Upload
michel-tremblay
View
212
Download
0
Embed Size (px)
Citation preview
A Study of the Determinants and of the
Impact of Flexibility on Employee Benefit
Satisfaction
Michel Tremblay,1,3 Bruno Sire,2 and Annie Pelchat1
The purpose of this study was to e xam ine the influe nce of individual
characte ristics and organizational justice on employee benefit satisfaction, and
to explore the role of flexible benefit plans. Employees from three Canadian
organizations were surve ye d. A total of 285 usuable questionnaires were
returned, for a response rate of 42% . The variables in the model accounte d for
more than 40% of the variance in benefit satisfaction. The findings showed that
while distributive and proce dural justice were use ful in predicting benefit
satisfaction, the concept of process justice had the greatest e ffect on satisfaction.
Among the variables, communication had the gre atest impact. The effect of
flexibility, although significant, was ambiguous. Sociodemographic factors had a
ve ry limited effect when perceptual variables were introduced into the equation.
The paper also sets out the limitations of the study and its practical implications,
and makes some suggestions for future research.
KEY WORDS: benefits; satisfaction; flexibility; communication; organizational
justice.
INTRODUCTION
The cost of benefits has increased conside rably in recent years. In the
U.S., it represents nearly 40% of salary (Kroll & Dolan, 1985¯1986; McCaf-
fery, 1988) , and in Canada between 30¯35% (Theriault, 1991) , although it
may be as high as 39.6% in some sectors (IRIR, 1994) . At least part of
the obse rved increase may be due to societal pressure , favorable tax treat-
Hum an Relations, Vol. 51, No. 5, 1998
667
0018-7267/98/0500-0667 $15.00/1 Ó 1998 The Tavistock Institute
1École des Hautes É tudes Commerciale s de Montréal, Service de l’Enseignement de la Ge s-tion des Ressource s Humaines, 3000, chemin de la Côte-Sainte-Catherine, Montréal, Québec,
H3T 2A7, Canada.2ESUG, Unive rsité de Toulouse 1, 2, rue Albert-Lautman, 31000 Toulouse, France.3Requests for reprints should be addressed to Michel Tremblay, É cole des Hautes É tudesCommerciale s de Montréal, Service de l’Enseigne ment de la Gestion des Ressource s Hu-
maines, 3000, chemin de la Côte-Sainte-Catherine, Montréal, Québec, H3T 2A7, Canada.
ment, paternalism on the part of employers, a desire on the part of em-
ployees to have more leisure time, the provision of economic protection
for employe es, and union pressure (Bergmann et al., 1994) . To control the
escalation of costs in today’s competitive environm ent, some employe rs
have decided to re duce the ir own contribution to certain benefits while
increasing the direct contribution of their employe es and monitoring the
use of benefits more agressively (McCaffery, 1988; Balkin & Griffeth, 1993;
Bergmann et al., 1994). Others have chosen a different path, offering the ir
employees flexible benefits plans. Their aim is to maintain satisfaction lev-
els and at the same time control costs (Cavagnac & Sire , 1994) .
The essential feature of flexible benefit plans is that they allow employ-
ees, within certain limits, to make their own choice s or to construct the ir own
benefits package . In this type of program, the employe r determines both the
budget allocate d for indirect compensation and the choices offered, but it is
the employee who decides which of the benefits offered he or she wishes to
receive . Generally speaking, the choices made are effective for 1 or 2 years,
and can be modifie d at the employe e’s request (Kroll & Dolan, 1985¯1986;
Besser & Franck, 1989). In North America, this type of benefit system was
first deve loped in the late 1960s, but did not become popular until the early
1980s (McCaffery, 1988). In the last 15 years, we have witnessed a steady
growth in the number of flexible benefit plans. In 1990, according to a study
by Hewitt Associate s, more than 50 employers in Canada and more than 1200
in the U.S. had introduce d flexible benefit systems. A recent study conducte d
in Canada by the same firm revealed that nearly 16% of employe rs had in-
troduce d this type of plan (Hewitt Associate s, 1995).
The three most popular flexible benefits plans are modular plans, core-
plus-options plans, and flexible spending accounts (Gomez-Mejia, Balkin,
& Cardy, 1995) . Modular plans consist of a series of different benefit bun-
dles or diffe rent leve ls of coverage designed for different subgroups of the
employee population. Employe es can select only one bundle , and coverage
leve l substitutions are not usually allowe d. Core-plus-options plans consist
of a core group of essential benefits together with a wide array of other
benefit options that employe es can add to the core . The core is designed
to provide a minimum level of economic security. Under plans such as this,
employees receive benefit credits that entitle them to purchase additional
benefits of value to them.
The flexible spending account is the most simple type of flexible bene-
fit plan. Here, employers allow their employe es to pay for certain categories
of eligible benefits that are not include d in the plan, using untaxe d dollars.
This type of plan is in fact a bank account used to pay claims and managed
by the employer.
668 Tremblay, Sire, and Pelchat
The growing inte rest in flexible benefit plans can be explaine d by their
many pote ntial advantage s. These plans allow employe es to choose the bene-
fits that best satisfy their personal needs (Beam & McFadden, 1988; McCaf-
fery, 1988; Milkovich & Newman, 1990), to unde rstand and appre ciate the
benefits offered and the relate d costs (Beam & McFadde n, 1988; Rose, 1988;
Milkovich & Newman, 1990; Hornsby et al., 1991), and to avoid pointle ss
duplication of benefits for couple s with two income s (Baker, 1988) . For em-
ploye rs, the flexible programs provide a better way of satisfying the changing
needs of salaried employe es (Beam & McFadden, 1988; Milkovich & New-
man, 1990), facilitate the recruitment and retention of employe es (Cable &
Judge , 1994; Rylan & Rosen, 1988; Rose , 1988), permit the introduction of
new, less costly benefits (Milkovich & Newman, 1990; Theriault, 1991), and
help increase employee satisfaction (Barbe r et al., 1992).
Despite the cost of benefits and the growing popularity of flexible plans,
employe e benefits have received little attention by personne l/human resource
researchers (Gerhart & Milkovich, 1992; Lengnick-Hall & Bereman, 1994).
Other than the theoretical work by Miceli and Lane (1991) , Danehower and
Lust (1992) , Lengnick-Hall and Bereman (1994) , and Cavagnac and Sire
(1994) , and the empirical studie s by Barber et al. (1992), Rabin (1994), Davis
and Ward (1995) , Poilpot-Rocaboy (1995) , and Williams (1995), very little
empirical research has been carried out to assess the validity of any or all of
these models and to test the influence of flexible benefit policie s. The goal
of the research described here is to test the validity of specific propositions
related to the determinants of benefit satisfaction and to explore the influ-
ence of flexible benefit plans on employee satisfaction.
THEORETICAL FRAMEWORK AND HYPOTHESES
Past studies of the determinants of benefit satisfaction have focused mainly
on the relationships between sociodemographic characteristics and level of sat-
isfaction, considered jointly as benefit satisfaction in a piecemeal fashion, and
have analyze d the variable s determining satisfaction with specific benefits. To
our knowledge, only Williams (1995) has tested a theoretical model of the an-
tecedents of benefit satisfaction. The results of all this work reveal that two
groups of variable s may have a direct or indirect effect on benefit satisfaction:
factors related to the individual and factors related to organizational justice .
Influence of Factors Related to the Individual
Studie s of benefit satisfaction have focused mainly on two individual
factors: sociodemographic variable s, in particular, sex, age , level of educa-
tion, size of family, seniority, salary and organizational leve l, and attitudinal
Flexibility and Em ployee Benefit Satisfaction 669
variable s related to value s, needs, desirability, and preferences regarding
benefits.
Past empirical studies of sociodemographic characteristics have produced
mitigated results. The results regarding sex, for example , are inconsiste nt. Lust
(1987), Scarpello et al. (1988), Hemmasi et al. (1992), and Rabin (1994) did
not observe any significant relationship between sex and overall benefit satis-
faction, whereas Balkin and Griffeth (1993) found that, generally speaking,
women were more satisfied than men with their benefits. The results concern-
ing the relationship between age and benefit satisfaction seem to be equally
inconsiste nt. Lust (1987), Dreher et al. (1988), Hemmasi et al. (1992), and Wil-
liams (1995) found no significant relationship between these two variable s. On
the other hand, Scarpello et al. (1988) observed a positive relationship, and
Balkin and Griffeth (1993), Rabin (1994), Poilpot-Rocaboy (1995), and Judge
(1993) all observed a negative relationship. Evidence of a link between level of
education and benefit satisfaction is equally ambiguous. For example , Lust
(1987) and Scarpe llo et al. (1988) found a positive relationship, Lust (1990) and
Balkin and Griffeth (1993) found a negative relationship, and Hemmasi et al.
(1992) found no significant relationship. Although very little work has been
done on the impact of family size, the findings are still inconsistent. Lust (1987)
and Poilpot-Rocaboy (1995) found no significant relationship, while Dreher et
al. (1988) observed a negative one. Evidence of the influence of seniority is
equally inconclusive . Lust (1987, 1990) found a positive relationship between
seniority and overall benefit satisfaction, Scarpello et al. (1988) and Rabin
(1994) found a negative one, while Dreher et al. (1988), Balkin and Griffeth
(1993), Hemmasi et al. (1992), and Williams (1995) observed no significant re-
lationship. The impact of level of salary on benefit satisfaction has been found
to be fairly constant and positive (Lust, 1987; Dreher et al., 1988; Balkin and
Griffeth, 1993). However, this same impact was limited when perceptual vari-
ables such as perception of justice were introduce d into the model (Hemmasi
et al., 1992; Williams, 1995; Poilpot-Rocaboy, 1995). The results regarding the
relationship between organization level and benefit satisfaction have also been
inconsiste nt. Scarpello et al. (1988) found a positive relationship, Lust (1987)
a negative one, and Williams (1995) observed no significant relationship.
In fact, the empirical results re garding the impact of socio-de mo-
graphic factors have been so inconclusive that it is difficult, at the present
time, to draw any conclusions on the existence and meaning of the vari-
ations in the relationships. In addition, the direct effect of personal factors
on benefit satisfaction seems to be limited when perceptual variable s are
include d in the mode ls. Despite their presumed relevance , we will therefore
not propose any specific hypothe ses regarding these factors, but will pursue
our inve stigations in this field.
670 Tremblay, Sire, and Pelchat
The weakne ss of the demographic approach in explaining employe e
attitude s has led some authors to examine employe e preference or need
(Milkovich and Newman, 1990) . Employee values are propose d as deter-
minants of benefit satisfaction in the mode ls of both Mice li & Lane (1991)
and Danehower and Lust (1992) . However, the ir role is uncle ar, and has
received little attention in empirical studie s of satisfaction (Danehower et
al., 1994). According to Locke (1976), the more importan ce individuals at-
tach to a give n aspect, the greater their satisfaction with that aspect. In
empirical work, Lust and Danehower (1992) found a positive relationship
between the importance attache d to benefits and the satisfaction with re-
spect to leave and retirement, and Williams (1995) obse rved a positive , but
not significant, relationship between benefit desirability and benefit satis-
faction. These results suggest that there may be a positive relationship be-
tween the importance attache d to benefits and employee satisfaction.
Hypothesis 1. The more importance individuals attach to benefits, the
more satisfied they will be with those benefits.
In the same way that value s are acquired through individual expe rience,
a benefit will be valued more highly if the individuals concerned have a
knowledge of the benefits that are offered to them. The study of Dreher et
al. (1988) showed that employe es who had an accurate view of the ir coverage
were more satisfie d with their benefits package s than employees who had an
inaccurate view. Such awareness may be a function of benefit history. Miceli
and Lane (1991) defined benefit history as an individual’s cumulative expe-
rience with benefits across organizations and over time. An individual who
has experienced several organizations during his or her career will probably
have acquire d a greater understanding of the benefits issue than someone
who has had only one employer. Moreove r, every move between employers
constitute s an opportunity to become better informe d and more aware of the
benefits package , and greater mobility provide s concre te points for compari-
son of compensation (Mice li & Lane , 1991). Research on the question of pay
satisfaction supports the importance of an historic standard (Hill, 1980) .
Thus, we can formulate the following hypothe sis:
Hypothesis 2. There is a positive relationship between interorganiza-
tional mobility and benefit satisfaction.
Benefits are aimed, among othe r things, at prote cting employe es and
the ir dependents from risks that could jeopardize their health and financial
security (Gomez-Mejia et al., 1995; Milkovich & Newman, 1990) , and pro-
viding continuity of income in a varie ty of circumstances and situations
Flexibility and Em ployee Benefit Satisfaction 671
(e.g., sickne ss, injury, une mployment, disability, retirement, death) . A bene-
fits plan perceived as providing a high leve l of security is more like ly to
increase employe e satisfaction. However, although the notion of protection
forms the core of all benefits plans, to our knowledge no research has yet
been done on the relationship between perception of security and attitude s
toward benefits. We therefore propose the following hypothe sis:
Hypothesis 3. There is a positive relationship between perception of
security and satisfaction with benefits.
Influence of Factors Related to Organ izational Justice
Distributive Justice
Researchers working on justice have identifie d two principal forms of
justice : distributive justice and procedural justice . The theoretical models
of benefit satisfaction (e.g., Miceli & Lane , 1991; Danehower & Lust, 1992)
suggest that these two components of justice may affect satisfaction.
The term “distributive fairne ss” means the perception of whether the
rewards received by individuals are fair. The explanatory model most often
cited in the literature for explaining the effects of distributive fairness is
that of Adams (1965) , which has been described as one of the most valid
frameworks for inte rpreting human motivations (Summers & DeNisi, 1990) .
According to this theory, individuals assess the fairne ss of their contribu-
tions and benefits by comparing with those of othe r people , known as “ref-
erents.” The referents may be the individuals themselve s or people with
the same or different jobs in the same company or elsewhere. Research
has shown that employees look to a varie ty of referents when assessing the
fairness of their salary, and that a give n referent will not necessarily have
the same significance in the eyes of all individuals (Summers & Hendrix,
1991; Summers & DeNisi, 1990; Tremblay et al., 1997) .
Several studies have shown that the perception of distributive justice
has a positive impact on pay satisfaction (Dyer & Theriault, 1976; Weiner,
1980; Berkowitz et al., 1987; Konovsky et al., 1987; Folger & Konovsky,
1989; Sweeney, 1990; Summers & Hendrix, 1991) and that satisfaction lev-
els vary according to the type of referent (Ronen, 1986; Scholl et al., 1987;
Summers & DeNisi, 1990; Tremblay & Roussell, 1996) . For example , Ronen
(1986) obse rved a stronger link between pay satisfaction and inte rnal equity
than between pay satisfaction and external equity. Similarly, Scholl et al.
(1987) and Tremblay and Roussel (1996) found that perceptions of internal,
external, and employe e equity had a significant influe nce on pay satisfac-
672 Tremblay, Sire, and Pelchat
tion, and that the strength of the relationship varie d according to the nature
of the referent.
Nevertheless, the link between distributive justice and benefit satisfac-
tion has received little attention. To our knowle dge , only two studies have
tested the impact of individual comparisons with other referents on benefit
satisfaction. Williams (1995) found a positive relationship between benefit
standard comparisons (e xternal, education, responsibility, and age ) and
benefit leve l satisfaction. Individuals who rated their current level of benefit
coverage as be ing better than the coverage of othe rs were more satisfied.
Davis and Ward (1995) found evide nce that employe e perceptions of dis-
tributive justice are important predictors of benefit satisfaction. Thus, we
propose the following hypothe sis:
Hypothesis 4. There is a positive relationship between the perception
of distributive justice and benefit satisfaction.
Research has shown that individuals attach diffe rent leve ls of impor-
tance to the diffe rent salary referents (Hill, 1980; Tremblay et al., 1997) ,
and Blau (1994) recently found that the effect of distributive justice per-
ception on satisfaction varies according to the importance attache d by the
employees to the salary referents. Thus, it is reasonable to think that if
individuals conside r benefits to be very important and have a positive per-
ception of benefit justice , the n the y should be more satisfied with the ir
benefits. If, on the othe r hand, they have a positive perception of the bene-
fits package but conside r benefits to be less important, they will be less
satisfied. This sugge sts that the perceived importance of benefits may affect
the relationship between distributive justice and benefit satisfaction. In this
respect, Harris and Fink (1994) pointe d out the need to include a mod-
erator variable in order to improve our understanding of benefit satisfac-
tion. We therefore propose the following hypothe sis:
Hypothesis 5. The perceived importance of benefits plays a moderator
role in the relationship between distributive justice and benefit satisfaction.
Procedural Justice
To unde rstand why individuals react in one way or anothe r to unfair
treatment, we must look at a second form of fairne ss in organizational mat-
ters: procedural justice . Procedural justice is concerned with individual re-
actions to the process used to establish the reward (Greenbe rg, 1990)—in
other words, the means rather than the ends (Sweeney & McFarlin, 1993) .
Research on this que stion has shown that employees are able to make a
Flexibility and Em ployee Benefit Satisfaction 673
clear distinction between “the ends” and “the means” (Thibault & Walker,
1975; Sheppard & Lewicki, 1987; Daile y & Kirk, 1992) and that these two
notions have independent effects (Alexande r & Ruderman, 1987; Folger
& Konovsky, 1989; Sweeney & McFarlin, 1993; Scarpe llo & Jones, 1996) .
Two procedural elements seem to be relevant from the compensation
standpoint, i.e ., degree of control over the process leading to pay-re lated
decisions, and degree of control over compensation decisions. Some authors
have shown that control ove r the process, for example , by giving employe es
the opportunity to choose a desired compensation form, to participate in
the design of a compensation system, to voice the results of compensation
decisions, or to receive accurate information, can produce a strong sense
of process justice and more positive attitude s to the results and the organi-
zation (Leventhal, 1980; Lind & Tyler, 1988; She ppard et al., 1992; Green-
berg, 1996) .
Early studies testing the e ffect of administrative components in the
field of compe nsation showed that perceptions of procedural justice ex-
plaine d a large portion of the variance in pay satisfaction (Dyer & Theri-
ault, 1976; Weiner, 1980) . Similarly, Jenkins and Lawler (1981) found a
link between involve ment in compensation decisions and pay satisfaction.
Folger and Konovsky (1989) also found a positive link between satisfaction
and the existence of an appe al proce ss in the de termination of pay in-
creases. However, the effect of procedural justice on benefit satisfaction
has not yet been clearly demonstrate d. Mulvey (1992) found that the power
to appe al against pay decisions and consistency in the application of pay
policie s—both criteria of procedural fairness—were positive ly and signifi-
cantly linked to benefit satisfaction. While Williams (1995) did not specifi-
cally test the effect of the employe e input variable , her results neverthe less
suggested that inclusion of that variable in the determination of the benefits
package may increase benefit satisfaction. It is therefore possible to con-
clude that giving employe es the chance to participate in decisions related
to benefits (e.g., choice) and taking account of the ir benefit preferences
(e.g., voice) will have a positive effect on satisfaction.
Hypothesis 6. There is a positive relationship between the perception
of involve ment in decisions related to benefits and benefit satisfaction.
Hypothesis 7. There is a positive relationship between the perception
that employe e preferences are conside red and benefit satisfaction.
The literature suggests that an elaborate communications program is re-
quire d if the compensation program is to be fully appre ciated (Gomez-Me-
jia & Balkin, 1992; Barocas 1993; Sire, 1993; Danehower et al., 1994) . The
theoretical mode ls of Mice li and Lane (1991) and Danehower and Lust
674 Tremblay, Sire, and Pelchat
(1992) highlighte d the pote ntial effect of informational justice on benefit
satisfaction. Folge r and Greenberg (1985) sugge st that communication of
this type of information creates a sense that procedures have been followe d
correctly, and generates confidence in the people making the decisions. The
provision of more information on benefits can also reduce the gap between
employee expectations and perceptions (Danehower et al., 1994) . The work
of Mulve y (1992) , Rabin (1994) , Poilpot-Rocaboy (1995) , Williams (1995) ,
and Davis and Ward (1995) has shown that the more information employ-
ees have , the more satisfie d they are with the ir benefits. Indire ctly, Dreher
et al. (1988) also highlighte d the predominant role of communication. They
found that employe es who receive precise , accurate information on the ir
coverage are more satisfie d than those who receive no relevant information.
It would therefore seem, as sugge sted by Barbe r et al. (1992) and Rabin
(1994) , that communication is more important in organizations that have
introduce d flexible benefits policie s. If an employe r offers a flexible benefit
plan, it must disseminate extensive information on the various options avail-
able , to enable its employe es to make informe d choice s. Although very little
comparative empirical data is available on the leve l of communications ef-
fort required for traditional plans and flexible plans, we neverthe less think
it is justified to take this variable into account. We therefore propose the
following hypothe sis:
Hypothesis 8. There is a positive relationship between communication
received and benefit satisfaction.
A flexible benefits policy provide s a rare opportunity for employees to ex-
ercise choice in the area of compensation. Farh et al. (1991) suggested that
allowing employe es to choose their compensation increases both their con-
trol and the ir chances of satisfying the ir needs, and thus the ir satisfaction
leve ls.
One argument often invoke d when such programs are establishe d is
they will he lp obtain a higher corresponde nce between individual prefer-
ences and the benefits offered. Inde ed, from a theoretical point of view,
if employe es perceive the system of compensation to be well matched to
the ir preferences, they are bound to be more satisfie d than if no such cor-
relation existed (Dyer & Theriault, 1976; Mulve y, 1992) .
To our knowledge , only two studie s have conside red the effect of the
introduction of flexible benefits plans. Barber et al. (1992), in a study of
110 employe es in a large service sector company, observed a significant
increase in benefit satisfaction following the introduction of a new flexible
benefit program. Anothe r study (Rabin, 1994) , conducted in a large private
company after the introduction of a flexible benefit program, revealed that
Flexibility and Em ployee Benefit Satisfaction 675
the employe es were more satisfie d with the ir benefits under the flexible
plan than under the former fixed benefit plan.
Folge r and Greenberg (1985) suggested that the act of choosing may
be a strong determinant of satisfaction. Thus, it is important to know
whether individuals are more satisfie d when they are able to choose the ir
benefits, or when benefits are imposed upon them. We should also consider
whether the act of choosing benefits has an effect on satisfaction after con-
trolling for the fact that benefits plans take preferences into account. To
our knowle dge, no study involving more than one plan has ever tested the
influence of benefit flexibility in a multivariable mode l. The above argu-
ments enable us to postulate that the power to choose benefits would have
a positive effect on employe e satisfaction.
Hypoth esis 9. There is a positive relationship between flexibility of
benefits and benefits satisfaction.
RESEARCH DESIGN AND METHODOLOGY
Sample
The data in this study were collected from three Canadian organizations:
an insurance company (A), a university institution (B), and a drinks manu-
facturing company (C). The insurance company had a traditional (fixed) sys-
tem of benefits, the educational institution a modular flexible plan, and the
drinks manufacturing company a benefits system resembling a “core-plus-op-
tion” plan, or basic protection plus options. A questionnaire was distribute d
to 672 employees. A total of 285 completed copies were returned, for an
overall response rate of 42.4% . In terms of a sociodemographic profile , 63%
of the responde nts were women, 30% had a unive rsity degree, the average
age was 37, and the average length of service was 6 years. Some differences
in respondent profile s emerged from the data. Organization C had a highe r
percentage of men, a higher level of education and more reporting leve ls,
and ave rage salarie s were higher than in organizations A and B. In organi-
zation B, responde nts were olde r and had more seniority than in organiza-
tions A and C. These differences will be discusse d in more detail later, in the
section on the limits of the research.
Measures
Individual Variables. Individual characteristics were assessed from meas-
ure ments recorded by the surve y participant s. Gender was treated as a
“dummy” variable (male = 0, female = 1). Age, seniority, and number of
676 Tremblay, Sire, and Pelchat
dependents were coded directly. Level of education was classifie d into three
categories (secondary level = 1 to unive rsity leve l = 3) and salary into 11
categories, from less than $30,000 (1) to $60,000 or more (11). Organizational
level was treated as a dummy variable and coded 0 for production and office
staff and 1 for professional and managererial staff. The ratio of benefits to
salary was coded into five categorie s, from 5% (1) to 25% or more (5). Im-
portance of benefits was measured on a 5-point scale (1 = not very important
to 5 = very important), using the following item: “To what extent is your
system of benefits important to you? ” Interorganizational mobility was meas-
ured by the number of employers for whom the responde nt had worked in
the course of his or her career. Perception of security was measured by two
items using a 5-point scale from “completely disagre e” (1) to “comple tely
agree” (5) (e.g., the benefits I receive from my employe r give me and my
family a sense of security, Cronbach alpha: 0.70).
Organizational Justice. The perception of distributive justice was meas-
ured using the scales develope d by Goodman (1974) and Tremblay et al.
(1997) . For the purpose s of this study, distributive justice linked to benefits
was broke n down into justice base d on needs, internal equity (related to
the responde nt’s immediate superior and colleague s) and external equity
(relate d to other companie s). For each of these referents, responde nts were
asked to locate their benefits on a 7-point scale (¯3 = conside rably less, 0
= more or less the same, + 3 = conside rably more). Here 0 represented
equity, ¯3 extreme negative inequity and + 3 extreme positive inequity. We
carried out a factorial analysis to see whether the different referents rep-
resented distinct constructs. The analyse s showed that the four aspe cts of
equity measured the same concept, with a Cronbach alpha of 0.79.
Level of involve ment was measured by one item. Responde nts were
asked to indicate the extent to which they were involve d in these decisions,
using a 5-point scale (1 = others make the decision, 3 = equal contribution
to the decision-making process, 5 = the decision is entirely up to me). The
variable representing conside ration of preferences was measure d by the fol-
lowing item: “To what extent doe s your organization take account of your
preferences in determining the different aspects of your benefits package ? ”This item was measure d using a 5-point scale from “takes no account what-
soever” (1) to “takes full account” (5). Communication effort was measured
by means of one item that sought to evaluate the degree of satisfaction
with the information received (“In your view, the information provide d by
your employe r on your benefits is:”). This item was measured on a 5-point
scale from “highly unsatisfactory” (1) to “highly satisfactory” (5).
Two indicators were used to measure the operationalization of flexi-
bility. For the first measurement, we distinguishe d between the company
that did not have a flexible plan (coded 0) and the two that had flexible
Flexibility and Em ployee Benefit Satisfaction 677
plans (coded 1). For the second, we distinguishe d the two companie s that
had flexible plans by a function of the degree of fle xibility of each; we
oppose d the modular plan (the less flexible of the two = 0) against the
“core-plus-options” plan (the more flexible = 1).
Dependent Variable. To measure benefit satisfaction, we drew on the
work of Lust and Danehower (1992). Using the ir model and inve stigative
instrument as a basis, we separated the creation of satisfaction into two
components: satisfaction with the cost and satisfaction with the quality of
benefits. The cost dimension was measured by four items (e.g., the amount
I pay for my benefits program) , and the quality dimension by three items
(e.g., the form of the benefits available to me). These items were adapte d
for French speakers and measure d using a 5-point scale. Here again, we
carried out factorial analysis to validate Lust and Danehower’s (1992) two
components of benefit satisfaction. The analyse s did not allow us to dif-
ferentiate the compone nts of cost and quality, since each item was placed
unde r the same factor. The degree of inte rnal consiste ncy for the aggregate
of the items, cost and quality, was found to be very satisfactory, with a
Cronbach alpha of .87.
Statistical Analyses
The research hypothe ses were tested using bivariate and multivariate
statistical technique s. We used a Pearson-type corre lation to test for a line ar
association between the variable s. Table I presents the mean, standard de-
viation, and correlation matrix of the variable s studied. The correlations
between the variable s range from .00 to .55, showing that there are no
serious multicolline arity proble ms. Nevertheless, give n the strong correla-
tion between the salary and organizational level variable s, we decided to
remove the latter from the mode l. We began by running a simple regres-
sion, entering all the variable s at once. To test the moderating effects of a
give n variable , we followe d the recommendations of Jame s and Brett
(1984) , introducing the interaction of distributive justice and importance
of benefits into the equation. We used usefulness analysis to asse ss the
incremental variations in satisfaction measured by predictors (Darlington,
1968) . Se veral combinations of variable inputs were tested to assess the
contribution of individual and organizational justice predictors.
RESULTS
The results of the simple regression, containe d in Table II, show that
all sets of variable s explaine d between 37% and 40% of the variance in
benefit satisfaction (F = 6.3 and 5.7; p < .001) . Sociode mographic variable s
678 Tremblay, Sire, and Pelchat
Ta
ble
I.D
esc
rip
tive
Sta
tist
ics
an
dC
orr
ela
tio
ns
Be
twe
en
Va
ria
ble
s
Me
an
SD
12
34
56
78
91
01
11
21
31
41
5
1.
Be
ne
fits
sati
sfa
ctio
n
20
.55
.1
2.
Se
x0
.60
.5.2
6*
**
3.
Ag
e3
7.2
9.1
.04
¯.05
4.
De
pe
nd
en
ts0
.80
.8.0
6¯.0
7.1
2*
5.
Ed
uca
tio
n2
.10
.9¯.1
1¯.2
6*
**
¯.25
**
*¯.1
7*
6.
Se
nio
rity
7.8
6.9
.05
¯.12
.55
**
*.0
3¯.2
6*
**
7.
Sa
lary
6.6
2.5
.21
**
*¯.5
7*
**
.28
**
*.1
1.3
9*
**
.32
**
*8
.B
en
efi
t
pro
po
rtio
n
3.1
1.3
.03
¯.09
.11
.01
.03
.18
**
.28
**
*
9.
Inte
rorg
an
-
iza
tio
na
lm
ove
s
3.1
1.6
.03
¯.07
.38
**
*.1
8*
*¯.1
4*
¯.11
.11
.01
10
.B
en
efi
tim
po
rta
nce
2.1
0.7
¯.15
.17
**
.09
.05
¯.28
**
*.1
8*
*¯.1
1.1
0.0
4
11
.S
ecu
rity
6.5
1.6
.51
**
*.0
1.0
8.0
7.0
5.0
7.0
4.1
6*
*.1
6*
*¯.0
51
2.
Dis
trib
uti
ve
just
ice
15
.72
.3.3
4*
**
.09
¯.03
.12
*.0
2.0
3.0
4.1
8*
*.0
5.0
5.3
9*
**
13
.P
art
icip
ati
on
2.1
1.3
.06
.08
.08
.05
.04
.01
.02
.07
.04
.19
**
.16
**
.15
*
14
.P
refe
ren
ces
2.6
1.0
.18
**
¯.05
.17
**
.11
¯.15
*.2
1*
**
.02
.05
.07
.11
.29
**
*.2
0*
**
.20
**
*1
5.
Co
mm
un
ica
-
tio
n
3.3
1.1
.58
**
*.0
9.0
7¯.0
2¯.0
1.0
3¯.0
5.1
5*
.05
.05
.52
**
*.2
3*
**
.11
*.2
4*
**
16
.F
lexi
bil
ity
0.8
0.4
¯.20
**
¯.22
**
*.0
9¯.1
0.3
4*
**
.03
.35
**
*¯.0
9.0
7¯.0
9¯.0
6.0
1.1
9*
*.1
1.1
0
*p
<.0
5.
**
p<
.01
.
**
*p
<.0
01
.
Flexibility and Em ployee Benefit Satisfaction 679
were found to play an unimportant role in explaining our dependent vari-
able : none (sex, age, dependents, education, salary, leve l, seniority, propor-
tion of benefits) had a significant independe nt effect beyond .10 on benefit
satisfaction. However, contrary to our hypothe sis 1, we observe a negative
relationship between the importance ascribe d to benefits and the leve l of
satisfaction (Beta = ¯.15 and ¯0.18; p < .05) . Contrary to our hypothe sis
2, we found an insignificant negative relationship between interorganiza-
tional mobility and satisfaction (Beta = ¯.15, p < .10) . Hypothesis 3, the
influence of security, was not confirmed. The results show a positive but
non-significant relationship between the perception of security and benefit
satisfaction (Beta = .15 and .17, p < .10) .
Table II. Results of the Regression Analysis Betwee n the De terminants and Benefit Satis-
faction (N = 285)
Variables Beta T Beta T
Individual variables
Sex .10a 1.18 .15 1.64Age .07 .67 .11 .83
Dependents ¯.02 ¯.23 ¯.03 ¯.36Education ¯.12 ¯1.26 ¯.10 ¯.88
Seniority ¯.07 ¯.68 ¯.09 ¯.69Level of salary .05 .52 .01 .03
Benefit proportion ¯.05 .76 ¯.08 ¯.93Interorganizational moves ¯.15" ¯1.63 ¯.14 ¯1.27
Benefit importance ¯.15* ¯2.03 ¯.18* ¯2.15Security .15" 1.93 .17" 1.63
Organizational justiceDistributive justice .14" 1.82 .20* 2.11
Distri ´ Importance .29 .61 .38 .69Participation in decision ¯.07 ¯.92 .11 1.24
Prefere nces .14" 1.83 .19* 2.20Communication .43*** 5.26 .46*** 4.80
Flex1 ¯.17* ¯1.94Flex2 .21* 2.05
R2 .46 .46
R2 adjusted .40 .37
F-value 6.7*** 5.27***Usefulness analysisb
Individual variable s beyond organizationaljustice variables
.05
Distributive justice beyond procedural justice .03**Proce dural justice beyond distributive justice .25***
aStandardized regression coefficients.bWhere the entry indicates x beyond y it means the increment in the square of the multiple
corre lation coefficient when x is added following y. Otherwise, the entry is the multiple cor-
re lation coefficient (Folger & Konovsky, 1989) .
“p < .10.
*p < .05.
**p < .01.
***p < .001.
680 Tremblay, Sire, and Pelchat
Hypothe sis 4, concerning the effect of distributive justice , was partly
confirme d by our results (Beta = .14, p < .10 and .20, p < .05) . The fairer
responde nts perceived the ir benefits to be, and the better the benefits com-
pare d with those of the referents, the more satisfie d responde nts were. Hy-
pothe sis 5, suggesting that the importance ascribe d to benefits may have a
moderating effect between distributive justice and satisfaction, was not con-
firmed. Hypothesis 6, concerning procedural justice , was not confirmed,
since participation in benefit decisions was found to have a nonsignificant
effect on benefit satisfaction. However, hypothe sis 7 was supporte d. The
perception that the company took account of employe e preferences was
found to be positive ly linke d to benefit satisfaction (Beta = .19, p < .05) .
Hypothesis 8, concerning the effect of communication , was fully confirmed
(Beta = .43 and .46, p < .001) . Communication was found to be the main
predictor of benefit satisfaction. This suggests that the more information
employees have , the more satisfie d they are with the ir benefits.
Our results gave mitigate d support to our hypothe sis 8. In the first
mode l, it would appear (flex 1) that the fact of having a flexible plan has
a negative effect on benefit satisfaction (Beta = ¯.17; p < .05) . In the
second mode l, however, when the two flexible plans are examine d (flex 2),
flexibility is positive ly relate d to satisfaction (Beta = .21, p < .05) .
DISCUSSION
The analyse s provide evide nce on two major isssues: (a) the influe nce
of individual factors on benefit satisfaction, and (b) the role of organiza-
tional justice in the prediction of benefit satisfaction.
Although previous research has shown that individual factors have an
effect on benefit satisfaction, our research revealed the ir influence to be
very limited. In fact, the sociodemographic variable s explaine d less than
1% of the variance . These results seem to confirm the models of Dane-
hower and Lust (1992) and Miceli and Lane (1991) , in that these variable s
have an indire ct effect on satisfaction. Similarly, they confirme d the studies
of He mmasi e t al. (1992) , and William s (1995) , in that the sociode -
mographic factors had a very limite d effect when perceptual variable s were
introduce d into the mode l. It is possible , however, that these variable s may
have a direct effect when specific benefit elements are studied. Furthe r-
more , we observed a negative relationship between the importance of bene-
fits and satisfaction. Dane howe r and Lust (1992) found a positive
relationship between these two variable s for individual benefits. Our results
did not support the hypothe sis of a significant inte raction between the im-
portance of benefits and distributive justice . It may be , as sugge sted by
Milkovich and Newman (1990) , that need and preference arise out of a
Flexibility and Em ployee Benefit Satisfaction 681
sense of perceived equity or ine quity. These authors justifie d this inde-
pendence by the fact that employe rs grant benefits for reasons other than
to satisfy the needs of the ir employees—for example , to avoid unionization
or to copy competitors. Further research is require d to clarify the influe nce
of needs, values, and desirability on both the general and specific compo-
nents of benefit satisfaction.
Our results indicate that, as in the case of satisfaction with regard to salary,
the perception of equity in the benefits policy is positive ly linked to benefit
satisfaction. This finding is consistent with the results of Williams (1995) and
Davis and Ward (1995), suggesting that employees are able to assess the fair-
ness of their benefits and that any perceived unfairne ss will affect satisfaction
levels. Our results also suggest that employees base their perception of fairness
on their referents. However, we are not able to draw conclusions on the sig-
nificance of any particular referent. It would be relevant in subsequent research
to assess the impact of internal and external referents, to see whether some, as
suggested by Danehower and Lust (1992), are more significant than others.
With respect to procedural justice , our results did not demonstrate that
participation in benefits-related decisions significantly improved benefit sat-
isfaction. This aspect of procedural fairne ss had a significant effect on satis-
faction only when communication was removed from the mode l. However, it
is somewhat surprising that perceived consideration of preferences was not
linked more strongly to benefit satisfaction. In this respect, additional analy-
ses showed that there was no significant difference between the three type s
of plans studied. This suggests that employe es covered by a fixed plan may
feel that their employers are taking account of the ir preferences even when
they cannot select the ir benefits. This finding is consistent with the results of
Williams (1995) , who found a positive relationship between employe e input
and benefit satisfaction in a numbe r of traditional and fixed benefit plans.
O ur research confirms the importance of communication in bene fit
manage ment, and is consiste nt in this respect with the findings of Mulvey
(1992) , Rabin (1994), Poilpot-Rocaboy (1995), and Williams (1995). Commu-
nication was found to be by far the most important predictor of benefit sat-
isfaction. This reinforces the idea of adopting a marketing orientation for
employe e benefits (Danehower et al., 1994), and justifie s the relevance of
research, such as that of Driver (1980), Haar and Kossack (1990) , and Milk-
ovich et al. (1994) on the effectiveness of communication methods for bene-
fits. Interestingly, we found a strong correlation between communication and
security (r = .52, p < .001). This suggests that employees may have a better
understanding of benefit package s where an effective communications pro-
gram exists, and thus an increased sense of security or protection.
Our findings partly contradict the results obtaine d by Barbe r et al.
(1992) and Rabin (1994) . In fact, we observed that the existence of a flex-
682 Tremblay, Sire, and Pelchat
ible plan was not sufficie nt, alone , to improve satisfaction. The relationship
appe ared to be negative . This can undoubte dly be explaine d by the fact
that information on remuneration is not adapte d to the complexity of flex-
ible schemes.
Although the theoretical mode ls highlight the positive effect of benefit
plan flexibility on satisfaction, it is also true that these plans increase the
complexity of the choice s, and thus the risks and the fear of making a mis-
take in the selection process. This negative effect must, however, be quali-
fied since we also found that, where flexible plans exist, the most flexible
plans generate the highest leve ls of satisfaction. It would be instructive to
study a greater varie ty of flexible benefit plans and to establish whether
there is a positive or negative relationship between the degree of flexibility
and the degree of satisfaction. Beside s the feature s of flexible plans, it
would also be interesting, as Rabin (1994) did, to examine the impact of
the choice s made , and the effect of the maturity of flexible plans. Another
interesting line of research would be to study the conseque nces of benefit
dissatisfaction to see whether, as sugge sted by Harris & Fink (1992, 1994) ,
dissatisfaction actually increases the desire to join a union or change jobs,
or reduces organizational commitment.
Studie s in the fie ld of organizational justice have consiste ntly shown that
distributive fairne ss has a higher explanatory power than process fairness in
salary satisfaction (e.g., Folger & Konovsky, 1989). In contrast, our usefulness
analysis and the results obtained by Williams (1995) suggest that, in the benefit
context, procedural compone nts play a more important role than distributive
components in generating satisfaction. The results of the usefulness analysis
showed that, in our study, the contribution of procedural justice was nearly
eight times that of distributive justice (0.25 and 0.03). One possible explana-
tion for this would be that employees are require d to judge benefit manage -
ment more often than salary manage ment. As benefit manage ment involve s
a numbe r of complex tasks and procedures (e.g., claims processing for indi-
vidual and dependent insurance s), and improve ments in benefit package s are
few in numbe r and infrequent, the administrative compone nts become more
important to employe es. Further research is required to see why the roles of
these two dimensions of organizational justice differ for pay and benefits.
Limitation s and Conclusion
A numbe r of limitations in this research, caused by the subje cts, the
measure s, and the analytical technique s, require comment. First, the ele-
ment of conve nience in selecting the sample should be note d. The scarcity
of flexible benefit plans restricted the size and representative ness of the
sample , and prevented us from using random selection, although the weak-
Flexibility and Em ployee Benefit Satisfaction 683
ness of the demographic variable s in the multivariate models attenuated
the conseque nces of the imperfectly-matche d populations. The sample ,
drawn from only three companie s, is too small to enable us to generalize
our conclusions on the effects of flexibility. Furthe r research is require d to
examine benefit satisfaction among homogeneous populations subject to
diffe rent flexible plans.
Second, we used single items to measure several of the variable s. This
may have affected reliability. On the other hand, the pattern of results ob-
tained from these single -item measure s across the three samples was gen-
erally consistent.
Third, the methods used to gather and analyze the data should be
conside red. The research design was cross-sectional, and it is thus not pos-
sible to infe r causal relationships. Because fairness judgme nts are a dynamic
phe nomenon and the effect of fle xibility may change ove r time (Hewitt
Associate s, 1995) , there may be a need to use a longitudinal design.
A further limitation is that all the variable s, both independent and
depende nt, derive from the same que stionnaire . This raises concerns about
share d variance due to a common measure ment method. No statistical
technique provide s proof of causation. A structural equation under LIS-
REL could be used to ove rcome this problem, as it would take account of
measure ment errors in variable estimate s (Brannick, 1995) .
In conclusion, the study provides some evidence that organizational jus-
tice is a relevant conceptual framework for explaining employee attitudes to
benefits. For practitione rs and managers in organizations, it has shown the
relevance of benefit comparisons when designing benefits package s: the fairer
employees perceive the ir benefit package to be, in comparison to others, the
more satisfied they will be with their benefits. Most importantly, the results
sugge st that managers should pay careful attention to benefit management. A
better communications strategy seems to be essential in order to improve at-
titudes toward benefits. Companie s that want to increase employee satisfac-
tion levels should invest in information on their benefits before broadening
the range of bene fits offered (Williams, 1995) or introducing completely dif-
ferent policie s (e.g., flexible plans). Despite ambiguous conclusions on the ef-
fect of fle xible plans, process justice and attitude s can be reinforce d in
organizations that take account of employee preferences and needs by offering
diffe rent methods of participation, such as survey and focus groups.
ACKNOWLEDGMENTS
We would like to acknowle dge the helpful comments and suggestions
of David Balkin, and two anonymous reviewers. The research was funded
by le Fond pour la Formation de Chercheurs et l’Aide à la Recherche
684 Tremblay, Sire, and Pelchat
(FCAR) du Québec and a Social Sciences and Humanitie s Research Coun-
cil (SSHRC) of Canada.
REFERENCES
ADAMS, J. S. Inequity in social exchange. In L. Berkowitz (Ed.), Advances in experimental
and social psychology (Vol. 2). New York: Academic Press, 1965, pp. 267-299.
ALEXANDER, S., & RUDERMAN, M. The role of procedural and distributive justice in
organizational behavior, Social Justice Research, 1987, 1, 1401-1420.
BAKER, C. Flex your benefits. Personnel Journal, 1988, 67(5) , 54-61.
BALKIN, D., & GRIFFETH, R. The déterminants of employee benefit satisfaction. Journalof Business and Psychology, 1993, 7(3), 323-339.
BARBER, A., DUNHAM, R., & FORMISANO , R. The impact of flexible benefits on em-
ployee satisfaction: A field study. Personnel Psychology, 1992, 45, 55-75.
BAROCAS, V. Benefit Communications: Enhacing the Employe r’s Inve stment. The Confer-
ence Board, Report Numbe r 1035, 1993, 52 pp.
BEAM, B., & McFADDEN, J. Employee benefits (2nd ed.). Irwin Home wood, 1988, 539 pp.
BERGMANN, T., BERGMANN, M., & GRAHN, J. How important are employee benefitsto public sector employees. Public Personnel Managem en t, 1994, 23(3) , 397-406.
BERKOWITZ, L., FRASER, C., TREASURE, P., & COCHRAN, S. Pay, equity, job gratifi-
cations, and comparaisons in pay satisfaction. Journal of Applied Psychology, 1987, 72(4) ,544-551.
BESSER, L., & FRANCK, G. Artificial intelligence and flexible benefits decision: Can expert
systems help employees make rational choice? Advance Managem ent Journal, 1989, 54(2) ,4-9, 13.
BLAU, G. Testing the effect of leve l and importance of pay referents on pay level satisfaction,Hum an Relations, 1994, 47(10) , 1251-1268.
BRANNICK, M. Critical comments on applying covariance structure modeling. Journal of
Organizational Behavior, 1995, 27(1), 56-72.
CABLE, D., & JUDGE , T. Pay prefere nce and job search decisions: A person¯organization
fit perspective. Personnel Psychology, 1994, 47, 317-348.
CAVAGNAC, M., & SIRE, B. Rémunération cafétéria et engagemen t de l’employeur. Revue
de G estion des Ressources Hum aines, 1994, 12, 17-25.
DAILEY, R. C, & KIRK, D. J. Distributive and procedural justice as antecede nt of job dis-
satisfaction and intent to turnover. Hum an Relations, 1992, 45(3) , 305-317.
DANEHO WER, C., & LUST, J. A conceptual model of the determinants of employee benefit
satisfaction. Hum an Ressource Managem ent Review, 1992, 3(3) , 221-238.
DANEHO WER, C., CELUCH, K., & LUST, J. Benefits managemen t and communication: Amarke ting orientation. Hum an Resource Managem ent Review , 1994, 4(2), 177-195.
DAVIS, E., & WARD, E. Health benefit satisfaction in the public and private sectors: The
role of distributive and procedurale justice. Public Personnel Managem ent, 1995, 24(3) ,255-270.
DARLINGTO N, R. Multiple regre ssion in psychological rese arch and practice . Psychological
Bulletin, 1968, 69, 161-182.
DRE HER, G., ASH, R., & BRETZ, R. Benefit coverage and employee cost: Critical factors
in explaining compe nsation satisfaction. Personnel Psychology, 1988, 41, 237-254.
DRIV ER, R. W. A determination of the relative efficacy of different techniques for employee
benefit communication. Journal of Business Communications , 1980, 17, 23-37.
DYER, L., & THERIAULT, R. The déterminants of pay satisfaction. Journal of Applied Psy-
chology, 1976, 61(5), 596-604.
FARH, J., GRIFFIN, R., & BALKIN, D. Effect of choice of pay plans on satisfaction, goals
setting, and performance . Journal of Organizational Behavior, 1991, 12, 55-62.
FOLGER, R., & GREENBER G, J. Procedural Justice: An interpretive analysis of personnel
systems. Research in Personnel and Hum an Resources Managem ent, 1985, 13, 141-183.
Flexibility and Em ployee Benefit Satisfaction 685
FOLGER, R., & KONOVSKY, M. A. Effect of procedural and distributive justice on reactions
to pay raise decisions. Academ y of Managem ent Journal, 1989, 32(1) , 1989, 115-130.
GERHART, B., & MILKOVICH, G. Employee compe nsation: Research and pratice. In M.
D. Dunette & L. M. Hough (Eds.), Handbook of Industrial and organizational psychology(Vol. 3) . Palo Alto, CA: Consulting Psychologist Press, 1992, pp. 481-569.
GOMEZ-MEJIA, L., BALKIN, D., & CARDY, R. Managing hum an resources . Prentice Hall,1995, 690 pp.
GOMEZ-MEJIA, L., & BALKIN, D. Compensation, organizational strategy, and firm perform -
ance. Cincinnati, OH: South-Western Publishing, 1992.
GOODMAN, P. S. An examination of referents used in the evaluation of pay. OrganizationalBehavior and Hum an Performance, 1974, 12, 170-195.
GREENBE RG, J. The quest for justice on the job. Sage Publications, 1996, 428 pp.
GREENBE RG, J. Organizational justice: Yesterday, today, and tomorrow. Journal of Man-agem ent, 1990, 16(2), 399-432.
HAAR, J., & KOSSACK, S. Employe e benefit packages: How understandable are they? Jour-
nal of Business Comm unications, 1990, 27, 185-200.
HARRIS, M., & FINK, K. Employee Benefits: Future Directions for Research. Pape r pre-
sented at the 52nd Annu al Meeting of Academ y of Man agem ent, Las Vegas, 1992.
HARRIS, M. M., & FINK, K. Employe e benefit programs and attitudinal and behavioraloutcomes: A pre liminary model. Hum an Resource Managem ent Review, 1994, 4(2), 117-
129.
HEMMASI, M., GRAF, L., & LUST, J. Correlates of pay and benefit satisfaction: The uniquecase of university faculty. Public Personnel Managem ent, 1992, 21(4) , 429-442.
HEWITT ASSOCIATES. Canadian flexible benefit programs and practices. He witt AssociatesLLC, 1995, 31.
HILL, F. The relevant other in pay comparisons. Industrial Relations, 1980, 19(3), 345-351.
HORNSBY, J., KURATKO, D., & WALLINGFO RD, C. Flexible benefit plans in smaller
firms. Compensation & Benefits Managem ent, 1991, 7(2) , 14-20.
JAMES, L., & BRETT, J. Mediators, moderators, and tests of a theory. Journ al of AppliedPsychology, 1984, 69, 307-321.
JENKINS, D., & LAWLER, E. Impact of employe e participation in pay plan deve lopment.
Organizational Behavior and Hum an Perform ance, 1981, 28, 111-128.
INSTITU T DE RECHERCHE ET D ’INFO RMATIO N SUR LA RÉ MUNÉ RATIO N
(I.R.I.R.). La com paraison de la rémunération globale des salaires de l’adm inistrationquébécoise et des autres salariés québécois. Dixième rapport dur les constatations de
L’I.R.I.R. Partie 1, Mai 1994.
JUDGE, T. A. Validity of the dimensions of pay satisfaction questionnaire: Evidence of dif-ferential prediction. Personnel Psychology, 1993, 46, 331-355.
KONOV SKY, M., FOLGER, R., & CROPANZ ANO, R. Relative effects of procedural and
distributive justice on employee attitudes. Representative Research in Social Psychology,1987, 17(1), 15-23.
KROLL, M., & DOLAN, J. Cafeteria benefit plans: A conce pt that fits the changing face ofwork. Sam Advanced Managem ent Journal, 1985¯1986, 50¯51, 4-9.
LENGNICK-HALL, M. L., & BEREMAN, N. A. A conce ptual framework for the study of
employee benefits. Hum an Resource Managem en t Review, 1994, 4(2) , 101-115.
LEVENTHAL, G. What should be done with equity theory. In K. J. Gergen, M. S. Gre enberg,& R. H Willis (Eds.), Social exchan ge: Advances in theory and research . New York: Plenum,
1980, pp. 27-55.
LIND, E., & TYLE R, T. The social psychology of procedural justice. Ne w York, Plenum Press,
1988, 267 pp.
LOCKE, E. The nature and causes of job satisfaction. In M. D. Dunette (Ed.), Handbook ofindustrial and organizational psychology, Chicago: Rand McNally, 1976.
LUST, J. Correlates of employee benefit satisfaction. Proceedings of the Sou thern Managem ent
Association 1987 Annual Meeting, 112-114.
LUST, J. The determinants of employee fringe benefit satisfaction: A replication and revision.
Benefits Quarterly, 1990, 6, 89-95.
686 Tremblay, Sire, and Pelchat
LUST, J., & DANEHO WER, C. The Dimensionality of Employee Benefit Satisfaction. Paper
presented at the 1992 Annual Meeting of the Academ y of Managem en t, 1992, 25 pp.
McCAFFERY, R. Employee benefit program s: A total com pensation perspective . Boston: PWS-
Kent, 1988, 250 pp.
MICELI, M., & LANE, M. Antecedents of pay satisfaction: a review and extension. Research
in Personnel and Hum an Resou rces Managem ent, 1991, 9, 235-309.
MILKOVICH, G., & NEWMAN, J. Compen sation (3rd ed.). Boston: BPI Irwin Home wood,
Boston, 1990, 627 pp.
MILKOVICH, G., STURNAN, M., & HANNO N, J. Using expert systems to aid employee s’flexible benefits decisions. ACA Journal, 1994, 3(1), 18-29.
MULVEY, P. Predicting Pay and Benefit Satisfaction: Can One Model Fit All Dimensions?Paper presented at the 52nd Annual Meeting of Academ y of Managem ent, Las Vegas, Au-
gust 1992.
POILPOT-ROCABOY, G. La satisfaction des individus à l’égard de la protection sociale com-pléme ntaire d’entreprise. Revue de G estion des Ressources Hum aines, 1995, 16, 25-35.
RABIN, B. Assessing employee benefit satisfaction under flexible benefit. Compensation andBenefits Managem ent, 1994, 10(3) , 33-44.
RONEN, S. Equity perception in multiple comparaisons: A field study. Hum an Relations, 1986,
39(4) , 333-346.
ROSE, R. Taking control of benefits. Across the Board , 1988, 7 and 8, 49-52.
RYLAN, E., & ROSEN, D. Attracting job applicant with flexible benefits. Personnel, 1988,65(3) , 71-73.
SCARPELLO, V., & JONES, F. Why justice matters in compe nsation decision making. Journalof Organizational Behavior, 1996, 17, 285-299.
SCARPELLO, V., VANDENBERG, R., & HUBER, V., Compensation satisfaction: Its me as-ureme nt and dimensionality. Journal of Applied Psychology, 1988, 73(2), 163-171.
SCHOLL, R., COOPER, E., & McKENNA, J. Referent selection determining equity percep-
tions: Differential e ffects on behavioral and attitudinal outcomes. Personnel Psychology,1987, 40, 113-124.
SHEPPARD, B., & LEWICKI, R. Toward ge neral principles of managerial fairness. Social
Justice Research , 1987, 1(2) , 161-176.
SIRE, B. G estion stratégique des rémunérations, France: É ditions liaisons, 1993.
SHEPPARD, B., LEWICKI, R., & MINTO N, J. Organizational Justice. Ne w York: Lexington
Books, 1992, 227 pp.
SUMMERS, T., & HENDRIX, W. Modelling the role of pay equity perceptions: A fieldstudy. Journal of Occupation al Psychology, 1991, 64, 145-157.
SUMMERS, T., & DENISI, A. In search of Adams’ other: Reexamination of referents usedin the evaluation of pay. Hum an Relations, 1990, 43(6) , 497-511.
SWEENEY, P. Distributive justice and pay satisfaction: A field test of an equity theory pre-
diction, Journ al of Business and Psychology, 1990, 4(3) , 329-341.
SWEENEY, P., & McFARLIN, D. Workers evaluation and the end and the means: An evalu-
ation of four models of distributive and procedural justice. Organizational Behavior andHum an Decision Processes, 1993, 55, 23-40.
THERIAULT, R. G uide Mercer sur la gestion de la rém unération: Théorie et pratique. Chicou-timi: G. Morin, 1991, 589 pp.
THIBAULT, J., & WALKER, L. Procedural justice: A psychological analysis. Hillsdale, NJ:Erlbaum, 1975, 145 pp.
TREMBLAY, M., & ROUSSEL, P. Modelling the Role of Organizational Justice: Satisfactionand Attitudes to Collectice Action. Working paper, Hautes É tudes Commerciales, Mont-
real, 1996, 27 p.
TREMBLAY, M., St.-ONGE, S., & TOULO USE, J. M. Determinants of salary referents rele-vance: A field Study of Managers. Journal of Business and Psychology, 1997, 11(4) , 463-484.
WILLIAMS, M. C. Antecedents of employee benefit leve l satisfaction: A test of a model.
Journal of Man agem ent, 1995, 21(6), 1097-1128.
WEINE R, N. Determinants and behavioral conseque nces of pay satisfaction: A comparison
of two models. Personnel Psychology, 1980, 33, 741-757.
Flexibility and Em ployee Benefit Satisfaction 687
BIOGRAPHICAL NOTES
MICHEL TREMBLAY is Associate Professor of Human Resource Management at the HEC
Business School of Montreal and researcher at CIRANO. He graduated from the manage mentdoctoral program at the University of Aix-Marseille. Professor Tremblay’s rese arch interests
include organizational justice, skill pay, care er manage ment, and workplace innovations. Hismost recent work has been published in Hum an Relations, G roup & Organizational Manage-
ment, and Journal of Business and Psychology.
BRUNO SIRE is Full Professor of Human Resource Management at the French Unive rsity
of Toulouse I. He is currently associated with the L.I.R.H.E., a rese arch unit affiliated to theNational Center for Scientific Research. His research interests include compensation manage-
me nt and competence developme nt. He is also President of the French Human ResourceManageme nt Association (AGRH) .
ANNIE PELCHAT received her MSc in Human Resource Manage ment from the HEC Busi-
ness School in Montreal.
688 Tremblay, Sire, and Pelchat