Content uploaded by Patrick C. Flood
Author content
All content in this area was uploaded by Patrick C. Flood on Oct 14, 2017
Content may be subject to copyright.
Journal of
Occupational
and Organizational
Psychology
(1993), 66, 213-223 Printed in Great Britain
213
© 1993 The British Psychological Society
An expectancy value analysis of the willingness
to attend union meetings
Patrick Flood*
Department
of
Business
Studies,
University of Limerick,
Department
of
Management
and Centre for Human
Resource
Management,
Limerick, Ireland
This paper reports the findings of
a
study of the willingness to attend union meetings
using the expectancy valence motivation framework of analysis
at
the individual level.
A goal motive application of Klandermans'
(\9SAa,b,c)
model of union meeting atten-
dance was tested on a sample of blue-collar union members using a range of attitudinal
measures. The willingness to participate was found
to
be strongly related
to
both goal
motives and role-related characteristics. The expectancy valence framework is shown
to
be relevant
for
procedural membership meetings. Pessimistic estimates
of
the likely
participation of
lay
members was found
to
increase the importance for office holders of
their own attendance.
Why do members engage
in
trade union affairs.'
This question
has
attracted both public
and
academic attention
for a
number
of
years.
Various studies (Hemingway, 1978; James, 1984) have described the effects
of
member-
ship involvement
on
internal union democracy while other investigators have focused
upon
the
determinants
of
involvement (Klandermans, 1986). Membership involvement
itself encompasses a wide range of
behaviours:
voting
in
union elections, holding office as
a union representative either paid or unpaid, attending union meetings both
at
workplace
and branch level, speaking
at
union meetings, seeking information about union affairs,
reading
the
union newspaper, raising grievances
and
participating
in
industrial action
which
can
include strikes, both official
and
unofficial, overtime bans, work
to
rules,
go
slows and so on (Anderson, 1979; McShane, 1986). The concept of membership involve-
ment
has
been operationalized
in
various studies
as
prospective participation (Blyton,
Nicholson & Ursell, 1981), actual participation (Klandermans, 1984c) and retrospective
participation (Nicholson, Ursell & Blyton, 1981). While a plethora of correlates of union
involvement
is to be
found
in the
literature (Huszczo, 1983), there have been very
few
attempts
to
classify
or
test them. As Klandermans (1986, p. 199) points out: 'from a psy-
chological viewpoint, most research
on
union participation
is
based implicitly
on
one
or
other
of
three theoretical approaches: frustration—aggression theory, rational choice
the-
ory or interactionist theory. The first approach points to dissatisfaction as the cause of par-
ticipation, the second
to
the perceived costs and benefits of participation and the third
to
the social context
of
individuals'. While there have been studies focusing
on
the frustra-
tion-aggression theory (Blyton et al., 1981) and
the
interactionist framework (Batstone,
*Requests for reprints.
214 Patrick
Flood
Boraston & Frenkel, 1977; James, 1984), there have been remarkably few studies which
systematically apply the expectancy framework to the case of union participation with the
exception of Klandermans (I984a,b,c; 1986). The aim of this paper is to re-examine the
efficacy ofthe expectancy framework in the case ofthe willingness to attend union meet-
ings,
apply it to the case of
a
routine procedural member meeting and partially test it on
a blue-collar sample of union members. As Klandermans (1984^^, p. 120) points out:
'research in which the willingness to take action is studied comparatively under different
circumstances, at different times and places, would be extremely valuable'.
The expectancy framework applied to the intention to attend union meetings
The willingness to attend union meetings according to Klandermans (1984«) is a func-
tion ofthe following elements:
Goal motives: (a) The importance that a person attaches to influencing trade union
policy.
(b) The expectancy that a member meeting can influence this policy.
(c) The expectancy that his/her presence will contribute to the success of
the meeting.
(d) His/her expectancy that others will attend as well.
Social motives: The expected reactions of family members, colleagues and supervisors
and the importance ofthese reactions.
Reward motives: The rewards and costs that a person associates with meeting attendance
and their importance.
According to Klandermans (1984«), willingness to participate is a necessary but
insufficient condition for union meeting attendance to occur. Two other factors—
(i) familiarity with the time and place ofthe meeting (this is often role-related in the sense
that such information is often transmitted via an informal network of activists at plant
level which can mean that ordinary members do not uniformly receive such information);
and (ii) the ability to attend—are highly important additional factors.
Combinations of the goal, reward and social motives occur depending on whether the
meeting is of
a
procedurally routine (i.e. regularly occurring) or action-orientated nature
(Klandermans, 1984<?). Procedural meetings are routine in the sense that they are not
arranged to formulate a programme of action. According to Klandermans (1984<«), the
goal motive is likely to be the most important motive in the case of
a
procedural member
meeting while reward and social motives will be uppermost in the decision to take action.
He also argues that the motives can therefore compensate for one another, i.e. depending
on the type of action or participation involved, one or other ofthe goal, reward or social
motives is likely to assume greater importance than the others.
Method
Subjects
Subjects were members ofthe Irish Transport and General Workers' Union (ITGWU), a geographically dis-
persed general union, whose area branch holds an annual general meeting (AGM) which is open to all of the
general (lay) membership in the region. The AGM is typically addressed by a prominent national officer of
Expectancy value analysis: Attendance at union meetings 215
the union. As part of a larger survey of union members in one ofthe geographically dispersed branches ofthe
ITGWU (having a combined membership of 1,400 members), it proved possible to include some questions
which allowed Klandermans' expectancy value model of union participation to be partially tested. As
expectancies are prospective in nature it was decided to focus attention on a forthcoming AGM ofthe entire
branch membership.
Familiarity with the time and place ofthe meeting
was
accomplished with the cooperation ofthe local full-
time officer of the union and shop stewards who posted notices in the various workplaces concerning the
AGM. Arguably, capacity to attend can be assumed, as the meeting was open to the general membership of
the branch (the monthly branch meetings
are
only open to the branch delegates who are normally elected shop
stewards) and the meeting was held on a Sunday at noon which would allow the majority of members to
attend if they so wished.'
Design and procedure
It was originally the author's intention to test a full goal, reward and social motive application ofthe cost
benefit model. However, it was only possible in the circumstances (because ofthe fact that it was part of
a
wider survey of membership attitudes a decision was made to keep the length ofthe questionnaire as short as
possible) to test a goal motive application ofthe expectancy model. The research is therefore limited by this
fact although Klandermans (1984a) does make the point that the goal motive is likely to be the most impor-
tant one in the case of
a
procedural member meeting such
as
the branch
AGM.
The survey questionnaires were
administered during working hours to all the union members in the respective 15 plants and were collected
on the following day. In total some 420 usable questionnaires representing a 33 per cent response rate on the
part ofthe members was achieved. Such response rates are typical in union member surveys (Nicholson
etal.,
1981) and subsequent comparisons with the larger branch membership in terms of sex and union tenure
revealed that a broadly representative sample was obtained.^
Dependent and independent variables
The dependent variable 'willingness to participate' was measured by the question: 'Are you going to attend
the branch AGM?' with three possible response categories: 'yes', 'not sure' and 'no'.
The independent variables included a measure ofthe valence ofthe expected outcome ofthe meeting, mea-
sures ofthe various expectancies involved and a position in the union variable which was used as a proxy for
familiarity with the time and place ofthe meeting which is, as discussed earlier, often a role-related variable.
This is particularly relevant in this study as the activists were responsible for posting up notices in the plants
concerning the time and venue ofthe AGM. A measure ofthe ability to attend the meeting was not included
' It could of course be argued that union members (particularly mothers) with children might find it more difficult to
attend a Sunday afternoon meeting due for example to babysitting difficulties. To test this, a regression analysis was car-
ried out to find out if sex-role differences and parental responsibilities might effect willingness to attend. However, no
significant effects were found. In aggregate, the sample was 41 per cent male, 61 per cent ofthe sample were married and
45 per cent ofthe sample had no children.
^
The branch population of union members (N = 1400) from which the sample was drawn was 67 per cent male while the
sample was 59 per cent male. The union tenure distribution ofthe branch population of union members when compared
with the sample revealed the following:
Union tenure Branch population (%) Sample (%)
0-5 years 23.7 37
6-11 years 32.9 32
12+years 43.4 31
It can be seen that members with less than
five
years' membership
are
over-represented while those with more than 12 years'
membership are under-represented in the sample relative to the population. This level of bias, however, does not
significantly affect the multivariate analyses reported in the paper. For mulcivariate analysis purposes the most important
factor is to obtain a broad spread on the dependent and independent variables which was obtained as can be seen from
Table 2. The branch population was calculated from the branch registet of union members.
216 Patrick
Flood
as it was assumed that the majority of members could attend the meeting if they wished which was held on
a Sunday and therefore was unlikely to interfere with their normal work schedule (see footnote 1). At least
one major obstacle, namely normal working, was removed for the majority of members.
Valence was measured by the question: 'How important is it to you to have a say in your union affairs?'(i.e.
influencing union affairs is viewed as the desired outcome of involvement) with three possible response cate-
gories—'very important, 'not sure, 'not important.
The expectancy that the meeting would influence union policy was subdivided (Klandermans, 1984f) into
three probability estimates: PI—the expectancy that ones own participation would contribute to a
successful meeting; P2—the estimate of
the
number of participants; and P3—the expectancy that a meeting
which is successful will have an impact on union decision making. PI was measured by two questions geared
towards establishing the respondents' opinion of the importance of their attendance to the outcome of the
meeting. These two questions were: 'It doesn't matter whether I attend the branch AGM or not' and
'It is important that I attend because by going I will make sure my views are heard'. There were three
possible response categories to these questions—'agree', 'not sure' and 'disagree'. P2 was measured by two
questions: (i) 'Would a good turnout of members be an advantage or a disadvantage in influencing union
policy?' which had a three-point response scale—'advantageous', 'not sure' and 'disadvantageous'; and (ii)
'How many people do you think will attend the AGM?' which had three possible response categories (0-29
persons), (30—59 persons) and (60+ persons). P3 was measured by the question: 'Do you think the branch
AGM can have an effect on union policy?' with three possible response categories—'definite effect', 'might
effect' and 'no effect'.
Scoring procedures for the expectancy-value model
The model was scored according to Vroom's (1964) scheme: Valence refers to the extent to which the goal of
participating is valued. It varies from +1 to —1 representing maximally positive outcomes and maximally
negative outcomes respectively. Neutral outcomes are scored 0. Expectancies vary from 0 to +1 with the mid-
point intermediate category receiving the score 0.5. The effort-performance expectancy was subdivided into
two estimates concerning the importance of one's own personal attendance and an estimate of how many other
persons would attend.
Motivational force is the product of these expectancies and valences, i.e. the valence of each outcome is mul-
tiplied by the strength of the expectancy that the act will lead to the attainment of the outcome. The alge-
braic sum of the resulting products is then taken and the formula for motivational force, AI is expressed as
/M =2 (£ X V) where £ represents expectancy and V represents the valence of the outcome.' Item content
for the measures are given in Table
1
while Table 2 details the scoring procedure for each measure used in the
analysis.
Results
Table 1 outlines the item content and the descriptive results for the goal motive applica-
tion. Table 2 summarizes office holder-rank and file member differences on these vari-
ables.
The
willingness to
attend
the branch
AGM
Slightly over one quarter of the members professed to be willing to attend the Branch
AGM. A total of 24 per cent were unsure and 49 per cent were not willing to attend. A
naive observer might expect that a similar number would be expected to attend the actual
' An expanded version of this formula was used in order to reflect the distinction between effort-performance (E-f)
expectancies and performance-outcome (P-0) expectancies and valence
(V)-
The formula used to compute motivation (Al)
to attend was tA =S {(£-Pj X 2 [(P-OXVO]).
Expectancy value analysis: Attendance at union meetings 217
Table 1. Klandermans' expectancy value model: Variables and item content (goal motive
application) (N = 420)
Dependable
variable
Willingness to
participate
Independent
variables
(a) Valence
(outcome)
Effort-performance
expectancy
(b) Own
participation
(c)
(d) Numbet of
patticipants
(e)
Performance-
outcome
expectancy
(0 Impact of the
meeting
Item
content
'Ate you going to
attend the btanch AGM?'
'How important is it to
you to have a say in
yout union affairs?'
'It doesn't mattet whethet
I attend the btanch AGM
ot not.'
'It is important that I attend
because by going I will make
sure my views ate heard.'
'Would a good tutnout of
membets be an advantage
or a disadvantage in
influencing union policy?'
'How many people do you
think will attend the AGM?'
'Do you think tbe btanch
AGM can have an effect
Yes
27
Very
important
(%)
59
Agree
55
50
Advantage
62
0-29
71
Definite
effect
Not
sure
24
Not
sure
20
Not
sure
21
22
Not
sure
17
30-59
16
Might
effect
No
49
Not
important
21
Disagree
24
28
Disadvantage
21
60+
13
No
effect
on union policy?'
Familiarity with
time
and place
ofthe
meeting
Position in the union''
16
Office holder
(%)
15
5430
Lay
members
85
Note.
This table is subsequently referenced throughout this paper as Table
1
(a), (b), etc. Thus Table
1
(a) refers to the data
in Table 1, subsection (a).
" Position in the union is used as a proxy for familiarity with the time and place ofthe meeting.
218 Patrick Flood
meeting. However, willingness to attend only represents the fitst step in the motivation
arousal chain. Valences and expectancies must also be positive for behaviour to occur. The
discussion which follows centres on the goal motives as this was the section of
Klandermans' theory which was tested.
Goal motives
Valence.
The valence or value of attending the branch AGM was assumed to be the
objective of exerting influence over one's union affairs. As we can see from Table 1 (a),
almost 60 per cent ofthe membets placed a high value on this goal. However, just one
quartet ofthe membership expressed willingness to attend the AGM and less than half
of
that numbet actually did attend the meeting (ascertained from union tecotds). The detet-
minants of valences are worth commenting on here. One determinant of valence (Staw,
1977) is the set of individual needs which a person possesses. Need strength ot intensity
Table 2.
Variable
Expectancies:Rankandfile(N = 310)
Rank and
X
—
office
file
SD
holder (N =
Office
X
70) comparisons
holdet
SD
Willingness to attend AGM" 1.66 0.79 2.25 0.88
-5.18**
Value of goal motive* 0.35 0.81 0.83 0.44
-8.98*
Influence of
AGM''
0.40 0.32 0.49 0.37 -1.86*
Advantage of high tutnout^ 0.69 0.40 0.75 0.40 -1.12
Numbers expected at AGM' 0.48 0.39 0.47 0.38 0.34
Importance of
own
attendance^ 0.30 0.39 0.52 0.48 -3.56**
Importance of expressing own
views*
0.56 0.43 0.75 0.39
—3.63**
*p< .05;*>< .01.
" 1 if subject is willing to attend, 2 if subject is unsure, 3 if subject is willing to attend the AGM.
* +1 if subject places high value on having a say in union affairs, 0 if subject is unsure,
— 1
if subject places low value on
having a say in union affairs.
''
1
if subject believes AGM can definitely affect union policy; 0.5 if subject believes AGM might affect union policy, 0 if
subject believes AGM will have no effect on union policy.
''
1
if subject believes
a
good turnout is
a
definite advantage, 0.5 if subject is unsure, 0 if subject believes
a
good turnout is
a disadvantage.
' 0 if subject believes 0-29 persons will attend, 0.5 if subject believes 30-59 persons will attend and 1 if subject believes
60+ will attend.
•'^O
if subject agrees, 0.5 if subject is not sure, +1 if subject disagrees.
* 0 if subject disagrees, 0.5 if subject is not sure, +1 if subject agrees.
varies between individuals and is often role-related. To test the ptoposition that valence
would be tole-related, a subanalysis was carried out between office holdets (i.e. lay repre-
sentatives) and ordinary membets. The value ofthe goal motive (valence) was found to be
significantly highet (^ < .05) for office holders (x = 0.83, SD = 0.44) than for rank and
file membets (x= 0.35, SD = 0.81). One possible explanation which is consistent with
Expectancy value
analysis:
Attendance
at
union meetings
219
the literature on motivation to become a shop steward (Moore, 1980; Nicholson 1976) is
that office holders possess stronger needs for involvement than tank and file membets. As
a result, procedural member meetings are much more likely to provide an avenue for need
fulfilment fot representatives than fot ordinary membets.
Own participation
The figures in Table 1 show that while 50 per cent of the membets believe that they
should go to the meeting in order to make their views known (Table lc), over 55 per cent
do not believe that the marginal contribution of their own attendance is worthwhile
(Table lb). Studies by Lieberman (1956) and Oliver (1984) have shown that role occu-
pancy has a significant effect on individuals' perceptions and attitudes. It was therefore
decided to test the effects of role occupancy upon this variable and a subanalysis of office
holdets versus lay members was carried out. Office holders {x ~ 0-49, SD = 0.37) wete
found to possess a significantly higher (p < .05) expectation that their attendance at the
meeting would influence union policy when compared with lay membets {x ~ 0.40,
SD = 0.32). This finding is similar to the findings of Olivet (1-984) and Moore (1980)
(in relation to the belief system of union/community activists) that office holders are
ftequently propelled into theit role by a belief that if they do not become personally
involved, pertinent issues will not be brought to a satisfactoty conclusion by virtue of
apathy on the part of rank and file members.
Number of participants
While over 60 per cent ofthe membets (Table Id) believe that a good turnout of mem-
bets would be an advantage in influencing decision making, over 70 per cent (Table le)
believe that less than 30 members will attend the meeting. This would seem to have the
effect of altering their expectancies about the vety point of attending such meetings. Rank
and file members unaccustomed to attending such ptocedural meetings are likely to view
them as irrelevant and to hold pessimistic opinions about their likely effectiveness. Office
holders ix = 0.75, SD = 0.40) did not differ significantly from tank and file members {x
= 0.69, SD = 0.40) in relation to the belief that a good tutnout of members would be an
advantage in influencing union policy (i.e. both office and non-office holders viewed a
good turnout as having a potentially positive impact on union policy). Office holders
{](
— 0.47, SD = 0.38) did not diffet significantly from rank and file membets (x = 0.48,
SD = 0.39) in relation to the numbet of patticipants expected to attend the AGM—botb
held equally pessimistic estimates. Thus, while both office holders and rank and file mem-
bets hold similar pessimistic views about the expected numbet of patticipants, office
holdets believe that theit attendance will influence union policy when compared with the
views of rank and file members (see previous paragraph).
Familiarity with
the time
and
place
ofthe
meeting
The variable 'position in the union' was used as a proxy for familiarity with the time and
place of the meeting as office holders are accustomed to attending monthly procedural
meetings in the local union hall where the AGM was held and they were involved in pub-
220 Patrick
Flood
licizing the meeting. The willingness to attend the AGM was found to be significantly
higher (p < .01) for office holders {x — 2.25, SD = 0.88) than for ordinary members
iX = 1.66, SD = 0.79).
Multivariate analysis
In order to test how well each component of the goal motive predicted willingness to attend
the AGM, a regression analysis was carried out. Position in the union was used as a ptoxy
for familiarity with the time and place of the meeting and as a 'check' variable because of
the strong role-related effects observed in Table 2. The results are shown in Table 3. The
computed score using the formula M = S (ExV) (see footnote 3) was also used in a separate
regression analysis to test how well it predicted the willingness to attend the AGM. Position
in the union was also included as per the earlier procedure. The results of this analysis are
shown in Table 4. The position in the union variable (0 = rank and file member, 1 = office
holder) contributed just 1 per cent of the overall variation in the dependent variable.
It can be seen from the results of
Tables
3 and 4 that the regression equation using the
individual components of the goal motive plus position in the union (a proxy for famil-
iarity with the time and place of the meeting) explains some 35 per cent of the variance
in the willingness to attend the AGM. The computed motivation score plus position in
the union is less successful, explaining only 26 per cent of the variance in willingness to
attend the AGM. Focusing on the results of
Table
3 yields some interesting insights.
Table 3. Regression of the six components of the goal motive and position in the union
on the willingness to attend the AGM: Standardized beta coefficients
Zero-order correlation Beta
Value of
goal
motive
.35***
.19
Influence of
AGM
.29***
.11
Advantage of high turnout
.26***
.05
Numbers expected at AGM
—
.08*
Importance of
own
attendance
.43***
.16
Importance of expressing own views .49 .28
Officeholder
.26***
.12
*«*
***
***
***
*
l>
< .05;"*/) < .01.
Table 4. Regression of the computed motivation to attend the AGM and position in the
union on the willingness to attend the AGM: Standardized beta coefficients
Zero-order correlation Beta
Motivation to attend the AGM .34 .44
Officeholder
.38***
.18*
R^
= .26.
*p <
.05;
'"p
<
.01.
Expectancy value
analysis:
Attendance
at
union meetings
221
The most important predictors are the expectancy in relation to importance of express-
ing one's views, the value of having a say in one's union affairs, the expectancy in relation
to one's own attendance, the expected number of participants at the AGM and the impact
of the meeting. However, the variable which measures whether a good turnout of mem-
bers is likely to be an advantage or disadvantage in influencing union policy is statisti-
cally insignificant. The fact that the number of expected patticipants has a negative
impact on the willingness to attend the AGM poses some interesting questions. After all,
one would expect that people would be more interested in attending a meeting where a
large attendance might be expected. The answer to this dilemma lies in the expected
marginal contribution of one's own attendance which was found to be much higher (see
Table 2) for office holders when compared to rank and file members despite the fact that
both office holders and rank and file members hold equally pessimistic estimates (see
Table 2) about the likely number to attend the AGM.
Discussion and conclusion
If office holders are equally pessimistic about the expected number of participants, why
are they still willing to attend? One possibility is that office holders inwardly compensate
by adjusting upwards their own expected marginal contribution to the meeting. The
results of Tables 2 and 3 would support such a hypothesis. Furthermore, it is consistent
with the findings of Oliver (1984) in relation to the factors which motivate neighbour-
hood activists where she found that activists' pessimistic expectation of the prospects for
collective action often acted as the propelling force to become personally involved in
neighbourhood campaigns, i.e. activists believed strongly in the adage: 'if you don't do it,
nobody will'. Theoretically, four general factors (Oliver, 1984) underlie participation in
collective action in public goods situations (where free riding behaviour is a frequent
problem).
Firstly, larger contributions can be expected from people who place a higher value on
the collective good. Secondly, larger contributions can be expected from people who expe-
rience lower costs or enjoy higher rewards as a result of their contributions. Thirdly, larger
contributions can be expected when social ties are strong amongst group members. The
fourth factor relates to the difference between large and small groups. In large groups an
individual may expect that his/her contribution is going to be very small and unlikely to
affect outcomes in a rational way. However, in a small group such as the inner cadre of
union activists who maintain the fabric of
local
union organization (and who attend AGM
type meetings), individual contributions do make a difference and are noticeable. Local
activism, as Oliver (1984) points out, exhibits diminishing marginal returns, i.e. once the
job is being done at all there is little marginal payoff for lay members to participate and
they 'free ride', secure in the belief that the local activists will ensure that essential work
is carried out. The data in this paper support this explanation. The setting is a procedu-
ral member meeting where high attendances are not as essential as, for example, in situ-
ations where a strike ballot might be taken (provided the elected stewards attend the
meeting, the meeting can be expected to go ahead and the main business of the day con-
ducted, i.e. the conditions of diminishing marginal returns to participation are present).
In relation to the four factors cited in explaining why larger contributions can be
expected from office holders there is also supporting evidence. Firstly, the office holders
222 Patrick Flood
were found to place a high value on influencing union affairs. Secondly, they may experi-
ence lower costs than ordinary members in the sense that they have internalized the costs
involved over time as they are accustomed to attending such meetings. In addition, the
fact that the AGM is usually addressed by a prominent national union official may mean
that selective incentives operate. Being present at the meeting means an opportunity to
meet and impress someone who may appoint you to a union position in the future.
Thirdly, social ties are strong amongst the stewards who meet regularly with the local
union officer and fourthly, the small group condition is also met.
The results show that analysis at the individual level using the expectancy-valence
framework does help to explain why some categories of members are more likely to be
willing to participate in union affairs than others. It has been shown that the analysis
applies to routine procedural type meetings which are an essential if tiresome part of
union organizational functioning. It was also shown that a pessimistic estimate of the
likelihood of others participating increases the expected marginal contribution of
participation for office holders (activists). This tuns counter to Klandermans' (1984f)
formulation which would suggest that a pessimistic estimate of others' participation
will decrease the expected marginal contribution of participation of office holder and
member alike. The computation of the motivational force score can act to disguise this
effect. The old adage of'If you don't do it, nobody else will' would seem to be a powerful
factor in the belief system of trade union activists. Future research could usefully
incorporate goal, reward and social motives in a fuller test of Klandermans' model. The
analyses presented here explain slightly over one third ofthe variation in the willingness
to attend which suggests that reward and social motives would help to explain the
remaining deficit.
Acknowledgements
The author wishes to thank David Guest, John Kelly, Bert Klandermans, Sarah Moore and two anonymous
referees for helpful comments on an earlier version of this paper. The PhD thesis from which this paper is
derived was funded under The British Council Foreign and Commonwealth Office Scheme,
References
Anderson, J. C, (1979), Local Union Participation: A Re-examination, Industrial
Relations,
18,
18-31,
Batstone, E,, Boraston, I, & Frenkel, S, (1977),
Shop Stewards
in Action. Oxford: Basil Blackwell,
Blyton, P, R,, Nicholson, N, & Ursell, G, (1981), Job status and white collar members' union acuwhy. Journal
of Occupational Psychology, 54, 33—45,
Flood, P (1988), Trade union government and membership participation in the Irish Transport and General
Workers' Union, Unpublished PhD, London School of Economics,
Hartley, J,, Kelly, J, & Nicholson, N, (1983),
Steel
Strike:
A CaseStudy in Industrial
Relations.
London: Batsford
Academic,
Hemingway, J, (1978),
Conflict
and
Democracy:
Studies in Trade Union
Government.
Oxford: Clarendon Press,
Huszczo, G, E, (1983), Attitudinal and behavioural variables related to participation in union activities.
Journal of Labor
Research,
4, 289-297,
James, L. (1984),
Power
in a Trade Union. London: Cambridge University Press,
Klandermans, P, G, (1984a), Mobilisation and participation in trade union action: A value expectancy
Appro'^ch. Journal of
Occupational
Psychology,
57, 107-120,
Klandermans, P, G, (1984^), Mobilisation and participation: Social psychology expansions of resource mobil-
isation theory, American
Sociological
Review,
49, 583—600,
Expeaancy value
analysis:
Attendance
at
union meetings
223
Klandermans, P, G, (1984f), Membership meetings and decision making in trade unions. Paper presented at
the E,G,O,S, Colloquium on Trade Unions in Europe: The Organisational Perspective, 11-13 October,
Amersfoort, The Netherlands,
Klandermans, P, G, (1986), Psychology and trade union participation: Joining, acting, quitting. Journal of
Occupational
Psychology,
59, 189-204,
Lieberman, S, (1956), The effect of changes in roles on the attitudes of role occupants. Human Relations, 9,
385-402,
McShane, S, L, (1986), The multidimensionality of union participation,JwrKa/ of
Occupational
Psychology,
59,
177-187,
Moore, R, J, (1980), The motivation to become a shop steward, British Journal of Industrial Relations, 18,
91-98,
Nicholson, N, (1976), The role ofthe shop steward: An empirical case study. Industrial Relations Journal, 1,
15-26,
Nicholson, N,, Ursell, G, & Blyton, P (1981), The Dynamics of White Collar Unionism. London: Academic
Press,
Oliver, P, (1984), If you don't do it, nobody else will: Active and token contributions to local collective action,
American
Sociological
Review,
49, 601—610,
Staw, B, M, (1977), Motivation in organizations: Towards synthesis. In B, M, Staw (Ed,), The
Psychological
Foundations
of Organization!
Behaviour.
Santa Monica, CA: Goodyear,
Vroom, V, (1964), Work and
Motivation.
New York: Wiley,
Received
15 February 1992;
revised version received
16 July 1992
