo mnso: no. no. mnumum Humane: own. no. coaumonvm no. oHH. xom mowumaumuomumso amsnu>wnsH mamas: sowumaouuoo moon novHOIouoN :ofiummaowunmmhwdauo> u:m«03 soaumaouuoo moon nonnOIouoN memes: souumflonnou moon noouOIouoN ooamoeouuqumuumwnaau< oHAmHum> usonsonmvaH Anew . zv «Hagan seam "sowumewowunmm sous: mo mason amaze you msoaumnom nowmmoumom manauanz nouwvumosmum was meowumuouuoo HooHOIouoN OH wanna 88 Hoe. v a . Ho. v n A mo.vno ooamouwwsmwm mo mumou moaumuloso uHo.m~ um~.¢ unm.n cannula Hs~\ms Ho~\na Ho~\ma no om.qo~ mo.mnu «c.0wo mmm Ho.oma om.¢ca mm.mm~ Mum oases ¢>oz< mm. ma. «N. . Somme mm. ma. mm. Nm «o.| umH.u so.| mo.l mo.t co.| owed on usousH owe. co. so. «a. acH.| oNH.I usoso>ao>eu ooh oHH.I 40H.I onH.u mmH.I «o.| eNH.I coauomwmwumm ooh oamsfinuxm no.1 No.| and. umu. eon. 6mm. mmosamom code: aw umonouoH no. so. no. umH. cod. ohm. anode: mo o=Hm> moonuauu< och\coa:= names: coqumaouuoo memes: soaumaowuoo names: couumaouuoo oaoouum> comm noouonouou moon nonmalouou soon nonmalonos unonoononoH coaummaofiunemAMWHuo> oocmmsouuflumuumasuan< \ Aumaaenaoov on «Home 89 As Table 10 reveals, the zero-order correlations between sex and each of the three forms of union participation are congruent with the observations reported in Table 9. But notice that with the other variables in the equations, the sex coefficients either become less positive or become more negative. Specifically, men are not significantly more active than women in union administra- tion when the other independent variables are entered; women are significantly more likely than men to attend union votes when other variables are controlled; and women are slightly more likely to attend union meetings when other factors are taken into account. Together, these findings suggest that these equations include fac- tors which suppress female involvement in all three forms of union activity. However, the reader should also be warned that the ad- justed effect sizes with respect to the sex variable are quite small and should be viewed cautiously. Based upon our review of the literature, it was predicted that education is positively associated with administrative par- ticipation but probably has no effect on the propensity to either attend meetings or vote in union decisions. The data in Table 10 confirm this hypothesis. When countervailing effects (such as value of unions) are discounted, union members with higher lev- els of formal education are significantly more involved in the administration of the union, whereas the associations are non- ' significant for meeting attendance and voting participation. The predictions regarding marital status and union partici- pation pertain mainly to the separate male-female subsamples and the associations for the combined sample are expected to be 9O nonsignificant. The data support this hypothesis. With the exception of the zero-order correlation for voting participation, the coeffi- cients are close to zero. Housework was discussed as a factor which reduces the opportunity for union members to participate in time-consuming activities and, ac- cordingly, was hypothesized to be inversely correlated with both ad- ministrative participation and meeting attendance.n 0n the other hand, hours of housework should be unassociated with voting participation. The findings in Table 10 support the direction of these predictions, but the values are quite modest. There is partial support for our predictions regarding seniority. The length of service is positively associated with voting partici- pation and especially with administrative participation. The zero- order correlation is statistically significant for meeting attendance, but the effect is greatly attenuated in the regression equation. These observations lend support to our theses that having a greater stake in the job encourages attendance at union votes and that sen- iority is a criterion in the passage of union members into the admin- istrative roles of the union. In Chapter III we predicted that temporary employees would be less involved than permanent employees in all three types of union participation under investigation. Not only would they be less will- ing to participate because of the limited payoff, but they are not permitted to participate in certain events. These hypotheses are largely supported in Table 10. Given the fact that temporary em— ployees are not permitted in the union constitution from voting in either the strike or contract ratification votes, the large 91 coefficients for voting participation should come as no surprise. Permanent employees are also more likely to attend union meetings and participate in the administration of the union, although the latter relationship does not quite reach a level of statistical significance in the regression equation. Union members who live 10 kilometres or more from the union hall are much less likely to attend union meetings and, albeit non- significantly, are less involved in the union administration. In contrast, the association between distance to the union hall and voting participation is actually positive (although nonsignificant). Thus, the data provide some support for our belief that this vari- able is a situational contingency which affects the opportunity to participate in certain union activities. Consistent with the results from previous research, people who participate in social forms of union activity tend to have more friends in the union. As hypothesized, the association is strong- est for union meeting attendance, although the relationship is al- most as large for voting participation. Presumably, attendance at the strike and contract ratification votes provides more perceived opportunity for social interaction than we had assumed. In contrast, only a modest association is reported between the number of friends in the union and administrative participation. This suggests, as we postulated in Chapter III, that involvement in the direct admin- istration of the union is probably not motivated by social needs whereas social needs do explain participation in the other two forms of union activity. 92 The hypothesis that the correlation between union attitudes (including value of unions and interest in union business) and union participation increases with the degree to which the union activity is central to the operation of the union local is supported by the data in the present study. As Table 10 shows, Value of Unions is a significant predictor of administrative participation, whereas the coefficients are smaller (and nonsignificant) for the other two forms of union activity. This trend is even more prevalent for interest in union business. Interest in union business is the best predictor of administrative participation, but covaries less strongly with meet- ing attendance, and is uncorrelated with voting participation. Extrinsic job satisfaction is hypothesized to be inversely cor- related with voting participation and, to a lesser extent, with meet- ing attendance. This variable was not expected to covary with admin- istrative participation, however, because this form of participation is only marginally instrumental to the fulfillment of job-related needs. These predictions are partially supported. Extrinsic job satisfaction is significantly and negatively associated with voting participation and meeting attendance, but the magnitude of the ef- fects are only marginally different. Consistent with our expecta— tions, extrinsic job satisfaction is not significantly correlated with administrative participation when other variables are taken into account. The reader will recall from Chapter III that job involvement refers to the degree to which a person identifies psychologically with his or her job and that two apparently conflicting theories have been advanced in the literature with respect to the relationship 93 between job involvement and union participation. However, when we reconceptualize union participation as being multidimensional, the theories are more compatible. Specifically, we predict that employ- ees with low job involvement are more likely to be involved in the administration of the union because these roles can potentially 'compensate' for uninspiring work in the company. Those with high job involvement, on the other band, should be more involved in spe- cial union votes and union meetings because they have a greater stake in their jobs. The findings in Table 10 clearly support the compensatory hypoth- esis that union members with low job involvement will be more active in the direct administration of the union local. Apparently, union leadership roles provide an alternate outlet for job involvement. Employees with low job involvement are also less likely to attend special union votes, although the relationship is much weaker than we had expected. Moreover, job involvement does not significantly covary with meeting attendance. Consequently, the idea that highly job-involved union members have a greater stake in the job is only marginally supported. On the other hand, the compensatory hypoth- esis between job involvement and administrative participation ap- pears to be well established. The respondent's intent to quit was hypothesized to be negatively associated with meeting attendance and voting participation because of the minimal payoff from participation in these activities whereas a nonsignificant association with administrative participation was by- pothesized. Although the coefficients reported in Table 10 are in the predicted direction, the effect sizes are marginal. 94 Consequently, the hypotheses regarding intent to quit and union participation are rejected based upon these data. Quite possibly, a two-year intention to quit time span offers many of these individ- uals the opportunity to receive some payoff from union involvement. Approximately one-fourth of the variance is accounted for by the 13 variables in the administrative participation regression equation. Involvement in this type of activity appears to be mo- tivated by growth needs (as indicated by the job involvement var- iable) as well as an interest in union business and a positive attitude toward unions in general (i.e. value of unions). Educa- tion, seniority, and (although not significantly) employment status and distance to the union hall are identifiable opportunities or constraints to administrative participation. Only 13 percent of the variance in meeting attendance is ex- plained by the regression equation presented in Table 10. The find- ings suggest that attendance at general union membership meetings is motivated by job-related needs (e.g. extrinsic job satisfaction and employment status), social needs (e.g. friends in the union), and union attitudes (e.g. interest in union business). The distance to the union hall is the most prevalent situational contingency, al- though housework might also have a slightly adverse effect on the number of meetings attended. Lastly, over one-half of the variance in voting participation can be understood in terms of the variables in the regression equa- tion displayed in Table 10. Attendance at the strike and contract ratification votes is clearly motivated by the individual's job- related needs (e.g. extrinsic job satisfaction, seniority). There 95 is also some evidence that voting participation fulfills social needs (e.g. friends in the union). Most predominant in this equation, howh ever, is the legal barrier that temporary employees have to voting in these special union decisions. A Male—Female Comparison of Union Participation In this section, we present the results of the male and female subsamples analysed separately. First, we compare the levels of edu- cation, housework, seniority, and other independent variables between the sexes to determine whether differences in union participation could be due to different endowment levels. Then, for each of the three forms of union participation, we look at the separate regression equations for men and women to analyse the possibility of structural differences. With the exception of marital status (for which we have proposed specific structural hypotheses), the latter analyses will be exploratory. Table 11 compares the differences in means between men and women for each of the 12 independent variables (the sex variable is, of course, excluded). As can be seen, only for hours of housework is there a statistically significant difference below the .05 level. Specifically, women in our sample perform an average of over 19 hours of housework per week compared with only 11 hours per week for men. If housework adversely affects union participation as several writers have argued (Cook, 1968; Glassberg et a1., 1980; Koziara & Pierson, 1980), then the significant difference reported ianable 11 would have partly explained the male-female difference. However, housework does not appear to significantly deter members from engaging in any of the 96 Table 11 Means, Standard Deviations, and t-statistics or Independent Variables for Men and Women Men Women Variable (N - 100) (N - 175) t-statistic Education 3.98 4.15 -1.114 ( 1.27) ( 1.12) Marital Status .68 .61 1.170 ( .47) ( .49) Hours of Housework 11.53 19.24 -5.931* ( 9.74) (11.39) Seniority 81.06 72.23 .963 (75.11) (69.52) Employment Status .76 .74 .368 ( .43) ( .44) Distance to Union Hall .19 .26 -1.366 ( .39) ( .44) Friends in the Union 2.57 ' 2.43 1.114 ( .98) ( 1.04) Value of Unions 43.50 44.45 - .762 (10.31) ( 9.26) Interest in Union Business 1.08 1.07 .072 ( 1.11) ( 1.10) Extrinsic Job Satisfaction 56.71 60.49 -l.452 (20.87) (20.58) Job Involvement 22.37 23.02 - .640 ( 8.32) ( 7.71) Intent to Quit 1.29 1.29 .000 ( 1.64) ( 1.59) Standard deviations in brackets One-tailed test of significance * p < .001 97 three forms of union participation in this study (see Table 10). Thus, we cannot support the view that housework prevents women from being in- volved in union activities. Two authors have either postulated or observed that women are less likely to hold union leadership roles because they have less for- mal education (Cook, 1968; Tooredman, 1981). As we observed earlier (Table 10), education is a major predictor of administrative partici- pation. But as we see in Table 11, women in the present sample have ‘gg£5_education on average than their male counterparts. Therefore, education is not a factor explaining male-female differences in this study (although it might be a valid factor in other union locals). Also contrary to our expectations is the fact that women in this sample are no more likely than men to have turnover expectations and are only slightly more likely to be temporary employees. Thus, neither of these variables explains male-female differences in union participation in this study. Even seniority does not significantly differ between men and women in the union local under investigation. To sum up the evidence from Table 11, none of the variables pre- dicted to explain male-female differences in the three forms of union participation actually support our predictions. Men and women signif- icantly differ on only one variable - hours of housework - but this factor is not a significant predictor of union participation. For other variables which do correlate significantly with any of the three dependent variables, the mean differences for men and women are negli- gible. The separate regression equations for the male and female subsamr ples are presented in Tables 12 - 14 for administrative participation, 98 Table 12 Zero-Order Correlations and Standardized Multiple Regression Equations for Administrative Participation: Male and Female Samples Male Sample (N - 100) Female Sample (N 3 175) Independent Zero-Order Beta Zero-Order Beta Variable Correlation Weight Correlation Weight Individual Characteristics Education .12 .31c .05 .19° Marital Status .21° .16 -.07 -.07 Hours of Housework -.01 -.14 .05 -.02 ‘Qgion/Job Characteristics Seniority .18° .08 .29c .25c Employment Status .25b .16 .13“ .07 Distance to Union Hall -.24b -.20° -.01 .02 Friends in the Union .22“ .11 .12 -.02 Qgégg/Job Attitudes Value of Unions .349 .24“ .22b .07 Interest in Union Business .33c .28b .34c .3r= Extrinsic Job Satisfaction -.09 -.08 -.12 -.02 Job Involvement -.03 -.13 -.15° -.20b Intent to Quit —.02 .00 -.1O -.07 32 .39 .25 RZadj. .31 .20 ANOVA Table SSR 171.76 125.34 SSE 264.83 372.66 df 12/87 12/162 F-ratio 4.70c 4.54“ One-tailed tests of significance clp<.05 bp<.0]. ‘ p'< .001 99 Table 13 Zero-Order Correlations and Standardized Multiple Regression Equations for Meeting Attendance: Male and Female Samples Male Sample (N - 100) Female Sample (N - 175) Independent Zero-Order Beta Zero-Order Beta Variable Correlation Weight Correlation Weight Individual Characteristics Education -.16 .03 .07 .16° Marital Status . 16 . 08 -. 03 . 02 Hours of Housework .00 -.10 -.01 -.O7 Union/Job Characteristics Seniority .12 -.08 .14° .08 Employment Status .265' .18 .12 .11 Distance to Union Hall -.26b -.21° -.20b -.19° Friends in the Union .24b .09 .13b .13 Union/Job Attitudes Value of Unions .31c .219 .10 -.04 Interest in Union Business .17“ .07 .27c .25b Extrinsic Job Satisfaction -.219 -.19 -.10 -.07 Job Involvement .08 .02 .01 .02 Intent to Quit -.14 -.O4 -.05 -.02 32 .25 .17 Rzadj. .15 .11 ANOVA Table SSR 106.07 86.64 SSE 320.44 424.47 df 12/87 12/162 F-ratio 2.40b 2.76b One-tailed tests of significance ° p <=.05 b p <1.01 C p <.001 Zero-Order Correlations and Standardized Mu1tiple Regression Equations for Voting Participation: 100 Table 14 Male and Female Samples Male Sample (N - 100) Female Sample (N - 175) Independent Zero-Order Beta Zero-Order Beta Variable Correlation Weight Correlation Weight Individual Characteristics Education -.11 .05 -.15° .04 Marital Status .31c .03 .11 -.03 Hours of Housework .10 .01 .14° .02 Union/Jobggharacteristics Seniority .26b .05 .40c .12° Employment Status .66‘ .65c .73c .68c Distance to Union Hall -.03 -.01 .10 .10° Friends in the Union .13 .09 .20b .12° UnionZJob Attitudes Value of Unions .10 -.O3 .05 -.07 Interest in Union Business -.05 -.09 -.01 -.O4 Extrinsic Job Satisfaction -.219 -.10 -.15° -.12° Job Involvement .09 .14 .00 .07 Intent to Quit -.08 .12 -.26c -.12° R2 .48 .62 22.81. .41 .59 ANOVA Table SSR 43.23 89.14 SSE 46.56 54.57 df 12/87 12/162 F-ratio 6.73c 22.05c One-tailed tests of significance ° p <1.05 5 p ‘=.01 c p <.001 101 meeting attendance, and voting participation, respectively. Most of the discussion that follows is based upon exploratory analysis. The exception is with respect to the marital status variable for which we have specific hypotheses. Looking first at the marital status beta weights in the three tables, the reader will see that none are significantly different from zero. Moreover, the direction and relative magnitude of the coefficients are not what we predicted. Married women were hypoth- esized to be less involved than unmarried women in union administra- tion and to attend fewer general membership meetings, but the data do not indicate that this is the case. Either marriage is not such a burden to prevent women from engaging in these union activities or, as we speculated for housework, women are able to schedule their re- sponsibilities in such a way that union participation is not crowded out. Examining only the bivariate marital status correlations for the male sample, it appears that married men are more active than un- married men in the union, especially in attending the strike and con- tract ratification votes. However, these effects wash out in each of the three multiple regression equations. It is noteworthy that the bivariate marital status coefficients for both men and women are consistent with what Tannenbaum and Kahn (1958) reported many years ago. But this covariance is actually due to the simultaneous effect of other variables in the equations as documented by the adjusted coefficients. The regression equations for men and women vary in the magni- tudes of the seniority coefficients. Particularly in Table 12 and Table 14, length of service is a significant predictor of 102 administrative and voting participation, respectively, only in the female sample. Quite possibly, union members give seniority more weight when considering women for union office than when choosing a male candidate. An explanation for the discrepancy in the seniority effects for voting participation might be that there is a large var- iance in the felt stake in the job between female employees who have been with the organization only a few years and female employees with many years of service. In contrast, male employees with little sen- iority might not have a substantially lower stake in the job than male employees with longer seniority. Ultimately, a more direct estimation of the felt stake in the job (i.e. job importance) needs to be intro- duced to verify these post hoc explanations. Another curiosity is the relative importance of value of unions and interest in union business in the regression equations for men and women. As can be seen in Table 12 and Table 13, the administra- tive participation and meeting attendance of women is highly predicted by the respondent's interest in union business but the more general union attitude (i.e. value of unions) has no significant effect. For men, either value of unions is the predominant predictor or both var- iables are relevant (as in the administrative participation equation). We interpret these observations as an indication that women union mem- bers engage in administrative duties and attend union meetings for: less ideological reasons than their male counterparts. A few other differences in the structural equations for men and women in Tables 12 - 14 are evident (e.g. distance to union hall, in- tent to quit), but the rationale for their existence remains elusive to this writer. Given the small subsample of men and the possibility 103 of idiosyncrasies within any sample, it would be unwise to overinterpret these structural differences. For the majority of variables, however, the zero-order and standardized regression coefficients do not differ significantly between men and women. Summary The only significant male-female difference in union participa- tion in the present study was in the administration of the union local. Men are more involved than women in union administration, although the difference is not as large as expected. No sex difference was observed in the number of meetings attended, and women are slightly (signifi- cantly in the regression equation) more likely to attend the special union votes. Many of the hypotheses pertaining to the model of union partici- pation were supported. The predictive power of union attitudes (i.e. value of unions and interest in union business) increases with the centrality of the union activity to the operation of the union local. Personal needs also add to our understanding of participation in dif- ferent union behaviours. Job-related needs covary with voting partici- pation and union meeting attendance. Social needs motivate attendance at group activities such as general membership meetings and special union votes. And prepotent growth needs apparently encourage involve- ment in the administration of the union. The opportunity to participate also appears to be a major factor in all three forms of union involvement. The effect of the legal bar- rier that temporary employees cannot vote in union decisions was quite apparent in the analysis. With respect to meeting attendance, the 104 distance from the respondent's home to the union hall was the predominant situational contingency. And for involvement in the administration of the union, it appears that education, seniority, and (although not significantly) employment status and distance to the union hall affect the opportunity to participate. In contrast to the empirical support received for the basic model, none of the the six predictions associated with the male- female difference in union participation was supported. WOmen in the sample perform nearly twice as much of the housework compared with men, but the number of hours of housework does not seem to ad- versely affect involvement in any form of union-related behaviour. In none of the other independent variables was the mean value for women significantly different from the mean value for men. Analysis of the separate regression equations for the male and female subsamples showed that married women are no less likely than unmarried women to participate in the administration of the union or attend union meetings. The seniority coefficients are substantially larger for women than for men with respect to administrative and voting participation. Also, it seems that women participate in the union administration and attend union meetings for less ideological reasons than do men. For most other variables, the coefficients do not substantially differ between men and women. CHAPTER VI DISCUSSION AND CONCLUSIONS Introduction In this final chapter we generalize the findings reported in Chapter V in the context of the three research objectives stated at the outset of this manuscript. In addition, several implications of our research for practitioners (union leaders) and researchers are discussed. we begin, however, by outlining the major limitations of the present study. Limitations Before proceeding to generalize the observations reported in Chapter V, it is first advantageous to remind the reader that this study is not without its limitations. The first, and perhaps fore- most, concern is that it was not possible to directly operationalize the personal need variables in the union participation model. The reason for this, as we have noted elsewhere, is that there presently does not exist a set of personal need measures which can be effi- ‘ciently employed in a survey format. Steers and Braunstein (1976) recently developed the 20-item Manifest Needs Questionnaire to meas— ure four needs "with specific reference to work settings and with minimal time requirements for completion" (pg. 251). Unfortunately, a recent review of nine studies which used this instrument concluded 105 106 that it lacks internal consistency reliability (Dreher & Mai—Dalton, 1983). Other need instruments are either not easily applied to our research design (e.g. projective tests) or are too lengthy for respon- dent acceptance. Two studies on union participation have included measures of various needs, but their reliability and/or validity are unknown (Glick et a1., 1977; Nicholson et a1., 1981). To offset this limitation, we have carefully outlined how various background variables (e.g. seniority, friends in the union) and prox- imate attitudes (e.g. extrinsic job satisfaction, job involvement) correspond to union participation through these personal needs. This not only serves to give added meaning to the variables which have been included in much of the earlier literature, but it also places them within a more theoretical framework for future investigations. A second limitation of this study is that the data have been col- lected from only one union local. We explained in Chapter IV that or- ganizational characteristics seem to contribute less than individual level variables to our understanding of local labour union participa- tion (Anderson, 1978; Tooredman, 1981) and that financial constraints prevented us from adopting a multi-unit research design. Nevertheless, this'is not to suggest that organizational factors are totally irrele- vant to our understanding of involvement in union locals. To the ex- tent that macro-level variables are pertinent, valid generalization of the findings reported here is reduced.. To minimize this adversity, the selection of the union local studied was, in our opinion, based upon careful criteria to enhance generalizability to the majority of Canadian union locals. Only further analysis in new research settings will determine the validity of this assumption. 107 A third potential weakness with this investigation is that a cross-sectional research design has been employed to investigate an essentially longitudinal phenomenon. In this regard, we have plenty of company. In fact, we know of no longitudinal studies of local labour union participation. Fortunately, the majority of variables (e.g. education, seniority, etc.) have a natural temporal precedence to the dependent variables; most others follow a sequence of causa- tion based on logic (Nicholson et a1., 1981: 58). Nevertheless, it is possible that value of unions, for example, is a function of union participation rather than vice versa (Gordon et a1., 1980). Ulti- mately, it is hoped that the author's future research will be able to more directly test these causal propositions in the union partici- pation model. Finally, a few readers might point to the proportion of union members responding to the questionnaire as another possible limita- tion on the validity of the research findings. Although one cannot be absolutely certain of such bias, the comparison between respon- dents and nonrespondents on several major variables (Table 5) sug- gests that our sample is quite representative of the designated population within the union local. Furthermore, we should point out that most other recent studies of local labour unions have had response rates well below 50 percent (Anderson, 1979; Andiappan & Chaison, 1983; Gordon et a1., 1980; Hagburg, 1966; Nicholson et a1., 1981; Portwood et al., 1981). Overall, we believe that the limitations described above do not severely restrict either the validity or the generalizability of the 108 results. However, they do point out aspects of the present research which could be improved in subsequent investigations. Discussion Although the primary focus of this study is to present a male- female comparison of participation in a local labour organization, several major hypotheses were also described with respect to the con- ceptualization of union participation and the validity of the model of union participation outlined in Chapter II. Accordingly, we shall discuss these three sets of findings separately. Multidimensionality of Union Participation. Throughout this manuscript, union participation has been conceptualized and opera- tionalized as a multidimensional (rather than global) construct. In other words, there are different typgg_of involvement in local labour organizations which shoUld be studied separately instead of being thrown together into a single measure. The reasons given for this reconceptualization were two-fold. First, union participation behaviours vary in the degree to which they are central to the operation of the union. Most writers on this subject have recognized, for example, that holding office and other administrative forms of involvement are the most central to the operation of the union local and, therefore, are substantively different from more passive types of union participation (Sayles & Strauss, 1953; Spinrad, 1960; Tannenbaum & Kahn, 1958; Vall, 1970). Indeed, attempts have been made to place union-related behaviours along a continuum representing their centrality to (and influence on) the operation of the union local (Nicholson, 1978). 109 The second reason why union participation should be conceptualized as a multidimensional construct is that these behaviours vary to the extent that they are influenced by different factors. In Chapters II and III we described the findings from numerous studies which indicate that the motives, attitudes, background characteristics, and opportune ities/constraints associated with administrative participation are dif- ferent from those associated with meeting attendance which, in turn, are different from those associated with voting behaviour. Two stud- ies have already called for a refinement of the union participation construct on these grounds (McShane, 1982; Portwood et a1., 1981). The results of the present study emphatically support the multi- dimensional reconceptualization of union participation. Evidence that administrative participation, meeting attendance, and voting partici- pation form three homogeneous clusters was presented in Chapter IV. To begin with, the estimates of internal consistency reliability for the two multi-item scales (i.e. administrative and voting participa- tion) were substantially higher than the marginally acceptable reli- abilities computed for more global indexes. Moreover, the correla- tions among the three dependent variables were much smaller than the correlations among the items within each multi-item index. Together, this information indicates that administrative participation, meeting attendance, and voting participation form three homogeneous groupings. Further support for the multidimensionality of union participa- tion came from the structure of each of the three regression equations display in Table 10. Specifically, variables which correlated with one form of union participation were either uncorrelated or inversely correlated with the other two forms of union participation in a 110 predictable way. Thus, it appears that union participation behaviours do, indeed, vary to the extent that they are a function of different influences. Finally, the proposition that union behaviours vary in the degree to which they are central to the operation of the union local was not directly examined, but there is indirect evidence to support it. Specifically, we explained that union attitudes would covary more strongly with central union activities than with activities which are peripheral to the union operation in terms of the time and energy ex- pended and the amount of influence generated. As predicted, union members who participate in the administration of the local union have more positive union values and interest in union business than others, whereas union attitudes are not significantly different between mem— bers who attend union votes and those who do not. To some extent we can interpret this as an indication of the centrality of union admin— istration relative to voting participation and, more generally, that union participation behaviours do vary in their degree of centrality to the operation of the union local. In summary, the present study supports the idea that union participation is a multidimensional con- struct. Validity of the Union Participation Model. Although consider- ably more research is required, the findings from this study lend support to the union participation model described in Chapter II. Briefly, the model states that union participation is a function of the willingness to engage in that behaviour and the opportunity to do so. The willingness to participate in a particular type of union 111 activity is determined by the prepotence of certain personal needs (i.e. job-related, social, growth) and the individual's attitude to- ward unions. The importance of opportunity variables in the propensity to engage in union activities is quite evident from the data analysed in this study. Education and seniority are inversely correlated with administrative participation, suggesting that members with lower ed- ucation and seniority are screened (or self-select themselves) out of these roles. The distance to the union hall is a major constraint against attending meetings and marginally deters involvement in the union administration. And for voting participation, the regression analysis clearly reflects the union constitution's prohibition of temporary employees from casting their ballot. Thus, gate keeping, time-space, and legal factors constrain involvement in different union activities. Surprisingly, hours of housework was not a sig- nificant deterrent to either administrative participation or meeting attendance as we had predicted. Quite possibly, housework can be rescheduled to avoid conflict with the timetable of union business. Our results also indicate that union attitudes are major factors in union participation. Specifically, we hypothesized that the as- sociation between union attitudes and union participation increases with the degree to which the union activity is central to the opera- tion of the union local. The evidence that we presented in Chapter V confirms this proposition. The correlation between union attitudes and administrative participation is strongest, followed by the corre- lations with meeting attendance and voting participation, respectively. 112 Although this information does not directly determine whether union attitudes affect the willingness to participate in the union, such a relationship has been documented in previous research (Glick et a1., 1977; Gordon et a1., 1980). Consequently, the position of union at- titudes in the model seems to be appropriate. Lastly, there is some support, albeit indirect, for the role of personal needs (and the expected fulfillment of these needs through union participation) in the union participation model. The signifi— cant effects of seniority and extrinsic job satisfaction suggest that job-related needs are most relevant to voting participation. This is as we had predicted since voting in the strike and contract ratificar tion decisions are highly instrumental means of serving job-related interests. However, there is some indication that voting participa- tion also serves a social function, as indicated by the significance of the friends in the union variable. we anticipated and found that social needs are associated with meeting attendance. As was discussed in Chapter III, earlier research has indicated that having union friends leads to meeting attendance rather than vice versa (Geare et a1., 1979; Hagburg, 1966; Spinrad, 1960). Thus, although the direction of causality has not been dir- ectly tested in this study, the development of social relations in the union tends to lead to meeting attendance. Our data also indicate that job-related needs influence meeting attendance. we had predicted only a marginal association with extrin- sic job satisfaction based upon our understanding that general member- ship meetings are time-consuming and are more directed toward the 113 internal operation of the union local than to directly improving job conditions (Sayles & Strauss, 1953). It seems, at least in this union local, that members who are dissatisfied with their job conditions are more likely to attend the general membership meetings.' Unlike meeting attendance and voting participation, administra- tive participation is not greatly affected by either job-related or social needs. Rather, we have presented indirect evidence, consis- tent with some earlier observations (Spinrad, 1960), that employees who are highly job-involved are less likely to participate in the ad- ministration of the union. This supports the compensatory hypothesis that those who do not hold intrinsically fulfilling jobs will seek out union positions to satisfy their growth needs. Of course, other out- lets are also available (e.g. holding key positions in community groups). If we were able to account for these alternate outlets of growth need fulfillment, the correlation between job involvement and administrative participation would be even stronger. To summarize, the results of our study provide substantial evi- dence in support of the union participation model. Opportunity var- iables play a prominent role in determining the level of participation in a given union activity. General union attitude is also an important component, especially with respect to union behaviours which are more central to the operation of the union organization. Finally, although only indirectly tested, it appears that personal needs are important influences on union participation. In particular, growth needs moti- vate administrative participation while social and job-related needs influence meeting attendance and voting participation. 114 A Male-Female Comparison of Union Participation. In contrast to the results in support of the multidimensionality of union participa- tion and the validity of the union participation model, this study has failed to identify many significant male-female differences either in the level of union involvement or in the factors contributing to union involvement. No sex difference was discovered for meeting attendance and women appear to be slightly more likely than their male coworkers to attend the strike and contract ratification votes. Only with re- spect to administrative participation are men more active than women, and even this distinction is marginal. These observations reflect the historical trend toward the greater involvement of women in their local labour unions. While much of the pioneering case studies reported an almost total nonin- volvement of women in labour unions (Purcell, 1954; Sayles & Strauss, 1953), recent investigations have found only slight (often nonsignif- icant) differences (Anderson, 1979; Huszczo, 1975; Kochan, 1978; Nicholson et a1., 1981; Portwood et a1., 1981). This trend is pos- sibly due to the changing role of women in the workplace. In the union local upon which our results are based, women have only slightly less seniority than men, are just as likely to have permanent posi- tions, are no less likely to be involved in their jobs, and have no higher intentions to quit their jobs. By inference, local union pro- cesses and outcomes probably have no less importance to women than men in our sample. Given the minimal differences in union participation, the rejec- tion of our hypotheses relating to the reasons for male-female differ- ences in union participation is hardly unexpected. As we pointed out 115 above, men and women in our sample did not vary significantly in the characteristics which might contribUte to different levels of union involvement (e.g. seniority, employment status, job involvement, turn- over intentions). Women are responsible for much more housework than their male counterparts, but housework is not a significant deterrent to any of the three forms of union participation (although it is in the predicted direction). With respect to the structures of the regression equations for men and women, a few minor differences have been identified. Senior- ity is an important predictor of administrative participation for women while the coefficient is nonsignificant for men. Perhaps union members (and union executive members in the case of appointed posi- tions) look more closely at the seniority of women before choosing them for union office. The seniority coefficient is also larger for women than men in the voting participation equation. Quite possibly, women vary more than men in their stake in the job and that this var- iance closely parallels seniority. Finally, it appears that women participate in the union administration and attend union meetings for less ideological reasons than men do. Overall,-women are slightly less involved than men in union ad- ministration, but not in other types of union activity. With the ex- ception of housework which did not have a significant effect on union participation, men and women were not substantially different on any of the variables cited in the literature as causes of the lack of female involvement in local labour unions. In the exploratory anal- ysis of the separate regression equations, men and women differed in only a few variables. 116 Conclusions The research findings presented in this manuscript have significant implications for both practitioners and researchers. These applications are discussed in this final section. Union participation research has frequently been justified on the grounds that member involvement enhances union democracy (Anderson, 1979; Nicholson, 1978; Ramaswamy, 1977; Strauss, 1977; Tannenbaum, 1965). To most union leaders today, however, union participation is essential for the very survival and maintenance of the local union organization. This is because a union local cannot function for very long when quorum is not reached at union meetings, when few show up to vote in special union decisions, and when important union posts are difficult to fill. Even in the union we studied, where union elections are often compet- itive and a large majority of the membership votes in.the strike and contract ratification decisions, the executive board members are still discouraged with the general apathy of the membership. Support for the union participation model and the multidimensional conceptualization of the construct provides useful insight into the dy- namics of union participation. Specifically, our findings suggest that different union activities serve different needs. This is an important consideration in any attempt to motivate union involvement. If the labour movement hopes to attract motivated individuals to lead union local operations, then these posts must provide for the fulfillment of growth needs, including challenge, responsibility, and power. If authority and resources are beyond the control of the union local ad- ministrators, these positions will be difficult to fill (Vall, 1970). 117 Many unions use incentives such as free food, door prizes, and even dues refunds to encourage meeting attendance, usually with little success (Miller & Young, 1955). Intermittently over the two years prior to our research, the union local we studied used door prizes to draw the membership to the monthly general meetings. In surveying the union meeting register, however, there did not appear to be any observ- able relationship between the number of members attending and the pres- ence or absence of a door prize. Part of the problem with these incen- tives, we surmise, is that they are not directly pertinent to the per- sonal needs that union members wish to fulfill through these functions. Unless the financial inducement to attend is quite large (an expensive proposition!) then attendance is not likely to increase noticeably if the meeting does not fulfill social and/or job-related needs. Accordingly, one solution to poor meeting attendance is to dev- ' elop social ties between new employees and more senior union members who are active in the union. This can be achieved through the simple process of informal introductions and discussion of union business during working hours. Indeed, several respondents in our survey com- plained that no one from the union introduced them to the labour organ- ization. As one member wrote: In 17 months of employment, I have not been informed of any union activities, elections, or decisions. The only contact that I've had...is the deduction of union dues without fail every two weeks. Another member recalls: When I joined the union, no one told me anything about my rights. As a result, many facts familiar to me were stumbled upon. 118 While these comments indicate a lack of communication between the administration and the membership (in spite of a regular newsletter), they also point to a lack of personal interaction with new union meme bers. Yet such interaction could help to socially integrate new em- ployees into the local labour union and thereby motivate attendance at general membership meetings and special union votes. A second method of increasing meeting attendance is to enhance the probability that desired job-related outcomes would result from such union activity. All too often, general membership meetings get bogged down with trivial items while issues important to the member- ship are not allocated sufficient time. A common complaint, for ex- ample, is that meetings are too boring. A few respondents in our sample also commented that the "important" issues have already been decided by the executive prior to the general meeting. In short, meeting attendance is often low because members do not see their job-related needs being fulfilled through this form of union par- ticipation. Our findings add to growing evidence that women are becoming more and more involved in their local labour organizations. In the present union, which we believe to be quite representative of many union locals across Canada, women are just as likely as their male counterparts to attend the monthly union meetings and they were ac- tually more likely than men to attend strike and contract ratifica- tion votes. Women were still less involved in the direct administra- tion of the union local, but the difference was marginal. This information suggests that complaints about female underrepresente- tion are not as well-founded as they once were. Indeed, even at the 119 national level, women are taking prominent union posts in Canada. At the time of this writing, the president of the country's largest union (the Canadian Union of Public Employees) and the vice-president (and, according to rumour, the next president) of the Canadian Labour Congress are women. During the period of research, the author was invited by the local union executive to attend the annual province-wide conference on women which is hosted by the national union. The conference not only mani- fests the awareness that many unions have toward issues of particular relevance to women, but the large number of women from local union ex- ecutive boards reinforces the generalizability of our findings to other labour organizations, at least in this sector of the economy. The implications of the present results for future research are very important because they indicate that new directions must be taken. First, it is important that future investigations treat different union activities as separate constructs rather than grouping them altogether into a global index. We have identified three clusters of union re- lated behaviour - administrative participation, meeting attendance, and voting participation. Quite possibly, there are other clusters of union participation to be discovered and studied. Future study of union participation should also attempt to meas- ure personal needs more directly and thereby provide a better test of the intermediate linkages in the union participation model. In par- ticular, it would be both interesting and useful to determine which types of growth needs (e.g. challenge, autonomy, responsibility, etc.) tend to motivate involvement in the administration of the union the most . 120 Research on this subject must also become longitudinal. As previously indicated, this writer is unaware of any empirical union participation analysis which is not based upon cross-sectional data. While this is not a concern where objective variables (such as edu- cation, seniority, sex, etc.) with unambiguous temporal precedence to the dependent variable are used, the increasing employment of at- titudinal constructs ultimately requires that data be collected at more than one time period in order to accurately assess causality. One excellent idea for such research would be to study new employees through their first few years of employment in a unionized company. Using multiple time period data collection, the researcher can an- alyse the conditions under which new employees become psychologically attached and integrated into the local labour union and begin to par- ticipate in that organization. Lastly, future research should attempt to explore male-female differences in union participation in other local labour unions. Our data reveal only marginal differences between men and women in union participation and cast into doubt the validity of several assumptions about women in the workforce. For example, in this union, paid em- ployment appears to be just as important to women as it is to men. No significant differences were found with respect to seniority, em? ployment status, job involvement, or turnover intentions. It would be useful to compare these observations with other union locals in Canada and internationally. The idea that women perform the bulk of the housework is strongly supported in this research, but the amount of housework apparently has little or no effect on the respondent's opportunity to participate in 121 even the most time-consuming activities. It would be useful to explore this rescheduling of domestic and union obligations further. For instance, to what extent do pro-union attitudes influence the willingness of union members to reschedule other duties to attend general membership meetings and other union events? Are some forms of housework (e.g. taking care of children) more difficult to re- schedule than others? If so, these discoveries would have practical implications for increasing union meeting attendance (e.g. the func- tionality of a babysitting service during meetings). Finally, the different predictive effects of seniority and union attitudes on the union participation of men and women beg further study. More direct tests should be conducted to discover if and why women participate for more pragmatic rather than ideological reasons than do men. Also, future research should look more closely at the relative importance of seniority in the participation of women in local labour unions. Is there a more extreme difference in job at- tachment between lowbseniority and high-seniority women that for men? Is seniority a more important criterion in the election and appoint- ment of women than men into union office? These and other questions deserve further attention. APPENDIX APPENDIX A YOUR UNION AND YOUR JOB A SURVEY OF ....... MEMBERS INSTRUCTIONS 1. PLEASE READ THE ATTACHED COVER LETTER CAREFULLY. 2. EACH QUESTION SHOULD BE ANSWERED COMPLETELY. PLEASE BE AS TRUTHFUL AS POSSIBLE. REMEMBER. THE INDIVIDUAL INFORMATION YOU PROVIDE WILL ONLY BE KNOWN TO THE RESEARCHER. 3. WHEN YOU HAVE COMPLETED ALL THE QUESTIONS. SEAL THE BOOKLET IN THE ATTACHED ENVELOPE AND RETURN IT THROUGH INTERDEPARTMENTAL MAIL TO THE UNION OFFICE. THE RETURNED QUESTIONNAIRES WILL BE OPENED AND SEEN ONLY BY THE RESEARCHER. NO ONE IN EITHER THE UNION OR MANAGEMENT WILL HAVE ACCESS TO THIS COMPLETED BOOKLET OR THE INFORMATION IT CONTAINS. 4. PLEASE DO NOT WRITE YOUR NAME ANYWHERE IN THIS BOOKLET. 5. IF YOU HAVE ANY OUESTIONS, PLEASE TELEPHONE ME (STEVE MCSHANE) COLLECT AT (517)-323-4222, PREFERABLY BETWEEN 6P.M. AND 11P.M. OR ON WEEKENDS. 6. THANK YOU VERY MUCH FOR YOUR ASSISTANCE IN THIS STUDY. 122 12.3 YOUR UNION AND YOUR JOB - ASURvEY OF------MEMBER$ A. UNION PARTICJIPATION um mmss 1. Before you complete the other questions in this section, it is a good idea to first tell us how long you have been a member of -----. YEARSAND mNTHSASAMEMBEROF----- 2. Within the past 12 months, did you vote in any of the following? Please circle either YES or NO for each type of vote. INTHEPASTTWELVE _______MCII IW' ' ' (Please circle) a) Voteinthe 1981 strike vote?.............. ...... NO YES b) vat. in th. 1981 contra: “tifiutmzseeseeesessssseessesssseeeeesmes No m c) Vote in the mien elections last Decedaer (for president and warden)?.. NO YES 3. Did you attend any of the general membership meetings held in the past 12 months? Please indicate your answer with a check mark (V)- ( ) N0 ( ) Its-1 (If YES) a) How many meetings did you attend in the past 12 months? NUMBER OF GENERAL MEMBERSHIP WINGS ATTEND- b) Thinking back to a typical meeting that you have attended in _- --_, to what extent did you ask guestim or state zgur views compared with others at that meeting? ()WCHPDREOMTHANOTHERSATTHEMEETING )SOMEWHATDDREOHENTHANOTHERSATTHEMEETING ( ) ABOUT THE SAME As OTHERS AT THE MEETING ()SOMEWHATLESSOFTENTHANOTHERSATTHEMEETM ()WCHLESSOFTENTHANOTHERSATTHEMEETING c) To what extent did you provide information requested at the meeting? ()MCHmREOFTmTHANOTHERSATTHEMEETING ()SOMEW‘HATIQREOF'TENTHANOTHERSATTHEBETING ()ABOUTTHESAMEASOTHERSATTHEMEETDIG ()SOMEWHATLESSOFTENTHANOTHERSATTHEMEETING ()MUCHLESSOFTENTHANOTHERSATTHEMEETING it. How far is your home from the union office on - - - - _ Road? MILES OR KILOMETRES FROM HG‘IE TO THE UNION OFFICE \l| e 124 People are occasionally unable or unwilling to attend union meetings for a variety of reasons. Please indicate the extent to which the following things prevent you from attending union meetings. TO WHAT EXTENT DO THE FOLLOWING Pntvmvoumouammmcm... L91°33° circle your answer) a) Distance from home to the place where mm ,m m m... the meetings are usually held. . . . . . .. . . ....... mum m mum use: some: swan: your: b) Lack of child care during meetings......... m m m m c) Conflict of hours of work with the time an“ m m 2 that the meetings are usually held............ m m mu m var m m m m. d) 1':- usually not interested in what goes on. m nature mum mus; e) I have other outside events (eg. school, ' other groups) at the same time as the meetings. :3: rm; fig ”a? f) My spouse, fanily, or friends discourage vm: a m m “.1. me from sttsndingunionmeetings.............. menus m mum null! 3) Home responsibilities don't allow me man m man: your: enough ti- togo to meetings................. HIV-u nay-m ass-m m h) Lack of information about union meetings... x a1: w w“; Below is a list of activities that some members of- _, - - _ are involved in or have recently taken part in. Please circle either YES or NO to indicate whether or not you have been involved in any of these activities in the past twelve months. IN THE PAST 12 MONTHS. HERE YOU...... {Please circle) s)Adelegate for-__-_ ataconvention or seminar? NO YES b) Answer ofaconitteein-____?.................................. NO YES c) A candidate in a -_--_election (whether or not you won)?.......... NO YES d) Anelected official in_____?...................................... NO YES e) A holder of any appointed position in _, -- - - not listed above?...... NO YES Women's issues have recently gained much attention among unions in Canada. In your opinion, what priorig should each of the following issues be given in 's negotiating agenda? Both men and women should answer this question. WHAT PRIORITY LEVELVOULD YOUGIVETO..... SPlease circle your a_nswer) a) Equal pay for equal work ‘32“, gm zmn b)Pudut.m:y1“"eeeeeeeeeeeeeeeeeeeeeeee gm 3“ gm :m C) Child Ctr. (1a.. 26 1101!: “y c‘n)eeeeeseeee $m m La,” M d) Affirmative action (i.e. ensuring that women are given ample opportunity to enter all mm m w- ,0 jobs at all levels in the workplace)........... mm moan-r mm mm a) Women's issues in general compared with m m as so mourn moans moan-r mourn Oth.rn.'°t“c1°n1““.00000000OOOOOOOOOOOOOOO 10. ll, 12. 125 _When you were growing up, what opinion did your parents or guardians have about labour unions? 30mm! FATHER ( ) IN FAVOUR OE UNIONS ') IN. FAVOUR 0F UNIONS ( ) NOT IN FAVOUR 0F UNIONS ( ) NOT IN FAVOUR or UNIONS ( ) NEITHER FOR NOR AGAINST UNIONS ( ) NEITHER FOR NOR AGAINST UNIONS ( ) ) NOT SURE WHAT aIS OPINION was THEN NOT SURE WHAT HER OPINION WAS THEN ( Have either of your parents or guardians ever been members of a labour union? m ELIE ( ) NO ( ) NO ( ) YES ( ) YES Approximately how many of your friends are members of a labour union (any labour union)? ()ALIDSTALLOITm ()mSTOFTm ()souzorrm ()ONLTAFEUOPTREM ()NOUEOPYEEH In general, what opinion do most of your friends hold of labour unions? THEY ARE IN FAVOUR OP UNIONS THEY ARE NOT IN FAVOUR OF UNIONS THEY ARE NEITHER IN FAVOUR NOR AGAINST UNIONS I DON'T mow WHAT MST OF H! FRIENDS THINK ABOUT UNIONS AAA“ Below and on page 4 are several statements dealing with labour unions and activities occasionally associated with them. We would like you to indicate the degree to which you agree or disagree with each statement. Please be honest; your answers will be kept in the strictest confidence. TO WHAT mm DO YOU AGREE ORDISAGREETEAI...... 1’1"" ‘1“:1' ’°“' M a) Unions are a positive force in Cansda........ 2.: g“! g 3:“ m a b) Strikes are too powerful in resolving sun-a m m m man man cmcr‘c: “‘put......00.00.000.000...00.0.0000... m m m ”M M m c) It bothers me when I see unions preventing people and vehicles from entering a struck plant. 2:." :2“ mm .‘mm imam: gum”. d) Workers who don't organize 9111 b. V011. m m sum snag m m Off m :h. 1“: MCCOOOCOOOOOOOCOOOO0.0.0.000... m m a am at“ am e) I would be willing to boycott my enployer's service or product if the medaership voted name an: mar m m to do “seeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeseeeee m m um 'm M f) Unions are too powerful in our society........ 2:." 22' 3:2,!“ m ”fig. 11 ‘1'. 1265 More Statements we would like to have your opinion on.... T0 WHAT EXTENT DO YOU AGREE OR DISAGREE THAT ..... a) Unions can be an effective means of protecting employees.......................... b) I would urge other members of this union to participate in a strike.................... c) Unions usually are 525 able to ensure that management treats employees fairly............ d) I am proud to be a member of the labour m:mMCOOOOOOOOOOOOO......OOOOOOOO e) I would not hesitate to participate in a work slowdown with other union members........ f) I believe that striking unions deserve the full support of all other union members ma. ‘r“OOOOOOOOOIOOO......OOOOOOOOOOOOOIOO 3) we need more laws to limit the power of mo“ MOOOOOOOOOOO......OOCOOOOOOCOCOOOO h) Even if a wildcat strike was encouraged by the union executive, I would 32; take P‘nm1:....-......00............OOOOOOOOOO. i) The growth of unionism has made our amt“, .tmnrOOOCOOOOOOO......OOOOCOOOOOO j) The right to strike is a valuable part of labour-management negotiations................ k) Too many grievances are filed by unions ch.“ “”000000000I.........OOOOOOOOOOOOOOOOO l) Acts of violence are sometimes important and useful means of acheiving union goals..... m) Employees have better wages and working conditions when they are unionised............ n) Unions impose too many restrictions m m1°y.n.........OOOICOOOOOOOO0.00.00... o) If I had to choose, I probably would not be .mr Of my man: m............‘...... p) Government regulations should change to give labour more power in labour-management nmtutw000000000I.......OOOOOOOOOODOOOOOO. Q) Workers who participate in illegal work .mm Md b. fir.d............CCCCCOOOO r) My wages and benefits would probably be just as good if there were no unions in chi.conga-......OOOOOOOOOOOOOOOOIOOOOOOOOOO ( ? l e a s e i r l e 3 r a n s w e r ) name smemea sane: name summer nmme amma amen amen nuumes enamel Onnmma name semen: nnmr name semen: nmae amen ems: amen pummel pummel anneal name smemea name’ name seamen name name Aflfll amen bummer nnnmma pummel :nmaa smemes :nnme :nnme smemes ‘nmms amen name umma unnmms pummel nnnmea name seamen name name' seamen name amen «mes amen mammal unneue avenue name summer :name name smemes :nmms ales umes amen enemas mammal mammal name summer name' name smemna name umes amen name enamel nausea messes :name amemea name name smemee name amen Aflll tame enamel acumen enemas :nmee smemea snnme :nnme summer :nmmn «mes all: umml inmeme mammal panama name seamen name’ name’ seamen name umua umea amen Inseam: mummea mammal :name seamen name' name summer name amen umea ems: manner mammal mammal name umemes name name smemea name emu: uals amen mummea nausea anneal name seamen name namn' seamen name same name name mum-ea mum-ma manner name smemea name name smemea name same name amen mammal maumea mammal name smemea ‘name name ~memues name uses umee amen mum-ms mammal messes :nmee smemna name name summer name use: «me: name tunnels mmumua onumma :name smemea name name seamen name all! amen «mes mnumes one-us pummel name smemea name name ~ smamne name amen Mill pummel summer anemia In. 15. Below, we want you to evaluate the performance of _y..._,_ oy indicating how satisfied or dissatisfied .127 you are with the union in each of the areas listed. How SATISFIgp ARE YOU WITH -..-.-.. IN TERMS OF ..... a) Handling members' grievances. ............ ... b) Providing feedback to members... ............ c) Getting better wages for members.... ..... ... d) Being a watchdog over management on a urcrd.’ but.........OOIOOOCOCCOOOO......IO. .) Maintaining job security.................... f) Getting management to improve the work 1c“1£...........0......OOOOOOOOOO0.0.0.0...... g) Listening to its members.................... h) Arranging social events..................... 1) Allowing members to have some say in the policies of the union local.................... 1) Getting better fringe benefits.............. k) Generally responding to the needs of the mfluPCOOOOIOOOOOOI......OOOOOOOOOOOOOOO... Please indicate the extent to which you either agree or disagree with each of the statements as they relate to _ _ _ - _ and its members. TO WHAT EXTENT DO YOU AGREE OR DISAGREE THAT..... a) This union local treats men and women u«al.00000000000O000000000......OOOOOOOOOOC b) Members of this union local are reluctant to vote for womem.in union elections........... c) Men and women are given equal opportunity to participate in this union local............. d) If I ran for union president and lost, one reason would be because I am a man/woman....... e) This union local is more interested in helping the women it represents than the men... f) When appointing people to committees, this union local does 225 discriminate against either women or men.................... g) In this union local, the membership encourages men to participate more than it menu's“ MOOOOOOOO......OCIOOOOOOOOOO ( P l e a s e c i r c l e y o u r a n s w e r ) vme summer .nnmr .name summer vme .nnn. amen urn. unmenn Hanan. Innnrn. vme summer name name summer nmr menu nan. nan. mnmens onnuei. maumn. nu! summer sane: name summer vme nun. man. nnn. uamun. unmen. mnment vme summer Tuner name summer me! man. nent menu ”menu. nnnmen unmenh nmr summer name name summer vme nan. nan. men. unmen. znnmnu. mnmen. vme summer lunar name' summer nee nan. menu nun. mnmenn unmenh Innenu. vme summer name name semen: mar nan. menu nan. memen. nnmen. Onaenn vme summer :unmr name summer vme nmn. nun. urn. mnmenu namun. nnnnen vme memem: name name semen: vme nun. menu nun. nnmenn unumn. unnunn vme summer name name: summer nmr nun. nnn. nun. nneenu unmana. enmen. vme summer :nnmr :nnmr summer nu: 'nnn. nuns suns. annenr. mament Imnnnn. ( P l e a s e 2: 3E a: c i r c l e name: m m A” naorr an mar Om your following a n s w e r ) 3m 17s 128 How willing would you be to participate in - .. - .. _ in each of the following ways? Please place a check mark (V) in the box above the appropriate number. Remember, your answer will be kept in the strictest confidence. sow stigma WOULD YOU BE ro.... a) Attend at least si_x NOT AI general mesmership ALL I 1 1 l L l I J VERY meetings per year. WILLING 1 2 3 4 S 6 7 WILLING b) Serve as an NOT AI VERY appointed mder on ALL I l J l l l J _J WILLING a union co-ittee? WILLING l 2 3 a 5 6 7 c) Run as a candidate in a union election NOT AI VERY for a position on the ALL I l J I l I I J WILLING executive board? SELLING 1 2 3 4 S 6 7 3. JOB ATIITUDES AND BEHAVIOUR Below are several adjectives which may or may not describe Eur present 10 . If the word or phrase does describe your job, circle YES; if. it doesn.t. circle NO. If you are not sure. circle the question mark (2). Please answer all of the adjectives listed. 1L IS you; won. . . . le. circle) I) ARE YOUR (IO-WORKERS. . . . (P1933; circle) Pascinating............. YES NO ? Stimulating”.................. YES NO Routine................. YES NO ? Boring............ ..... ........ YES NO Satisfying.............. YES NO 2 Slow.. YES NO Boring YES NO ? Ambitious YES NO Good YES NO ? Stupid YES NO Creative................ YES NO ? Eesponsible.................... YES NO Respected............... YES NO ? East YES NC Set YES NO ? Intelligent.................... YES NO Pleasant................ YES NO ? Easy tomakeenemies...........YES NO Useful....... ..... YES NO ? Talk too mch YES NO Iiresome................ YES NO ? Smart YES NO Eealthful............... YES NO ? Lasy YES NO Challenging............. YES NO ? Unpleasant YES NO Onyour feet............ YES NO ? No privacy........ ...... YES NO Frustrating..............YES NO ? Active. YES NO Simple. YES NO ? Narrow interests............... YES NO Endless........... ...... YES No ? Loyal YES NO Givesasense of Hard to meet YES so accomplishmt.......... YES NO ? ? ? 129 18. As in the previous question. please indicate if the following adjectives are descriptive of your present job. The statements below deal with your pay and promotional opportunities. WOULD YOU DESCRIBE HOW ABOUT YOUR. YOUR PAY AS.... PROMOTIONAL JPPORTUNI‘IIES... (Please circle) (Please circle) Income adequate for Good opportunity for normal expenses. ............... YES NO ? advancement. ....... . ....... YES NO .7 Less thanIdeserve............ YES NO ? Opportunity somewhat m m No ? lmt“0000000000000 ....... YES NO ? Promotions based on 7 ““1, pdd.OOOCOOOOIQIIOOOOOOO us No a “1110.................... as No ? , Satisfactory profit sharing.... YES NO . I: l-end-job............... YES NO ? , Barely enough to live on....... YES NO . Good I c. for p tion.. YES NO ? 9 MOOOOOOOOOOIIOOOOO0.0.0.0.... YES No C Unf‘ir Frantic“. ”11”... YES NO ? , Income provides luxuries....... YES NO . In! t p tions...... YES NO ? 7 . md.wdd.0000000000000 ..... 0.. m N0 0 h Jr P cl ......... as No ? Fairly good chance for _ PrmcionOOOOOOOOIOOOIOOIOO YES NO ? 19. Here are a few statements reflecting how some people view their jobs. Please indicate how strongly you agree or disagree that each statement reflects your attitude toward your job with the II I I I I I I I I I I I I. rowwmmnorou 91 1 1 ) 'acattonntsasaztrm.... —-(——“" ° '° ‘JW" ““w" a) My job is one of the major satisfactions ml, 11:.eeeeeseeeeeeeeeseeeeeeeeeeeeeeseeeee a m m m hm BM 5) There are many things that are more can mmnarr namr new man eases important tomethsnuyjob................... “'3 “'3 m M armour nears c)Hyjobisonlyasnllparto£whoIam....g :‘3 g 33:: m 33:“ em d) Overall, I am very much involved in my job. m =2“! 3:3 a m a e) I would not want my job to be a central m m nape sum: man mum part ofmylife..'............................. mar arr aura annals rem bum t) I live, eat, and breathe my job............ same man name m m same amsr mars arr beams: rem nears g) The most important things that happen to I. m mum " “'01"nJabeeeeeeeeesse‘eeeeeeeeeeeeeeeee % ”a m m im 9““ 20. What is the probability that you will guit your job for whatever reason with theI I I I . I I I I I I I I I within the next two years? ( ) 1002; I AM ABSOLUTELY OMAR! THAT I £14; 3! QUI‘ITING ( ) 801 ( ) 602 ( ) 602 ( ) 20: ( ) 01; I AM ABSOLUTELY cmam THAT I WILL NOT BE QUI'ITING 130 21. For each of the following adjectives. please indicate if the word or phrase is descriptive of your present ‘ob. The statements below deal with supervision. IS YOUR SUPERVISOR... (Please circle) IS YOUR SUPERVISOR... (Please circle) Intelligent. ............... YES NO ? Quick-tempered ............ .. YES NO ? Hard to please....... ..... . YES NO ? Tactful............. ........ YES NO ? Impolite ....... ....... .. YES NO ? Tells me where I stand. ..... YES NO ? Asksmyadvice............. YES NO ? Annoying. YES NO ? Praises good work.. ..... YES NO ? Stubborn.................... YES NO 7 Influential................ YES NO ? Knows job well.............. YES NO ? Up-to-date YES NO ? Bad... YES NO ? Doesn't supervise enough... YES NO ? Leavesmeonmyown......... YES NO ? Around when needed......... YES, so 2 Lexy YES so 2 22. How important to you is it that you eventually hold a mags-ent job? ()I‘IISVERYIHPOMTOHITOEVMUALLYBEINW ()I‘IISWMYDQOMIOHBIOWIBRINW ( ) IT IS ONLY SLIGHTLY'IHPOR‘IAN'I 1'0 HETO murmur BE IN mm ()I‘IISNO‘IAIALLDIPORIMTOHITOMALLYBIINW 23. Over the last couple of pages. we asked you several questions about your job. How we would like to learn more about your attitudes toward the Corporation as an employer. Please indicate how strongly you agree or disagree with each of the following statements. TOMWDOYOU moapymmr.“ Please circle your answer) I a) I feel a sense of pride working for th“ cmn’OOOOOOOOOOOCOOOO0.000.000......... w w w a m um 1’) “7 values I114 UN 60“!!!" values are same menu naar near our name EV." .ml‘tOOIOCO.......COOOOOOOOIIOOCIOO m m m am am om c) This company's problems are my problems.... g :31 fat-fl a roller “'0" d)Hyloyaltyistomywork,_no_ttothis m Issuer nasar nan: summer mums cmmYOOOOOC......OCOOOOOOOOOO......OIOIOOCCO m m m um um b) I am willing to put a great deal of effort beyond that normally expected of an “PM” in "3“: to make “‘1‘ “"9“” mm. more m m Inner new 'ucu.fueeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee a m m m um um f) I feel little loyalty to this cowany...... an in. g :32. m a C. OTHER WHO" 2“- Are you male or female? ()PEALE ()HALE 25. What is the highest level of education that you have had? 131 ) GRADE EIGHT OR LESS ) SOME HIGH SCHOOL ) COMPLETED HIGH SCHOOL ) COMMUNITY COLLEGE (whether or not completed) ) ) ) GRADUATE COURSES AT UNIVERSITY . OTHER EDUCATION (Please specify: A A A A f‘\ A f"\ UNDERGRADUATE UNIVERSITY (whether or not completed) 26. Do you have any children living at home? 27. Below is a list of household tasks. Host of us are held responsible for doing at least part of we would like you to estimate the percentage of each task that you are If you live alone, you would probably circle 1001 for If other members of the household do a particular task more than you, you some of these chores. responsible for doing where you live. most of the tasks. ( ) NO (If YES)' ( ) l: 2: 3 5:_________ 6: 7 ‘- Please write in the ggg_of each child living at home with you. Begin with the age of the youngest child. would circle 40!, 202. or possibly 02. WENT PERCENTAGE OF THE FOLLOUING DO YOU DO... a) b) c) d) a) f) S) h) i) J) 28. Overall. about how many hours ygr week do you spend doing household chores? Gathering/taking out the garbage............ 1002 Wishing/drying the laundry.................. 1002 Buying the groceries........................ 1002 Washing and drying the dishes............... 1001 Gardening/shovelling snow/mowing the lawn... 1002 Buying liquor/hardware/sundry items......... 1002 Cleaning the bathroom/kitchen............... 1002 Vacuuming/sweeping/dusting floors 8 furniture 1002 Handling bills/going to the bank............ 100! Preparing/serving meals and snacks.......... 1002 HOURS PER WEEK DOING HOUSEHOLD CEORES P l e a s e c i r‘c l e y o u r agg_g_g_g_y_) 802 60: 402 201 02 802 602 602 20: 02 802 602 602 202 02 801 602 402 202 O: 801 602 401 20! 02 801 602 601 20! 02 802 602 402 202 02 802 602 602 202 02 802 602 402 202 02 802 602 402 20% OZ 132 29. That is your marital status? If ) MARRIED AND LIVING WITH SPOUSE QR 'LIVING TOGETHER' ( ) NEVER BEEN MARRIED (and not currently 'living together ( ) SEPARATED. UIDOWED. DIVORCED ') (If MARRIED or LIVING TOGETHER) a) Has your spouse ever been a union member? ()NO ()rts b) What opinion does your spouse have of labour unions? ) IN FAVOUR OP UNIONS ) NOT IN FAVOUR OF UNIONS ) NEITHER FOR NOR AGAINST UNIONS ) NOT SURE "RAT OPINION MY SPOUSE HAS 0! LABOUR UNIONS c) STRONGLY ENCOURAOES IT SW ENCOURAGES IT SWEAT DISCOURAGES IT STRONGLY DISCOURAGES IT DOESN'T SEEM TO EVE AN OPINION ONE HAY OR THE OTHER t is the employment status of your spouse? does your spouse feel about you becoming very active in this union? CURRENTLY DIPLOYED OR SELF-EMPLOYED > TMORARILY LAID OH LOOKING FOR NOR! About how many hours per week does he/she work? HOURS PER WEEK RETIRED EULLTIMEBONEMAKER vvvvv VVVVV e) The statements presented below are on the subject of marriage and the family. the degree to which you personally agree or disagree with each. TO WHAT mm TO YOU AGREE OR DISAGREE THAT... 1) Parents should spend most of their free (Please time with their children g :3 ii) Husbands and wives are expected to be mm:- mm" with “Ch 0th.: too MOOOOOOOOOOOOOOOOOOOOI. m m iii) My greatest satisfaction is spending :1. “ch m.wu.0000000......QOOOOOOOIOOOO. : m iv) I often feel that I ought to be at home rem I WNIYOPOMOLIIE honeeseeeeeeseseeeeeeeee m m v) A strong family spends most of its c1. cog.th.r000.0.0.0.0......OOOOOOOOOOOOOIOO g m vi) People should spend as much time away from m m their spouses as with them in a good marriage. m an circle ’0 naar sears Please indicate answer) m seams ewes m same pram nears 13:3 In this final question. please rate yourself on a scale from one (1) to ten (10) in each of the following areas. Place a check mark (V) in the box above the appropriate number. a) As a member of, _ - _-_, how active are you in union business? I as ONE OF I AM ONE OF TEE LEAST THE MOST ACTIVE stuntns I I, I I I I I I I I l ACTIVE stanzas IN THIS UNION ' l 2 3 a 5 6 7 a 9 Io IN THIS UNION b) Overall, you militant are you as a union member? I as our or TEE I as ONE or TEE LEAST MILITANT MOST MILITANT UNION sassrns I, I I; I I I I I L I 1 UNION stanzas IN THIS ONION l 2 3 4 s 6 7 8 9 10 IN THIS UNION c) In general, how greatly do you value unionism? I VALUE UNIONISM A VALUE UNIONISM MUCH LESS THAN INCH MORE THAN 01m mans I L I I I I I I I | I ...—W mans IN TRIS UNION l 2 3 4 5 6 7 8 9 10 IN THIS UNION Thank you for taking the time to complete this questiomaire. Below, we have left space for you to add any coal-ants you would like to make about any of the issues raised in this booklet or about the questionnaire itself. You may be assured that your consents will be read and taken into account. IMPORTANT ll! HAVE YOU COMPLETED ALL OF THE QUESTIONS (EXCEPT THOSE YOU WERE INSTRUCTED TO SKIP)? FLEASEDONOTSIG‘YOURNAMEANYWREREINTRIS BOOKLET. SEALTEEBOOKLETINTNE ATTACHED ENVELOPE AND sum IT THROUGH INTERDEPARTMENTAL MAIL TO THE UNION OFFICE. _iTIAiK YOU VERY—Ml-fn FOR CONTRIBUTING TO THIS STUDY. BIBLIOGRAPHY BIBLIOGRAPHY Abicht, M.M. Women's leadership roles in two selected labor unions in the United States and Belgium: A comparative, descriptive study. Dissertations Abstract International, 1976. Allen, R.E., & Keaveny, T.J. Correlates of university faculty in- terest in unionization: A replication and extension. Journal of Applied Psychology, 1981, 99, 582-588. Anderson, J.C. A comparative analysis of local union democracy. Industrial Relations, 1978, 11, 278-295. Anderson, J.C. Local union participation: A re-examination. Industrial Relations, 1979,.18, 18-31. Anderson, J.C., & Milkovich, G.T. Propensity to leave: A prelim- inary examination of March and Simon's model. Relations Industrielles,_1980, gg, 279-292. Andiappan, P., & Chaison, G.N. The emerging role of women in the national union governance: The results of a Canadian study. Paper presented to the International Industrial Relations Association, Sixth World Congress, March 1983. Berheide, C.W., Berk, S.F., & Berk, R.A. Household work in the suburbs. Pacific Sociological Review, 1976, 12, 491-517. Bigoness, W.J. Correlates of faculty attitudes toward collective bargaining. Journal of Applied Psychology, 1978, 9;, 228-233. Brayfield, A.H., & Crockett, W.H. Employee attitudes and employee performance. ngchological Bulletin, 1955, 22, 396-424. Cook, A.H. Women and American trade unions. The Annals, 1968, 375, 124-132. Cook, J.D., Hepworth, S.J., Wall, T.D., & Warr, P.B. The experience of work: A compendium.and review of 249 measures and their use. London: Academic Press, 1981. Cronbach, L.J., & Meehl, P.E. Construct validity in psychological tests. Psychological Bulletin, 1955,‘§2, 281-302. 134 135 Dean, L. Interaction, reported and Observed: The case of one local union. Human Organization, 1958, 18, 36-44. Dean, L.R. Social integration, attitudes, and union activity. Industrial and Labor Relations Review, 1954, 8, 48-58. DeCotiis, T.A., & Le Louarn, J.Y. A predictive study of voting behavior in a representation election using union instrumen- tality and work perceptions. Organizational Behavior and Human Performance, 1981, 21, 103-118. Dillman, D.A. Mail and telephone survgys: The total design method. New York: John Wiley & Sons, 1978. Divers, A.J. An analysis of the relationship Of teacher job satis- faction and teacher union activism in the Detroit public school system. Dissertation Abstracts International, 1981,‘fll, 3795-A. Dolan, S. Determinants of officers' militancy: The case of the national union of teachers in Israel. Relations Industrielles, 1979, 33, 287-311. Dreher, G.F., & Mai-Dalton, R.R. A note on the internal consis- tency of the Manifest Needs Questionnaire. Journal of Applied Psychology, 1983, Q8, 194-196. Dunlop, J.T. Industrial relations systems. New York: Holt, Rinehart, & Winston, 1958. Economic Council of Canada. People andgjpbs. Ottawa: Minister of Supply and Services Canada, 1976. Flanders, A. Management and Unions. London: Faber & Faber, 1970. Form, W.H., & Dansereau, R.R. Union member orientations and pat- terns Of social integration. Industrial and Labor Relations Review, 1957, 11, 1-12. Geare, A.J., Herd, J.J., & Howells, J.M. Women in trade unions: A case study of participation in New Zealand. Wellington, N.Z.: Victoria University of Wellington, 1979. Geoffroy, R. & Sainte-Marie, P. Attitude of union members to women in industry. Study #9 for the Royal Commission on the Status of Women in Canada. Ottawa: Minister Of Supply and Services Canada, 1971. Getman, J.C., Goldberg, S.B., & Herman, J.B. Union representation elections: Law and reality. New York: Russell Sage Foundation, 1976. 136 Glassberg, E., Baden, N., & Gerstel, K. Absent from the agenda: A report on the role of women in American unions. Coalition of Labor Union Women, September 1980, mimeo. Glick, W., Mirvis, P., & Harder, D. Union satisfaction and partic- ipation. Industrial Relations, 1977, $9, 145-151. Gordon, M.E., Philpot, J.W., Burt, R.E., Thompson, C.A., & Spiller, W.E. Commitment to the union: Development of a measure and an examination of its correlates. Journal of Applied Psychology (Mbnoggaph), 1980, fig, 479-499. Hagburg, E.C. Correlates of organizational participation: An ex- amination of factors affecting union membership activity. Pacific Sociological Review, 1966,.2, 15-21. Hagburg, E.C., & Blaine, H.R. Union participation: A research note on the development of a scale. Industrial and Labor Relations Review, 1967, 21, 92-95. Hellriegel, D., French, W., & Peterson, R.B. Collective negotia- tions and teachers: A behavioral analysis. Industrial and Labor Relations Review, 1970, 3;, 380-396. Harrison, M. Participation of women in trade union activities: Some research findings and comments. Industrial Relations Journal, 1979, 19, 41-55. Herman, J.B. Are situational contingencies limiting job attitude - job performance relationships? Organizational Behavior and Human Performance, 1973, $95 208-224. Huszczo, G.E. The relative importance of variables related to participation in union activities. Unpublished Ph.D. dis- sertation, Michigan State University, 1975. Kochan, T.A. Contemporary views of American workers toward trade unions. Washington, D.C.: Assistant Secretary for Policy, Evaluation, and Research (Labor), U.S. Department of Commerce, September 1978. Kochan, T.A., LipSky, D., & Dyer, L. Collective bargaining and the quality of work: Reflections of union activists. ‘Egg- ceedings of the 27th Annual Meetingyof the Industrial Rela- tions Research Association. Madison, Wisconsin: IRRA, 1974, 150-162. Kowachyk, R.S. Local teacher organizations: The relationship between teachers' perceptions of their union and job satis- faction. Dissertation Abstracts International, 1982, 33, 323A. 137 Koziara, R.S., & Pierson, D.A. Barriers to women becoming union leaders. Proceedings of the 33rd Annual Meeting of the Industrial Relations Research Association. Madison, Wisconsin: IRRA, 1980, 48-54. Kyllonen, T.E. Social characteristics of active unionists. American Journal of Sociology, 1951,.26, 528-533. Lawler, E.E. III, 8 Hall, D.T. Relationship of job characteristics to job involvement, satisfaction, and intrinsic motivation. Journal of Applied Psychology, 1970, 53, 305-312. Lipset, S.M., Trow, M.A., Coleman, J.S. Union democracy: The inter- nal politics of the International Typographical Union. Glenoc, Illinois: Free Press, 1956. Lodahl, T.M., & Kejner, M. The definition and measurement of job in- volvement. Journal of Applied Psychology, 1965, 42, 24-33. Marchak, P. Women workers and white-collar unions. Canadian Review of Sociology and Anthropology, 1973,‘lg, 134-147. Maxey, C., & Mohrman, S.A. Worker attitudes toward unions: A study integrating industrial relations and organizational behavior per- spectives. Proceedings of the Thirty-third Annual Meeting. Wisconsin: Industrial Relations Research Association, 1980. McShane, S.L. Participation of women members in local labour organi- zations: An empirical test of hypothesized barriers - Results of a_pretest analysis. Unpublished report prepared for the Ontario Ministry of Labour, Toronto, January 1982. McShane, S.L. The socialization of attitudes toward labor unions. Paper presented at the 20th Annual Conference of the Canadian Industrial Relations Association, June 1983. McShane, S.L. Women in unions: Their numbers are growing. Worklife, 1981,‘2, 10-11. Meissner, M., Humphreys, E.W., Meis, S.M., & Scheu, W.J. No exit for wives: Sexual division of labour and the cumulation of household demands. Canadian Review of Sociology and Anthropology, 1975, 13, 424-439. Miles, R.H., & Perreault, W.D. Organizational role conflict: Its antecedents and consequences. Organizational Behaviour and Human Performance, 1976,.11, 19-44. Miller, G.W., & Young, J.B. Membership participation in the trade union local. American Journal of Economics and Sociology, 1955, .15, 36-45. 138 Mowday, R.T., Porter, L.W., & Steers, R.M. Employee-organization linkages: Theypsychology of commitment, absenteeism, and turn- over. New York: Academic Press, 1982. Nelson, J.I., & Grams, R. Worker interaction in occupational com- munities. Rural Sociology, 1978, 43, 265-279. Nicholson, N. Mythology, theory and research on union democracy. Industrial Relations Journal, 1978, 3, 32-41. Nicholson, N., Ursell, C., & Blyton, P. The dynamics of white collar unionism. London: Academic Press, 1980. Ontario. Royal Commission on the Status of Pensions in Ontario. Design for retirement. Toronto: Queen's Printer, 1980. Perline, M.M., & Lorenz, V.R. Factors influencing member participa- tion in trade union activities. American Journal of Economics and Sociology, 1970, 22, 425-438. Peter, J.P. Construct validity: A review of basic issues and mar- keting practices. Journal of Marketing Research, 1981,.18, 133-145. Portwood, J., Pierson, D., & Schmidt, 3. When less is more: A study of the participation/satisfaction relationship in unions. In K.H. Chung (Ed.), Academy of Mansgement Proceedings, Forty-first annual meeting, San Diego, California, August 1981. Purcell, T.V. The worker speaks his mind on company and union. Cambridge, Mass.: Harvard University Press, 1954. Rabinowitz, S., & Hall, D.T. Organizational research on job involve- ment. Psychological Bulletin, 1977, 84, 265-288. Ramaswamy, E.A. The participatory dimension of trade union democracy: A comparative sociological view. Sociology, 1977, 11, 465-480. Rose, A. Union solidarity -- The internal cohesion of a labor union. Minneapolis: University of Minnesota Press, 1952. Rosen, H., & Rosen, R.A. The union member speaks. New York: Prentice- Hall, 1955. . Saleh, S.D., & Hosek, J. Job involvement: Concepts and measurements. Academy of Management Journal, 1976, 12, 213-224. Sayles, L.R., & Strauss, G. The local union: Its place in the indus- trialyplant. New York: Harper & Brothers, 1953. Schriesheim, C.A. Job satisfaction, attitudes toward unions, and voting in a union representation election. Journal of Applied Psychology, 1978, 6;, 548-552. ‘ "illllllllll