BUDGETARY PARTICIPATION, MOTIVATION AND MANAGERIAL PERFORMANCE

Working paper alfred p. Sloan school of management



BUDGETARY PARTICIPATION, MOTIVATION
AND
MANAGERIAL PERFORMANCE



by



WP 1389-83



Peter Brownell

and

J. Morris Mclnnes



January 1983



MASSACHUSETTS

INSTITUTE OF TECHNOLOGY

50 MEMORIAL DRIVE

CAMBRIDGE, MASSACHUSETTS 02139



Budgetary Participation, Motivation And Managerial Performance

by

Peter Brownell
and
J. Morris Mclnnes
WP 1389-83 January 1983



Abstract

The paper proposes an "indirect-effects" model linking participation in
budget-setting with managerial performance. Motivation was selected as an
intervening variable to explore the proposed model in a field study
setting. A significant positive association between participation and
performance was observed. However, the indirect path linking participation
and performance through motivation was found to explain only a small amount
of the relationship. In the study, motivation was measured using an
expectancy model. Further analysis revealed that greater participation in
budget-setting was positively associated with the extrinsic components of
motivation, but negatively with the intrinsic components. Contrary to
wide-held beliefs, the results seemed to suggest that participation
reinforces the contractual nature of the budget at the expense of the
personal satisfactions derived from goal-directed behavior and
goal-accomplishment. The managerial implications of the "indirect-effects"
model, and of the findings of the study, are discussed in the final section
of the paper.



Budgetary Participation, Motivation And Managerial Performance

Early theorists (e.g., Argyris, 1952; Becker and Green, 1962) tended to
the view that participation in the budget-setting process was a major means
available to organizations to gain the commitment of managers to budgets,
with consequent improvement in performance. Hopwood (1976) summarized this
earlier sentiment when he suggested that participation was viewed by many as
"...a panacea: a cure for all the many ills which have been associated with
traditional Dudgetary systems" (p. 74). Nonetheless, even in the
non-empirical realm in which much of this debate was conducted disagreement
was evident. The reply to Becker and Green by Stedry (1964), together with
Becker and Green's (19o4) rejoinder, illustrates the controversial nature of
the debate.

As subsequent empirical evidence accumulated, it became clear that these
earlier disagreements were well founded, l^ile evidence emerged of positive
relationships between participation and criteria such as job satisfaction
(Cherrington and Cherrington, 1973), attitude towards job (Milani, 1975),
motivation (Hofstede, 1967; Searfoss and Monczka, 1973), and attitude
towards the budget system (Collins, 1973; Kenis, 1979), no consistent
relationship with performance was found.

A possible explanation for the lack of an observed association between
participation and performance revolves around the role of moderating
variables, that is, variables which interact with participation jointly to
affect performance. This view gained formal recognition in the literature
with the emergence of the contingency framework for control system design
(see, for example, Waterhouse and Tiessen, 1978; Otley, 1980). Specifically
with regard to participation, Brownell (1982a) has reviewed and synthesized
the literature within a framework which explicitly provides for interactive effects of participation with organizational, interpersonal, and individual
variables. Recent empirical work (Brownell 1981, 1982b) has produced
results consistent with this "moderated-ef f ects" framework.
A second possible explanation, and the one pursued in this paper,
emerges from literature in organizational behavior. In a review of
literature in the area of goal setting and routine task performance, Locke
and Schweiger (1979) could find no evidence of a systematic association
between participation and performance. Locke, et. al. (1981, p. 138) went
on to propose that to the extent it is found to be associated with
performance, participation may be influencing this through the joint
mechanisms of inducing higher goals and enhanced goal-commitment on the part
of workers. In other words, these authors imply an "indirect-effects" model
in which some other condition, aroused by participation, is positively
associated with performance.

The difference between the "moderated-ef f ects" and "indirect-effects"
frameworks is significant, and deserves brief elaboration. The "moderated-
effects" model posits the existence of variables which may interact with
participation to affect performance; but these variables are viewed as
neither directly affecting performance, nor being affected by
participation. To illustrate, Brownell (1981) found that the individual
variable, internal/external locus of control , interacted with
participation to affect performance. Internals were found to prefer, and
perform better, under conditions of high participation while externals
preferred, and performed better, under low participation conditions.
However, participation and locus of control were themselves unrelated, and
no effects of the latter on performance were found. By contrast, the
"indirect-effects" model posits the existence of variables which are
associated with participation and which, in turn, affect performance. In

-2-



this framework, it is possible that statistically significant associations
between participation and some intervening variable on the one hand, and
between the intervening variable and performance on the other, could combine
in such a way as to yield an insignificant association between participation
and performance.

In this study, motivation was chosen as an intervening variable in an

2
empirical investigation of the indirect-effects model. Apart from the

intuitive plausibility of motivation as an intervening variable, there is

ample justification in the literature for its choice. This is dealt with in

the next section of the paper.

Review of Previous Literature

Hofstede (1967), in his seminal study of budgeting, observed that, "Of
all variables studied, budget participation is the one with the strongest
effect on all measures of motivation" (pp. 181-182). Hofstede used four
measures of motivation, dealing with the perceived relevance of the budget
and the existence of favorable attitudes towards the budget on the part of
budgeted managers. Similar results were found by Searfoss and Monczka
(1973) who studied the effects on motivation of participation by general
foremen in the development of their budgets; motivation was measured through
the use of subordinates' ratings of the level of effort expended by the
foremen in goal-directing and evaluative activities based on the budget.

In both these studies the concept of motivation was operationalized in
terms' of acceptance of, and effort expended to achieve the budget. In this
study motivation is operationazlied by use of the expectancy model (see, for
example, Galbraith and Cummings [1967], House [1971], and Ronen and
Livingstone [1975]). The model, shown in Equation (1), is considerably more
general in its interpretation of motivation than the constructs used by
Hofstede and by Searfoss and Monczka. It posits that individuals will exert

-3-



Equation 1



Expectancy Model of Motivation:



n

M = IV, + P, (IV +1 P„. EV.)
b 1 a . , 2i 1
1=1

Where: M is motivation

Ivjj is the intrinsic valence associated with
goal-directed behavior

Px is the instrumentality, or probability, that
goal-directed behavior will result in goal accomplishment

IV^ is the intrinsic valence associated with goal accom-
plishment

P2i is the expectation, or probability, that goal accom-
plishment will result in the ith extrinsic outcome,
i = 1, ,n.

EVi is the valence of the ith extrinsic outcome,
i = 1, , n.



effort towards goal achievement to the extent that such effort is perceived
likely to be instrumental in achieving the goal, and additionally, to the
extent that positively valent (desirable) outcomes are perceived likely to
result from goal-directed behavior and from goal-achievement.

The intrinsic valences, or rewards, arise from within the individual -
essentially the personal satisfactions derived from goal-directed behavior and
goal-accomplishment. The extrinsic valences are proferred by some agent
external to the individual, such as the organization (e.g. additional pay,
promotion, etc.), peers, subordinates, or parties outside the organization.

Some possible effects of budgetary participation on the elements of the
expectancy model have been suggested by Ronen and Livingstone (1975). They
speculate, for example, that ego-involvement associated with pursuing a



-4-



self-set goal will enhance IV , and possibly also IV, . To the extent this
is true, however, attendant adverse conseqences for the EV's may arise,
following from attribution theory (see, for example, Staw, [1980] and
Kruglanski 11978]), which posits that extrinsic and intrinsic rewards are in
some state of psychological balance in respect to an individual's
self-perception. In other words, a strengthening of the attribution of effort
to the receipt of intrinsic rewards is a possible consequence of the removal
or reduction of extrinsic rewards, and vice versa.

A relationship of participation with P is also likely. Raia (1965) has
suggested that more difficult goals emerge from participative budgeting. This
might translate into reduced perceptions of P , but might increase IV^ due
to the enhanced internalized value of achieving more difficult goals (Locke
et. al., 1981, pp. 127 - 129).

How, or whether these opposing tendencies balance out in their effects on
LDotivation is unclear. Overall, however, an assessment of the possible
effects of participation suggests that there are more opportunities for
enhancing than for diminishing motivation. This judgment is consistent with
the empirical findings referred to earlier.

Turning to a consideration of the effects of motivation on performance,
this question is dealt with largely within the domain of organizational
behavior. Nevertheless, in the accounting literature some attention has
recently been given to the issue (Ferris, 1977; Rockness, 1977). Ferris,
using a relatively simple construction of motivation, found inconclusive
results regarding the association between motivation and performance. On the
other hand, Rockness, in an experimental study, and using a more complex
formulation of the expectancy model, found relatively strong support for a
positive relationship.

In the organizational behavior area, much of the research examining the

-5-



relationship between motivation and performance stems from the work of Porter
and Lawler (1968). They posit that motivation should contribute to an
explanation of performance and also, under specified conditions, to job
satisfaction. A great deal of subsequent research has been conducted and is
summarized in several review articles (Mitchell, 1974; Wahba and House, 1974;
Connolly, 1970). llitchell (1979) concludes that most of the research results
support a positive relationship between motivation and performance.

Method

Data for the study were collected by a survey questionnaire from 224
middle-level managers drawn from three separate companies - two in the
electronics industry and one in the steel industry. The managers held
positions in a variety of functions, including marketing, production,
research, and administration. One selection criterion was used, namely that
activities of the managers should be controlled, at least in part, via the use
of budgeting. Final sample selection was left to top management of each
company; thus the sample was not strictly random.

The questionnaire measured participation in budget-setting, motivation and
performance.

Budgetary Participation

The participation measures developed by Milani (1975) and by Hofstede
(1967) were both employed in this study. The Milani measure is a six item
Likert type scale, each item calling for a response from one to seven. The
scale is designed for an additive construction of the overall score, and a
previously performed factor analysis of the scale (Brownell, 1982b) provides
adequate confirmation of the single factor nature of the measure. The
Cronbach alpha reliability coefficient computed from its use in this study

-6-



was 0.76.

The Hofstede measure is an eight-point, single-choice scale, with full
verbal anchoring. It was used in this study simply to validate the Milani
measure, with which it correlated at 0.59 (p
Motivation

The questionnaire items were designed to elicit measures on each construct
of the model presented in Equation (1). In the following discussion, we first
address valences, and then instrumentalities/expectancies.

Valences: The approach to the measurement of the valences was adapted

from the procedure developed by Lawler and Suttle (1973) in such a way as to

distinguish between the three classes of valence (IV,, IV , and EV's).

b a

Seventeen outcomes were used, classified a priori as either intrinsic (eight
outcomes) or extrinsic (nine outcomes). The outcomes are listed in Appendix

A. For each outcome, respondents were twice asked to indicate on a scale from

3

one to nine (extremely desirable to extremely undesirable) the strength of

their preference for that outcome. First respondents were asked to value each
outcome as it might result from "working hard" (goal-directed behavior), and
then to value each outcome as it might arise from "meeting or beating budgeted
goals" (goal-accomplisl"UDent) . IV, was measured by averaging the eight
responses to the intrinsic items from the first set of responses. IV was
measured by averaging the responses to the same eight items, but from the
second set of responses. The use of the same eight intrinsic outcomes in the
measurement of IV, and IV pre-supposes that a given outcome may arise
from either goal-directed behavior or goal-accomplishment; and moreover, that
the value placed on it might vary depending on how it arises.

The valences associated with the nine extrinsic outcomes (EV's) were
assessed from the second set of responses (outcomes resulting from

-7-



feoal-accomplishment) •

Instrumentalities/Expectancies: Equation (1) requires the assessment of
nine specific instrumentalities associating goal-accomplishment with extrinsic
outcome (Pn's). These instrumentalities were assessed by asking respondents
to indicate on a scale from one (never) to seven (always), how often "meeting
or beating the budget" would result in each outcome. Three additional
questionnaire items elicited measures of P-, , the perceived probability that
goal-directed behavior would result in goal-accomplishment. For ease of
interpretation, the instrumentality responses were converted to probabilities
in the range of zero to one. The three assessments of P, were significantly
(p < 0.01) correlated, with a mean correlation of 0.35. A single measure of
P, was derived by averaging the three responses.

The final measure of motivation was then obtained by aggregating the
individual constructs following Equation (1).

Performance

A self-rating version of the performance measure developed by Mahoney
et.al. (1963, 1965) was used. The measure calls for ratings on a nine-point
scale for each of eight separate dimensions of performance , together with a
single overall rating. In developmental work, Mahoney et al. (1963 pp.
106-107) indicated that the eight separate dimensions were conceived of as
being independent, and that jointly, the dimensions should explain at least 55
per cent of the overall rating (the remainder being explained by job-specific
factors). Each of these claims was tested in the current study. To assess
dimensional independence, a rule of thumb suggested by Pindyck and Rubinfeld
(1976, p. 68) was used. Multi-collinearity among an independent variable set
is likely to be a concern if the sample correlation between two of the
independent variables (dimensions) is larger than the correlation of either or

-8-



both with the dependent variable (in this case, the overall rating). Of 28
possible comparisons, only three intercorrelations violated this criterion,
indicating that the eight dimensions are reasonably independent. To test the
second claim, the overall performance rating was regressed on the eight
sub-dimensions. The regression explained 78 percent of the variance in the
overall ratings, implying that only 22 per cent of the functions critical to
effective performance were job-specific in the current sample.

Approach to Anal ysis

Because of the hypothesized collinearity between participation and
motivation, ordinary multiple regression cannot be used in the analysis.
Instead, partial regression, in the spirit of path analysis, was used. The
major benefit of path analysis is that it allows a decomposition of the
relationship between two variables in a structural model into the direct
effects of one on the other, as well as the indirect effects of the first on
the second, via one or more intervening variables.

The basic form of the structural model hypothesized earlier is depicted
in Figure 1, where participation, the exogenous variable in the model, is
denoted as X, , motivation as X~, and performance as X^. The path
coefficients in the model are denoted p. ., and R, denotes the
unexplained portions of the endogenous variables, motivation and performance.



-9-




Figure 1: Structural Model



The equations to the structural model are as follows;



Xo =



^21 1 '^Zu u



(2)



X. =



p^^X + p^„X + p R
31 1 32 2 ^3v V



(3)



The solution for p„ , can be directly assessed by computing the
correlation ^-i^' A path coefficient between two variables will equal the
zero-order correlation between them in circumstances where a variable is
viewed to be dependent on a single cause and a residual. This is true for
X (motivation) which, in the model, is conceived of as being dependent
only on X, (participation).

The solutions to p^^ ^"^"^ Pin ^^^ ^^ obtained in several equivalent
ways. One possibility (Stokes, 1974) is by means of an instrumental
variable procedure. This involves multiplying equations by suitable



-10-



instruments, taking expected values, and solving. Suitable instruments are
variables contained in a particular equation which are uncorrelated with the
residual variable of that equation. In equation (3), for example, X and
X^ are suitable, but X-, is not.

A second approach, and the one used here, involves partial regression.
If, as above, it is assumed that the residuals are uncorrelated with the
explanatory variables in the equations in which they appear, and all
variables are expressed in standard form, then the p. .'s can be estimated
by ordinary least squares procedures; they are in fact equal to the
standardized partial regression coefficients. Equations (2) and (3) can be
rewritten as follows:

^2 = e2iX^ + P2, R, (4)

^3= ^31.2^1^^32. 1^2 +P3v^ ^^^

The residual path coefficients can be calculated using the general form:

p. = / (1 - r^)
le e

2
where r is the square of the appropriate multiple correlation

coefficient. In the present context, we have:

P2u = ^ ^^ - ^12^ (^)

P3v - •/ ^1 - ^3.12) <7)

Finally, the decomposition of the total effect of one variable in the

model on another is obtained using the computed path coefficients and
zero-order correlations:

^12 = P2I ^8)

rj_3 = P31 + P32ri2 (9)

r23 = P32 + P3;l^12 ^l^^

-11-



The second components on the right hand side of equations (9) and (10)
capture the indirect effects, while the first components, the path
coefficients themselves, capture the direct effects.



Results

Of 224 questionnaires distributed, 140 were returned, a response rate of
02.5%. Of the 140 returned questionnaires 32 were excluded because of
improper or incomplete responses. The final sample size was therefore 108.

The zero order correlations among the three variables, participation,
motivation, and performance, are shown in Table 1. The probabilities
reported in Table 1 are based on a two-tailed test, as are all the other
results presented in the paper.



INSERT TABLE 1 HERE



The set of path coefficients of equation (3) were computed following
the procedure outlined in the previous section. The approach involves
regressing X on X and X in a multiple regression in which all
variables are standardized. The coefficient on X is p^, , and on X„
is P32* The results of the regression are presented in Table 2. Both
path coefficients, and the overall regression itself, are positive and
significant.



INSERT TABLE 2 HERE



-12-



The path coefficients for the errors in Figure 1 are derived by the

application of the formulae in Equations (6) and (7) to the results obtained:

p., = 0.990
2u

P3^ = U.984

Finally, the decomposition of total effects within the model presented
in Figure 1 is performed following Equations (9) and (10), and the
decomposition is summarized in Table 3.

INSERT TABLE 3 HERE

The results indicate that of the total effects of participation on
performance, the effect through motivation accounts for only a small
proportion, the majority appearing to be direct. The lack of explanatory
power of motivation as a mediator of the effects of participation was, of
course, first hinted at in Table 1, where the zero-order correlation between
the two was found to be insignificant. It is, nonetheless, premature to
conclude that the bulk of the effects of participation on performance are
necessarily direct. Other variables, operating like motivation but
unmeasured in this study, could account for some or all of what in the study
has been accounted for as an apparent direct effect.
Further Analysis and Interpretation

In the aggregate analysis, participation and motivation are not
significantly associated with one another. To examine this finding further,
the associations between participation and the components of the expectancy
model were explored. The results of the analysis are shown in Table 4.



-13-



INSERT TABLE 4 HERE



In the first column the simple correlation coefficients between
participation and the nine extrinsic valences are shown. Four of the nine
are negative and five are positive. There is no a priori reason to suppose
that extrinsic valences should be influenced by the degree of participation,
and this is borne out. However, the P„'s display an interesting pattern
in relation to participation; the coefficients are shown in the second
column. While only five of the nine coefficients are positive, three are
significant, namely, the probability of pay raise, high pay and promotion.
The third column shows the correlation coefficients between participation
and the products of the ^ j' s and the extrinsic valences. It seems as if
participation, largely through its effect on the P^'s, enhances the
contribution of the extrinsic rewards to motivation - particularly those
rewards within the official power of the organization to bestow.

An interpretation of this is that participation serves to legitimize the
budget as a basis for performance evaluation and contingent administration
of organizational rewards. Such a conclusion is consistent with previous
findings (Brownell, 1982b), to the effect that a "budget-constrained" style
of evaluation was appropriate in circumstances of high participation in
budget-setting; conversely, managers were found to respond poorly to
participation when the budget was not used as a salient basis for evaluation.

Column's 4 and 5 show the correlation coefficients between participation

and the intrinsic valences, the eight IV 's and the eight IV, 's. For

a D

both of these groups, six of eight of the coefficients are negative (p <

-14-



0.11, using a binomial test); overall, 12 of 16 correlations between
participation and intrinsic valences are negative (p < 0.03). Contrary to
the belief that participation is likely to enhance goal-commitment by
increasing the level of ego-involvement associated with the pursuit of
self-set goals, these results suggest that it appears to be acting in the
opposite manner. Indeed, from an intrinsic perspective, participation
appears to be associated with significant devaluation of the task and
reduction of the satisfaction derived from goal-accomplishment. In line
with the previous suggestions (Staw [1980] and Kruglanski [1978]),
strengthening of the instrumentality between goal-accomplishment and
extrinsic rewards appears to diminish, in some degree, the intrinsic
valences.

The correlations between participation and the three measures of P,
(the probability that goal-directed effort will lead to goal accomplishment)
were all found to be positive. Participation correlates with the overall
measure of P-, at 0.18 (p < 0.08). Two explanations of this would be
plausible. First, perhaps participation acts to strengthen goal commitment,
effort, and therefore the perceived probability of goal-achievement. There
is considerable evidence from the organization! behavior literature (see
Locke et al., 1981) that the intrinsic value of goal accomplishment tends to
increase as the level of goal difficulty increases. Indeed, this mechanism
is central to the notion of goal-setting and motivation. In the current
study, however, the negative association between participation and the
intrinsic valences relating to goal-accomplishment is at odds with this
explanation. A second explanation is that participation may provide an
opportunity to negotiate easier goals; that is, budget slack may be
introduced by participation (Schiff and Lewin, 1970; Onsi, 1973). This
would be consistent with the observed negative association of participation

-15-



with the IV 's, since achieving an easier goal would be likely to lead to
less intrinsic satisfaction. Moreover, since participation appears to be
connected with the use of the budget to evaluate performance as a basis for
bestowing organizational rewards, it would be perfectly rational for
managers to perceive it to be in their self-interest to negotiate as low a
budget as possible (Lowe and Shaw; 1968).

The preceding interpretation of the research findings raises important
questions about the benefits of participation in budget-setting.
Nonetheless, the study does indicate a significant positive relationship
between participation and the overall measure of performance. To explore
this further, each of the eight sub-dimensions of performance was correlated
with participation, and the results are shown in Table 3.



INSERT TABLE 5 HERE



All eight correlations are positive, and significant (p < 0.05). An
examination of the pattern among the coefficients suggests that the
dimensions of managerial performance which are most strongly associated with
participation are precisely those which might be expected to be strengthened
by judicious use of a budget as a managerial tool, in particular planning,
investigating and evaluating.
Limitations and Conclusions

The aim of the study was to assess whether, compared with a
"direct-effects" model, an "indirect-effects" model linking participation
and performance has the potential to add to our understanding of the
consequences of participation. In the interests of parsimony, a single variable, motivation, was included as intervening between participation and
performance. The results of the study show that the indirect path through
motivation contributes relatively little to an explanation of the observed
association between participation and performance. This evidence, however,
is insufficient grounds for abandoning the "indirect-effects" model. Other
intervening variables could enter the model. For example, ability and role
clarity have been suggested as co-determinants, with motivation, of
performance (Porter and Lawler, 1968; Lawler, 1971). Further research is
necessary to resolve this matter.

The form of the expectancy model of motivation adopted in the study also
requires further research and possible refinement. For example, Staw (1977)
offers a rationale for recognizing the possibility of extrinsic valences
being associated with goal-directed behavior, rather than simply with
goal-accomplishment. Moreover, multiple extrinsic valences enter into the
equation additively, whereas both the intrinsic valences enter the equation
as single mean values. This has implications for the relative weightings
accorded extrinsic and intrinsic valences respectively in the calculation of
motivation.

Then there is the question of the appropriateness of the measure of
performance in a research study such as this one. The criterion variable of
ultimate concern is an organization's creation of profit potential and the
realization of this through efficient operations. However, budgeting is
primarily concerned with the regulation of managerial work within an
organization. In this limited context, the appropriate criterion variable
of direct interest is managerial performance, even if the connection between
managerial performance and organizational effectiveness remains a broader
concern. This still leaves the issue of the reliability of self-ratings of
performance, compared with superior, peer or subordinate ratings. While

-17-



self-ratings have been shown to exhibit leniency error, provided the error
is not systematic among the respondents in a manner which is collinear with
any of the independent variables, the research findings will be unaffected.
Furthermore, Kavanagh et al. (1971) have suggested that self-ratings provide
greater discrimination among dimensions of performance than do other sources
of rating, and may therefore be more reliable in analyzing the components of
performance which contribute to the overall assessment. Having said this,
however, exhortations (e.g. Steers, 1977) to devote more research effort to
the development of better techniques for measuring performance are clearly
valid.

Finally, there is the issue of causality and the direction in which
effects among variables in the model proceed. The analysis done in the
study does not permit any inference about the causalities; these can only be
constructed from supporting theory. In this study, it is implied that
perceived participation in budget-setting may cause higher levels of
motivation, which in turn may cause higher levels of performance. But, in a
dynamic, inter-temporal context a rationale could be constructed to argue
for a reverse chain of causality - i.e. high performance leading to high
motivation leading to a perception (and indeed perhaps a reality) of high
participation in budget-setting. For example, it is entirely possible that
participation might be viewed as a luxury affordable only to organizational
units already performing satisfactorily.

i^ven recognizing the limitations of the study, the findings are
potentially of significance in the design of control systems. Participation
is a costly control strategy, since it consumes managerial resources. The
findings of the study indicate that the control system cost may be
compensated by improved managerial performance. But, on the other hand, the
findings also indicate some latent costs in the participation process,

-18-



arising in two ways. First, some potential performance may be being
sacrificed through the creation of budgetary slack. And secondly, costly
organizational rewards, contingent on the achievement of budgets, may be
supplanting intrinsic rewards. These latter rewards could be viewed as
being costless to the organization; stated even more strongly, their
suppression could induce hidden costs through, for example, diminished
creativity within the organization.

It is precisely because of these hidden consequences that the
"indirect-effects" model is potentially of considerable practical, as well
as theoretical, interest. A greater understanding of the mechanisms through
which participation affects performance should provide management with a
basis for designing the participation process in such a way as to capture
all the benefits, and to drive out the latent costs which the results of
this study have indicated.



-19-



TABLE 1
Zero-Order Correlations Between Variables

Participation(Xi)/Performance(X3) , r = 0.369, pParticipation(X]^)/Motivation(X2) , r = 0.139, n.s.
Motivation(X2)/Perforroance(X3) , r = 0.248, p
TABLE 2
Calculation of Path Coefficients P3-]^ and P32



Standard
Coefficient Value Error t



Intercept


0.000


0.088


0.00


n.s.


P31


0.341


0.089


3.81


<0.01


P32


0.201


0.089


2.25


<0.05



r2 = 0.176, F2^i05 = 11.20, p <0.01







TABLE


3






De^


com


iposition of Total


Effects in


Mod


el


Linkage




Direct (=p)


Indirect




Total (=r)


X1/X2




0.139







0.139


X1/X3




0.341


0.028




0.369


X2/X3




0.201


0.047




0.248



-20-



TABLE 4
Correlations between Participation and
the Elements of the Expectancy Model



Outcome*


EV.

1


^2i


^2i^^


IV
a


'\


1


.143


.232


.232


.025


.041


2


.075


.196


.159


-.033


-.072


3


.060


-.052


.037


-.042


.086


4


-.141


-.049


-.106


-.136


-.112


5


-.201


-.097


-.136


-.016


-.058


6


-.010


.104


.001


-.133


-.044


7


.049


-.201


.057


.011


-.098


8


-.024


.070


.061


-.042


-.155


9


.128


.192


.206


N/A


N/A



* The numbered outcomes are listed in Appendix A.

N/A Not Applicable

.191 > r > .160, p < 0.10

.248 > r > .191, p < 0.05

r > .248, p < 0.01



-21-



TABLE 5
Correlations between Participation and
the Eight Sub-Dimensions of Performance



Planning


0.297


Investigating


0.219


Coordinating


0.209


Evaluating


0.30A



Supervising


0.201


Staffing


0.284


Negotiating


0.217


Representing


0.174



-22-



Appendix A



List of Outcomes



Eight intrinsic and nine extrinsic outcomes were used in this study.



Extrinsic Intrinsic

1. Pay Raise 1. Personal growth and development

2. High Pay 2. Setting higher standards for yourself

3. Respect from boss 3. Giving help to others

4. Respect from other employees 4, Time at work passing fast

5. Receiving more compliments 5. Feelings of security '

6. Greater chances for independent 6. Setting higher standards for others
thought and action

7. Fewer chances to make friends 7. Feelings of accomplishment

8. Special reward or recognition 8. Being tired

9. Promotion



-23-



Appendix B

The solution to equation (3) can be derived through the use of
instrumental variables. As suggested earlier, X, and X„ serve as
suitable instruments since they are assumed uncorrelated with R .

Multiplying both sides of (3) by X, and then by X„ gives:

hh = P3l4 ^ P32V2 ^ ^3v\K ^'^

'"' hh - ^31^1^2-^ P32^2 ^ P3v^2\ ^^^^

Taking the expected value of each of (i) and (ii) gives

r^3 = P31 + P32ri2 ^^^^^

^^23 = P31^12 + P32 ^^^^

(iii) and (iv) follow from (i) and (ii) since, for standardized

variables,

(a) E(X.X.) = r. .

^'^^ (b) E(X^) = 1.

1

The terms involving R disappear since

E^^i^v) = ^X.R = 0-
1 V

(iii) and (iv) represent a system of two equations in two unknowns
(poi and Pon)« Solving (iii) and (iv) simultaneously gives

_ ^^13 "^ ^^12^23 (v)

P3I \ 2

1- r^2



P



^ ^23 "^ ^12^13 (vi)

32 , 2
1 - r^2



-24-



Substituting the values from Table 1 for the r's in (v) and (vi) gives
p, = 0.341, and ^^^ " 0*201, values identical to those shown in Table 2.



■25-



Footnotes

1. An internally controlled individual attributes the outcomes of his o\^m
actions to himself, while an externally controlled individual tends to
attribute these outcomes to chance, luck, or fate. See Rotter [1966]
for details of this variable, together with the instrument used to
measure it.

2. Of course, we do not intend to imply, by its choice as a single
intervening variable, that motivation is the only candidate for
inclusion.

3. These raw scores were reversed and rescaled by subtracting five from all
scores. As a result of this procedure, "neutral" responses score zero,
while responses in the "desirable" direction score positively (one to
four) and responses in the "undesirable" direction score negatively
(minus one to minus four).

4. The dimensions were planning, investigating, coordinating, evaluating,
supervising, staffing, negotiating and representing.

5. This procedure is also used to test the suitability of more parsimonious
models which exclude one or more linkages. In general, the assessment
of suitability is based on whether the original matrix of
intercorrelations among the variables in the model can be reproduced by
a more parsimonious model. For elaboration, see Land (1969).

6. As previously mentioned, it was not necessary to solve equation (2)
since P2i is equivalent to r]^2> the zero-order correlation between
participation and motivation.

7. The equivalence of the coefficients derived here with those resulting
from the instrumental variable procedure is demonstrated in Appendix B.



-26-



References



Argyris, C. , The Impact of Budgets on People , Ithaca, NY: School of Business
and Public Administration, Cornell University, 1952.

Becker, S. W. and D. Green, "Budgeting and Employee Behavior," Journal of
Business (October 1962), pp. 392-402.

, "Budgeting and Employee Behavior: A Rejoinder to



a 'Reply'," Journal of Business (April 1964), pp. 203-205.

Brownell, P., "Participation in Budgeting, Locus of Control, and

Organizational Effectiveness," The Accounting Review (October 1981),
pp. 844-860.

, "Participation in the Budgeting Process: When it Works and



When it Doesn't," Journal of Accounting Literature , (Vol. 1, 1982a),
pp. 124-153.

, "The Role of Accounting Data in Performance Evaluation,



Budgetary Participation and Organizational Effectiveness,"
Journal of Accounting Research (Spring 1982b), pp. 12-27.

Cherrington, D. J. and J. 0. Cherrington, "Appropriate Reinforcement
Contingencies in the Budgeting Process," Empirical Research in
Accounting: Selected Studies , Supplement to Journal of Accounting
Research (1973), pp. 225-253.

Collins, F., "The Interaction of Budget Characteristics and Personality
Variables with Budgetary Response Attitudes," The Accounting Review
(April 1978), pp. 324-335.

Connolly, T., "Some Conceptual and Methodological Issues in Expectancy Models
of Work Performance Motivation," Academy of Management Review
(No. 1, 1976), pp. 37-47.

Ferris, K. R. , "A Test of the Expectancy Theory of Motivation in an

Accounting Environment," The Accounting Review (July 1977), pp. 605-615.

Galbraith, J. and L. L. Cummings, "An Empirical Investigation of the

Motivational Determinants of Task Performance: Interactive Effects
between Instrumentality-Valence and Motivation-Ability,"
Organizational Behavior and Human Performance , (No. 2, 1967),
pp. 237-257.

Hofstede, G. H., The Game of Budget Control , Assen: Van Gorcum, 1967.

Hopwood, A. G. , Accounting and Human Behavior , Englewood Cliffs, N.J.:
Prentice-Hall, 1976.

House, R. J., "A Path-Goal Theory of Leader Effectiveness," Administrative
Science Quarterly (September 1971), pp. 321-338.



-27-



Kavanagh, M. J., A. C. MacKinney and L. Wolins, "Issues in Managerial

Performance: Multitrait-Multimethod Analyses of Ratings," Psycholgica l
Bulletin (January 1971), pp. 34-49. ""

Kenis, I., "Effects of Budgetary Control Characteristics on Managerial
Activities and Performance," The Accounting Review (October 1979),
pp. 707-721.

Kruglanski, A. W., "Issues in Cognitive Social Psychology" in H. R. Lepper
and D. Greene (eds.) The Hidden Costs of Reward; New Perspectives on
the Psychology of Human Motivation . New York: Wiley and Sons, 1978,
pp. 19-29.

Land, K. C, "Principles of Path Analysis," in E. Borgatta (ed.).

Sociological Methodology , San Francisco: Jossey-Bass, 1969, pp. 3-2 7.

Lawler, E. E., Motivation in Work Organizations , Belmont, California:
Brooks/Cole, 1973.

Lawler, E. E. and J. L. Suttle, "Expectancy Theory and Job Behavior,"

Organizational Behavior and Human Performance (June 1973), pp. 482-503.

Locke, E. A. and D. M. Schweiger, "Participation in Decision-Making: One
More Look," in B. M, Staw (ed.). Research in Organizational Behavior
(Vol. 1), Greenwich, Ct.:, JAI Press, 1979, pp.

, K, N. Shaw, L. M. Saari and G. P. Latham, "Goal-Setting and



Task Performance: 1969-1980," Psychological Bulletin (No. 1, 1981),
pp. 125-152.

Lowe, A, E, and R. W, Shaw, "An Analysis of Managerial Biasing: Evidence
from a Company's Budgeting Process," Journal of Management Studies
(October 1968), pp. 304-315.

Mahoney, T. A., T. H. Jerdee and S. J. Carroll, Development of Managerial
Performance: A Research Approach , Cincinatti: Southwestern Publishing
Company, 1963.

, "The Jobs of Management,"



Industrial Relations (February 1965), pp. 97-110.

Milani, K. W. , "The Relationship of Participation in Budget-Setting to
Industrial Supervisor Performance and Attitudes: A Field Study,"
The Accounting Review (April 1975), pp. 274-284.

Mitchell, T. R. , "Expectancy Models of Job Satisfaction, Occupational

Preference and Effort: A Theoretical, Methodological, and Empirical
Appraisal," Psychological Bulletin (81, 1974), pp. 1053-1077.

., "Organizational Behavior," Annual Review of Psychology



(1979), pp. 243-281.

Onsi, M. , "Behavioral Variables Affecting Budgetary Slack," The Accounting
Review (July 1973), pp. 535-548.



-28-



Otley, D. T. , "The Contingency Theory of Management Accounting: Achievement
and Prognosis," Accounting, Organizations and Society (1980),

pp. 413-428.

Pindyck, R. S. and D. L. Rubinfeld, Econometric Models and Economic
Forecasts , New York: McGraw-Hill, 1976.

Porter, L. W. and E. E. Lawler, Managerial Attitudes and Performance ,
Homewood, Illinois: Irwin, 1968.

Raia, A. P., "Goal Setting and Self-Control: An Empirical Study,"
Journal of Management Studies (No. 1, 1965), pp. 34-53,

Rockness, H. 0., "Expectancy Theory in a Budgetary Setting: An Experimental
Examination," The Accounting Review (October 1977), pp. 893-903.

Ronen, J. and J. L. Livingstone, "An Expectancy Theory Approach to the

Motivational Impact of Budgets," The Accounting Review (October 1975),
pp. 671-685.

Rotter, J. B., "Generalized Expectancies for Internal versus External Control
of Reinforcement," Psychological Monographs: General and Applied
(1966, Whole No. 609), pp. 1-28.

Schiff , M. and A. Y. Lewin, "The Impact of People on Budgets," The Accounting
Review (April 1970), pp. 259-268.

Searfoss, D. and R. Monczka, "Perceived Participation in the Budget Process
and Motivation to Achieve the Budget," Academy of Management Journal
(December 1973), pp. 541-554.

Staw, B. M. , "Motivation in Organizations: Toward Synthesis and

Redirection," in B. M. Staw and G. R. Salancik (eds.). New Directions
in Organizational Behavior , Chicago: St. Clair Press, 1977.

, "Intrinsic and Extrinsic Motivation," in Leavitt, H. J., L. R.



Pondy and D. M. Boje, Readings in Manger ial Psychology , University of
Chicago Press (1980), pp. 23-61.

Steers, R. M. , Organizational Effectiveness: A Behavioral View , Santa
Monica, Calif.: Goodyear Publishing, 1977.

Stokes, D. E. , "Compound Paths: An Expository Note," American Journal of
Political Science (February 1974), pp. 191-214.

Stedry, A. C. , "Budgeting and Employee Behavior: A Reply,' Journal of
Business (April 1964), pp. 195-202.

Wahba, M. A. and K. J. House, "Expectancy Theory in Work and Motivation -
Some Logical and Methodological Issues," Human Relations (27, 1974),
pp. 121-147.

Waterhouse, J. H. and P. Tiesssn, "A Contingency Framework for Management
Accounting Systems Research," Accounting, Organizations and Society
(1978), pp. 65-76.

-29-



kSS3 019



v^



MIT LIBRARIES



3 TDflD DD4 S7h 143



"SM^Aobe: i5 oH dA^ (^^^^^sf^
Brownell (1981) found that the individual An interpretation of this is that participation serves to legitimize the budget as a basis for performance evaluation and contingent administration of organizational rewards
Untuk menggambarkan, Brownell ( 1981) yang ditemukan [bahwa/yang] yang individu [Adalah] suatu penafsiran ini adalah keikutsertaan itu melayani untuk mengesahkan anggaran [itu] sebagai basis untuk evaluasi capaian dan administrasi ketidak-tentuan [dari;ttg] penghargaan organisatoris. Kesimpulan seperti itu adalah konsisten dengan penemuan sebelumnya