Content uploaded by Chris Stride
Author content
All content in this area was uploaded by Chris Stride
Content may be subject to copyright.
Transformational Leadership and Employee Safety Performance:
A Within-Person, Between-Jobs Design
Michelle Inness
University of Alberta
Nick Turner
University of Manitoba
Julian Barling
Queen’s University
Chris B. Stride
University of Sheffield
We investigated the extent to which the safety performance (i.e., self-reported safety compliance
and safety participation) of employees with 2 jobs was predicted by their respective supervisors’
transformational leadership behaviors. We compared 2 within-person models: a context-specific
model (i.e., transformational leadership experienced by employees in 1 context related to those
same employees’ safety performance only in that context) and a context-spillover model (i.e.,
transformational leadership experienced by employees in 1 context related to those same em-
ployees’ safety performance in the same and other contexts). Our sample comprised 159
“moonlighters” (73 men, 86 women): employees who simultaneously hold 2 different jobs, each
with a different supervisor, providing within-person data on the influence of different supervisors
on employee safety performance across 2 job contexts. Having controlled for individual differ-
ences (negative affectivity and conscientiousness) and work characteristics (e.g., hours worked
and length of relationship with supervisor), the context-specific model provided the best fit to the
data among alternative nested models. Implications for the role of transformational leadership in
promoting workplace safety are discussed.
Keywords: moonlighting, safety, safety compliance, safety participation, transformational leadership
Despite attempts to improve workplace safety,
high rates of job-related injuries persist worldwide
(Ha¨ma¨la¨inen, Takala, & Saarela, 2006), constraining
many employees’ physical capacity to work and neg-
atively affecting their subsequent work-related atti-
tudes and behaviors (e.g., Barling, Kelloway, & Iver-
son, 2003; Roberts & Markel, 2001). Gaining an
understanding of the determinants of employee safety
performance that precede such injuries can poten-
tially facilitate improvements to workplace safety
(Neal & Griffin, 2006).
The first goal of the study was to extend existing
research by investigating the influence of generalized
transformational leadership on employee safety perfor-
mance. In the present study, we conceptualized em-
ployee safety performance as a bidimensional, facet-
specific aspect of job performance. Following prior
research (e.g., Griffin & Neal, 2000), we suggest that
employee safety performance can be operationalized as
two types of safety behaviors: safety compliance and
safety participation. Safety compliance refers to behav-
iors focused on meeting minimum safety standards at
work, such as following safety procedures and wearing
required protective equipment. Safety participation re-
fers to behaviors that support workplace safety, such as
helping coworkers with safety-related issues or volun-
tarily attending safety meetings. As such, safety com-
pliance and safety participation parallel two types of
general work performance: task performance and con-
textual performance, respectively (Borman & Motow-
idlo, 1997).
Our second goal was to examine whether transfor-
mational leadership motivates safety compliance and
participation not only within the work context in
Michelle Inness, School of Business, University of Al-
berta; Nick Turner, Asper School of Business, University of
Manitoba; Julian Barling, Queen’s School of Business,
Queen’s University; and Chris B. Stride, Institute of Work
Psychology, University of Sheffield.
We thank the Social Sciences and Humanities Research
Council of Canada for financial support. Versions of this
study were presented at the 2006 European Association for
Occupational Health Psychology conference in Dublin, Ire-
land; the 2007 Society for Industrial and Organizational
Psychology conference in New York; the 2009 American
Psychological Association’s Annual Convention, Toronto,
Canada; and the 2010 Annual Meeting of the Academy of
Management, Montre´al, Canada.
Correspondence concerning this article should be ad-
dressed to Michelle Inness, Department of Strategic Man-
agement & Organization, School of Business, University of
Alberta, Edmonton, Alberta, T6G 2R6, Canada. E-mail:
michelle.inness@ualberta.ca
Journal of Occupational Health Psychology
2010, Vol. 15, No. 3, 279–290
© 2010 American Psychological Association
1076-8998/10/$12.00 DOI: 10.1037/a0019380
279
which transformational leadership is experienced, but
whether it provides a learning experience that trans-
lates into improved safety compliance and safety
participation in another work context. In so doing, we
investigate whether transformational leadership is a
within-context motivator of employee safety perfor-
mance, whether employees learn to adopt a higher
standard of safety performance that transcends the
work context, or both. We proposed two main mod-
els: one in which the effects of transformational
leadership on employee safety performance are spe-
cific to the work context in which an employee and
supervisor interact (i.e., context-specific), and one in
which transformational leadership in one work con-
text affects employee safety performance across mul-
tiple work contexts (i.e., context-spillover).
Transformational Leadership and
Employee Safety Performance
Transformational leadership has received consid-
erable conceptual and empirical attention in recent
years (Bass & Riggio, 2006). Transformational lead-
ership comprises four leader behaviors. Idealized in-
fluence is when leaders demonstrate high standards of
moral conduct in their own behavior. Inspirational
motivation occurs when leaders communicate a pos-
itive, value-based vision for the future state of the
organization and its employees. Intellectual stimula-
tion is when leaders encourage employees to chal-
lenge organizational norms and think creatively.
Lastly, individual consideration is leaders recognize
the unique needs of followers.
There are several lines of evidence showing that
transformational leadership can influence employee
safety compliance and safety participation. Transfor-
mational leadership can motivate superior employee
task and extrarole performance by creating a positive
vision of the organization’s future, empowering em-
ployees, and placing importance on their needs (Con-
ger & Kanungo, 1998). Empirical findings support
the relationship between transformational leadership
and enhanced task performance (e.g., Barling, We-
ber, & Kelloway, 1996; Howell & Avolio, 1993) and
contextual performance (e.g., organizational citizen-
ship; Koh, Steers, & Terborg, 1995). We argue that
transformational leadership exerts the same effects
on safety-specific performance, such as following
rules and helping improve employee safety, by gen-
erating motivation to achieve positive change and
prioritizing employee well-being. Zohar (2002)
found that employees exposed to transformational
leadership had higher levels of safety compliance (as
measured by earplug use) when compared with a
control group, and Clarke and Ward (2006) found
that transformational leadership was positively re-
lated to employee safety participation.
Several studies have found relationships between
safety-specific transformational leadership (i.e.,
transformational leadership specifically focused on
enhancing individual and organizational safety) and
safety-related outcomes, including perceived safety
climate, safety events, safety consciousness (Kello-
way, Mullen, & Francis, 2006), and safety citizenship
behavior (Conchie & Donald, 2009). However, sev-
eral issues remain with safety-specific transforma-
tional leadership.
First, the salience of safety as an important out-
come in the presence of safety-focused leaders is
understandable; there is longstanding research (e.g.,
Cohen, 1977; Zohar, 1980) showing that organiza-
tions in which leaders take an active role in promot-
ing safety enjoy better organizational safety records.
Safety-specific transformational leadership, when
used as the sole predictor of safety outcomes, con-
founds safety and transformational leadership; the
possibility remains that a safety climate (Wallace &
Chen, 2006; Zohar, 1980) rather than transforma-
tional leadership behaviors per se explains variance
in employee safety performance.
Second, item content is shared across measures of
safety-specific transformational leadership (e.g.,
Barling, Loughlin, & Kelloway, 2002) and employee
safety performance (e.g., Neal, Griffin, & Hart, 2000)
when predictor and criterion variables are collected
from the same source (i.e., employees)—all of the
scales contain derivatives of the word safety in every
item—potentially inflating the relationship between
predictor and criterion. One study (Mullen & Kello-
way, 2009) has tested the relative effects of safety-
specific and generalized transformational leadership
training on employee safety performance, and found
an increase in safety-specific transformational lead-
ership behaviors in the safety-focused training group.
However, it was less clear whether there was an
increase in generalized transformational leadership in
either training group, and the relationship between
generalized transformational leadership and em-
ployee safety outcomes was not reported.
Third, testing the extent to which generalized
transformational leadership predicts employee safety
performance reflects the lived reality of supervisors
whose daily priorities reflect a range of issues, and
not safety alone. Thus, understanding the relationship
between generalized transformational leadership and
280 INNESS, TURNER, BARLING, AND STRIDE
employee safety outcomes not only avoids confound-
ing predictor with criterion, but also more compre-
hensively reflects the way supervisors behave across
a range of work situations and contexts.
Hypothesis 1a: Transformational leadership ex-
perienced by employees in one job is related to
employees’ safety compliance in that same job.
Hypothesis 1b: Transformational leadership ex-
perienced by employees in one job is related to
employees’ safety participation in that same job.
Context-Specific or Context-Spillover?
The question of whether the effects of transforma-
tional leadership on safety compliance and safety
participation are specific to that leader (i.e., context-
specific) or whether transformational leadership ex-
perienced in one job can affect safety compliance and
safety participation in a different job (i.e., context-
spillover) has practical and conceptual relevance.
In terms of practical relevance, 5.2% of employed
people in the United States hold multiple jobs simulta-
neously (U.S. Bureau of Labor Statistics, 2006), and
more than 7% work for different companies in the same
time period or simultaneously work for an employer
and contractor. More generally, individuals increasingly
change jobs over their working lifetime (U.S. Depart-
ment of Labor, 2006). Thus, whether experiences with
a specific leader in one job affect employee behaviors in
another job has social significance.
Conceptually, this question has implications for
the generalizability of transformational leadership.
Specifically, it addresses whether transformational
leadership serves to help elicit safety performance
from employees in a particular context, or whether it
provides more generalized social learning that moti-
vates employees to achieve higher safety perfor-
mance across work contexts. Given that a transfor-
mational leader can generate employee enthusiasm
for the future of the organization and a connection to
the leader, employees may be motivated for high
levels of safety performance exclusively in the con-
text in which that leader–follower relationship exists.
Evidence that individuals’ behavior tends to be con-
sistent within but not across contexts is substantial
(Mischel & Shoda, 1999; Ross & Nisbett, 1991).
This is also consistent with the notion that organiza-
tions are social systems that provide cues for the
appropriateness of behaviors in a specific context
(Salancik & Pfeffer, 1978). With respect to leader-
ship, Lieberman (1956) and Conger (1989) suggested
that individuals distinguish their experiences with
one leader from another leader, and thus the “routi-
nization” of leadership is context-specific.
At the same time, experiences with leaders in one
job may transcend context. Organizational leaders are
socialization agents given that they introduce em-
ployees to the expectations of the workplace and
broader norms about desired employee behaviors
(Jablin, 2001). There is some evidence of sequential
leader transference, whereby prior experiences with
one leader can influence expectations of another
leader, especially if the two leaders are similar (Ritter
& Lord, 2007). Employees who feel motivated by
their leaders to behave in a particular way may feel
encouraged to behave in the same way in other con-
texts, from both a sense of consistency and moral
autonomy (Maclagan, 2007). Thus, there is evidence
for both the context-specific and context-spillover
models, and we examined this as an exploratory
research question.
Within-Person Differences and Employee
Safety Performance
A methodological contribution of this research is
its study design, which enables separation of the
influence of multiple supervisors and individual dif-
ferences on employee safety performance. First, with
respondents providing perceptions of two concur-
rently held jobs, we examined the relationships of
interest while holding within-person differences con-
stant.
Second, we modeled negative affectivity and con-
scientiousness as theoretically relevant constructs re-
lated to employee safety performance. Conscien-
tiousness is a stable trait that manifests itself through
hard work, orderliness, and conformity (Hogan &
Ones, 1997). Meta-analytic evidence demonstrates a
positive relationship between conscientiousness and
safety-specific performance (Christian, Bradley,
Wallace, & Burke, 2009). A positive association has
also been found between conscientiousness and gen-
eral contextual performance (e.g., Witt, Burke, Bar-
rick, & Mount, 2002). As such, conscientious em-
ployees are more likely to exhibit safety
participation.
Negative affectivity refers to the chronic experi-
ence of negative emotional states, feelings of ner-
vousness and worry, and perceiving events in a neg-
ative light (Watson & Clark, 1984). Although no
previous research has focused on a link between
negative affectivity and either safety compliance or
281TRANSFORMATIONAL LEADERSHIP AND SAFETY
safety participation at the individual level of analysis
(cf. Neal & Griffin, 2006), negative affectivity has
been positively related to workplace injuries (Iverson
& Erwin, 1997), as well as with lower task and
contextual performance (Kaplan, Bradley, Luchman,
& Haynes, 2009). We suggest that negative affectiv-
ity is related to safety compliance and safety partic-
ipation.
Method
Participants and Recruitment
We sent recruiting advertisements to 3,600 em-
ployed people via Study Response, an online research
service run by Syracuse University. We used a sam-
ple that cut across many occupations because it was
not possible to isolate a readily identifiable popula-
tion of moonlighters. At the time of data collection,
Study Response had a roster of more than 95,000
members reflecting the U.S. population in terms of
ethnic background and education.
We invited individuals to participate in the study if
they simultaneously held two jobs with a different
supervisor in each job. Of those, 159 people (M
age ⫽37.4 years, SD ⫽10.25; 73 men, 86 women)
self-identified as moonlighters and completed the
questionnaire. Given that 5% of people in the United
States hold two jobs simultaneously (U.S. Census
Bureau, 2006), we estimate that approximately 180
of the 3,600 people contacted were eligible to partic-
ipate, yielding a conditional response rate of 88.3%.
Respondents worked an average of 38.2 hr per week
in their “primary jobs” and 17.8 hr per week in their
“secondary jobs.” Participants’ primary jobs included
blue-collar work (12.8%; e.g., auto repair, farm work,
janitor), clerical work (16.0%; e.g., secretary, recep-
tionist), managerial work (9.6%), professional work
(22.4%; e.g., attorney, engineer, consultant), techni-
cal work (25.6%; e.g., IT jobs, lab technician), sales
and service (11.6%; e.g., sales, retail, customer ser-
vice), and semiprofessional work (2.0%; e.g., chef,
educational counselor). Participants’ secondary jobs
included blue-collar work (20.7%), clerical work
(10.7%), management (7.1%), professional work
(18.6%), technical work (17.9%), sales and service
(15.0%), and semiprofessional work (10.0%).
Study Measures
We asked respondents to identify one of their two
jobs as their primary job and the other as their sec-
ondary job on the basis of the number of hours they
worked per week in each job. Respondents completed
scales of items measuring transformational leader-
ship, safety compliance, safety participation, and
some potentially confounding variables (e.g., safety
concern, hours worked, length of relationship with
supervisor) for each of the jobs; scales measuring
negative affectivity and conscientiousness were com-
pleted once, along with demographic variables (e.g.,
age, gender).
Generalized transformational leadership was mea-
sured with four items from the Multifactor Leader-
ship Questionnaire (Bass & Avolio, 1995) using a
5-point Likert-type scale ranging from 1 (not at all)
to5(frequently or always). Each transformational
leadership behavior (i.e., idealized influence, inspira-
tional motivation, individualized consideration, intel-
lectual stimulation) was measured with one item
(e.g., “Emphasizes the importance of having a col-
lective sense of mission”).
Safety compliance was measured with three items
used by Neal and Griffin (2006). An example item is
“I use the correct safety procedures for carrying out
my job.” Items were rated using a 5-point Likert-style
scale ranging from 1 (strongly disagree)to5
(strongly agree), with higher scores indicating
greater safety compliance.
Safety participation was assessed with three items
used by Neal and Griffin (2006); for example, “I
voluntarily carry out tasks or activities that help to
improve workplace safety.” Items were rated on a
5-point Likert-type response scale ranging from 1
(strongly disagree)to5(strongly agree), with higher
scores indicating greater safety participation.
Recognizing that participants in the present study
hold jobs in a wide variety of occupations, there will
be variability in the extent to which workplace safety
is perceived to be a safety concern in that context. We
controlled for this using a single item for each job,
which asked participants to rate, for that job (on a
4-point response scale: 1 ⫽low concern to4⫽high
concern), the extent to which safety is a concern
(hereafter safety concern). Given the potential over-
lap in interpretation of the middle categories (2 ⫽a
bit,3⫽moderate), responses to this item were
subsequently grouped into three categories: low,a
bit/moderate, and high, with a bit/moderate used as
the reference category when this variable was
dummy coded for analysis purposes. Twenty-three
percent of the sample rated their primary job as
having low safety concern, 47% rated their primary
job as having a bit/moderate safety concern, and 30%
rated their primary job as having high safety concern.
In comparison, 37% of the sample rated their sec-
282 INNESS, TURNER, BARLING, AND STRIDE
ondary job as having low safety concern, 41% rated
their secondary job as having a bit/moderate safety
concern, and 22% rated their secondary job as having
high safety concern.
Negative affectivity was assessed using Fortunato
and Goldblatt’s (2002) 20-item Revised Strain-Free
Negative Affectivity Scale (e.g., “I get resentful
when someone expects me to do something I really
don’t want to do”). Items are rated on a 5-point
Likert-type scale (1 ⫽strongly disagree,5⫽
strongly agree).
Conscientiousness was measured using seven pos-
itively worded items from the NEO-Five-Factor In-
ventory (Costa & McCrae, 1992), in which partici-
pants are asked to rate on a 5-point response scale
(1 ⫽strongly disagree,5⫽strongly agree) the
extent to which they engage in a series of conscien-
tious behaviors (e.g., “I have a clear set of goals and
work toward them in an orderly fashion”).
For each of the scales, internal consistency reli-
ability was high; Cronbach’s alpha coefficients are
given in Table 1.
We also measured gender (0 ⫽female, 1 ⫽male),
age, hours worked per week in each job, and time
with supervisor in each job (the logarithms of the
latter pair of variables were used because of the
extreme positive skew of their distributions).
Results
Descriptive statistics and intercorrelations for
study variables appear in Table 1.
Measurement Models
Given that scales were repeated over multiple mea-
surements (i.e., in the primary and secondary jobs),
and that the transformational leadership, safety com-
pliance, and safety participation scales were derived
from subsets of items from larger scales, it was
important to establish that distinct constructs were
being measured for these three constructs and two
jobs, and that each scale was structurally invariant
across jobs. We did this in two ways.
First, to assess the quality of the proposed six-
factor measurement model (transformational leader-
ship, safety compliance, and safety participation
items loading on distinct factors for each job), we
computed a confirmatory factor analysis to test its fit
to the data. The resulting model,
2
(45) ⫽185.00,
p⬍.01, suggested an excellent fit to the data on the
range of fit indices recommended by Hu and Bentler
(1998), with the comparative fit index (CFI) ⫽.98,
Tucker–Lewis Index (TLI) ⫽.97, root mean square
error approximation (RMSEA) ⫽.04, and standard-
ized root mean square residual (SRMR) ⫽.05. Al-
ternative four-factor (safety compliance and safety
participation merged to form a single safety factor for
each job) and three-factor models (each of the three
measure forming a single factor across the two jobs)
yielded dramatically weaker fits to the data (CFI ⫽
.75, TLI ⫽.69, RMSEA ⫽.16, SRMR ⫽.17; and
CFI ⫽.68, TLI ⫽.61, RMSEA ⫽.18, SRMR ⫽.14,
respectively). Correlations between factors in the six-
factor model ranged in absolute value from .01 to .56.
A restricted model examining invariance of item-
factor loadings by fixing them to be equal across the
two jobs for the respective factors did not obtain
significantly worse fit than the initial model,
2
(152) ⫽194.00, p⬍.01; ⌬
2
(7) ⫽9.00, p⬎.05.
Second, we examined the convergent and discrimi-
nant validity of these factors via the average variance
extracted (AVE) method (Fornell & Larcker, 1981).
The AVE scores of transformational leadership for
primary and secondary jobs were .68 and .78, respec-
tively; for safety compliance, .75 and .70; and for
safety participation, .63 and .79. This satisfied For-
nell and Larcker’s criterion (1981; AVE ⬎.50) for
convergent validity. The variance shared (i.e., the
squared correlation) between any pair of factors
ranged from .01 to.31, and was less than the lowest
AVE score across the factors, satisfying Fornell and
Larcker’s discriminant validity criterion.
Structural Models
We used observed variable path analysis (Kello-
way, 1998) to test our hypotheses via the estimation
of the full model (depicted in Figure 1) and a series
of nested restricted competing models (Table 2).
Given the constraints of our sample size and the
additional variables used as controls in these models,
we were unable to directly extend the measurement
model described above to a structural equation mod-
el; instead, we computed observed mean scale scores
over the respective sets of items for transformational
leadership, safety compliance, and safety participa-
tion (as well as for conscientiousness and negative
affectivity) and used these rather than their latent
equivalents. In all model comparison and path coef-
ficient testing, the 95% confidence level of statistical
significance was applied (i.e., p⬍.05), with two-
tailed tests used in all analyses; the models were
estimated using full information maximum likeli-
hood, enabling the use of the full sample (N⫽159)
283TRANSFORMATIONAL LEADERSHIP AND SAFETY
Table 1
Study Variables: Descriptive Statistics, Intercorrelations, and Scale Reliability
Variable MSD␣1234567891011121314151617
1. Safety participation (Job 1) 3.55 .99 .83
2. Safety participation (Job 2) 3.23 1.19 .92 .37
3. Safety compliance (Job 1) 3.79 1.08 .90 .14 .09
4. Safety compliance (Job 2) 3.92 1.11 .94 .18 ⫺.02 .61
5. Transformational
leadership (Job 1) 3.01 1.09 .92 .31 .06 .10 .04
6. Transformational
leadership (Job 2) 2.87 1.23 .95 .15 .51 .14 ⫺.07 .29
7. Gender .46 .50 — ⫺.07 .03 .00 .01 ⫺.06 .06
8. Age 37.36 10.25 — .10 .12 .15 .13 ⫺.08 ⫺.07 ⫺.05
9. Hours worked per week
(Job 1) 38.20 12.73 — .13 .03 ⫺.14 ⫺.09 .04 ⫺.01 .29 ⫺.04
10. Hours worked per week
(Job 2) 17.79 12.93 — .15 .29 .14 .00 .05 .20 .02 .04 .13
11. Time with supervisor (log;
Job 1) 1.27 .65 — .01 .06 .08 .08 .02 .00 ⫺.01 .11 ⫺.03 .19
12. Time with supervisor (log;
Job 2) 1.13 .72 — .09 .33 .04 ⫺.08 .04 .20 .06 .13 .04 .28 .37
13. Negative affectivity 2.98 .78 .93 ⫺.14 ⫺.10 ⫺.27 ⫺.27 ⫺.19 ⫺.09 ⫺.03 ⫺.22 ⫺.01 ⫺.08 .00 .09
14. Conscientiousness 3.87 .63 .86 .37 .16 .32 .25 .27 .19 .12 .16 .04 .01 ⫺.02 .00 ⫺.22
15. Level of safety issues in
Job 1: Low
a
.23 .42 — ⫺.24 ⫺.27 .09 .11 .02 ⫺.10 ⫺.05 ⫺.21 ⫺.06 ⫺.10 ⫺.14 ⫺.24 .00 ⫺.08
16. Level of safety issues in
Job 1: High
a
.30 .46 — .24 .23 .11 .16 .02 .08 ⫺.05 .01 .06 ⫺.11 .06 ⫺.03 ⫺.04 .16 ⫺.35
17. Level of safety issues in
Job 2: Low
a
.37 .48 — ⫺.11 ⫺.31 .04 .19 ⫺.04 ⫺.20 ⫺.07 ⫺.02 ⫺.18 ⫺.24 .11 ⫺.13 ⫺.01 ⫺.12 .41 ⫺.16
18. Level of safety issues in
Job 2: High
a
.22 .42 — .14 .30 .22 .14 .00 .12 .05 .11 .18 .12 .02 .12 .03 .21 ⫺.18 .35 ⫺.40
Note. log ⫽log transformation; Job 1 ⫽primary job; Job 2 ⫽secondary job. 126 ⬍n⬍159.
a
Versus reference category a bit/moderate amount.
284 INNESS, TURNER, BARLING, AND STRIDE
throughout rather than losing cases to attritional ran-
dom missing data.
Table 3, which shows the path coefficients from
the full model, indicates that transformational lead-
ership is a statistically significant positive predictor
of safety participation. Specifically, transformational
leadership in the primary job predicted safety partic-
ipation in the primary job (⫽.19, p⬍.05) and
transformational leadership in the secondary job pre-
dicted safety participation in the secondary job (⫽
.43, p⬍.05). In the primary job, safety participation
was also predicted by conscientiousness (⫽.44,
p⬍.05), the number of hours worked per week (⫽
.02, p⬍.05), and the level of safety concern within
that job (the path coefficient for the dummy variable
for low concern; ⫽⫺.37, p⬍.05) indicated
significantly less safety participation among workers
within this group than for the reference category (i.e.,
JOB 1
JOB 2
Transformational
leadership
Safety
compliance
Safety
participation
Transformational
leadership
Safety
compliance
Safety
participation
Gender
Age
Conscientiousness
Negative Affect
Time with supervisor
Hours worked/week
Figure 1. Hypothesized model. Solid single-headed lines indicate proposed context-specific
or within-person effects. Dotted single-headed lines indicate proposed context-spillover
effects. Dotted double-headed lines indicate proposed covariance.
Table 2
Model Comparison: Full Model Versus Nested Models With Restricted Subsets of Paths From
Transformational Leadership Behavior to Employee Safety Performance
Path from transformational leadership to employee safety
performance
2
df
⌬
2
against
Model 1
⌬df against
Model 1
1. Full model (all paths)—see Figure 1 33.26 18 — —
2. Paths from transformational leadership to safety performance only 38.84 22 5.58 4
3. Paths from transformational leadership to safety compliance only 83.06 22 49.80
a
4
4. Paths from transformational leadership (Job 1) to safety
compliance and safety performance (Job 2) only 35.59 20 2.33 2
5. Transformational leadership (Job 2) to safety compliance and
safety performance (Job 1) only 35.19 20 1.93 2
6. No cross-job job paths 37.39 22 4.12 4
Note. Job 1 ⫽primary job; Job 2 ⫽secondary job. N⫽159.
a
The restricted model is significantly worse than Model 1, p⬍.05.
285TRANSFORMATIONAL LEADERSHIP AND SAFETY
a bit/moderate concern). In the secondary job, safety
participation was also predicted by level of safety
concern; the coefficients for the dummy variables for
low concern (⫽⫺.35, p⬍.05) and high concern
(⫽.43, p⬍.05) indicated significant differences in
safety participation compared with the reference cat-
egory; again, the sign of the respective coefficients
demonstrates the positive association between level
of safety concern and safety participation.
In contrast, transformational leadership in the pri-
mary job did not predict safety compliance in the
primary job (⫽⫺.03, ns), nor did transformational
leadership in the secondary job predict safety com-
pliance in the secondary job (⫽⫺.06, ns). In the
primary and secondary jobs, safety compliance was
predicted by conscientiousness (⫽.47, p⬍.05,
and ⫽.40, p⬍.05, respectively), negative affec-
tivity (⫽⫺.26, p⬍.05, and ⫽⫺.30, p⬍.05,
respectively), and level of safety concern in the job.
The effect of the latter differed from that which it had
previously exhibited on safety participation, with the
positive coefficients of both the low concern and high
concern dummy variables versus the reference cate-
gory indicating that it was curvilinear in form; low
and high safety concern prompted higher levels of
safety compliance than when they were moderate.
With regards to context-spillover effects, transfor-
mational leadership in the primary job did not predict
safety compliance (⫽⫺.02, ns) or safety partici-
pation (⫽⫺.11, ns) in the secondary job. Simi-
larly, transformational leadership in the secondary
job did not predict safety compliance (⫽.09, ns)or
safety participation (⫽.02, ns) in the primary job.
The model comparison stage of the analysis began
by removing the paths between transformational
leadership and safety compliance in both jobs (i.e.,
restricting the effect of transformational leadership to
safety participation only; Model 2). We then tested
the alternative restriction of removing the paths be-
tween transformational leadership and safety partic-
ipation in both jobs (i.e., restricting the effect of
transformational leadership to safety compliance
only; Model 3). Next, we removed only the paths
from transformational leadership in the secondary job
to safety compliance and safety participation in the
primary job (i.e., restricting the context-spillover ef-
fect of transformational leadership in the primary job
to employee safety performance in the secondary job;
Model 4). We then removed just the paths between
transformational leadership in the primary job and
both safety compliance and safety participation in the
secondary job (i.e., restricting the context-spillover
Table 3
Path Coefficients for Final Model
Outcome
Safety participation Safety compliance
Job 1 Job 2 Job 1 Job 2
Predictor variable B SE B SE B SE B SE
Controls
Gender ⫺.27 .15 ⫺.10 .16 ⫺.01 .17 .00 .17
Age .00 .01 .01 .01 .01 .01 .01 .01
Negative affectivity ⫺.05 .10 ⫺.12 .11 ⫺.26
ⴱ
.11 ⫺.30
ⴱ
.11
Conscientiousness .44
ⴱ
.12 .06 .13 .47
ⴱ
.14 .40
ⴱ
.14
Hours worked per week (Job 1) .01
ⴱ
.00 — — ⫺.01 .01 — —
Hours worked per week (Job 2) — — .01 .01 — — .00 .01
Time with supervisor (log; Job 1) ⫺.02 .10 — — .01 .10 — —
Time with supervisor (log; Job 2) — — .26
ⴱ
.11 — — ⫺.10 .10
Level of safety issues in Job 1: Low
a
⫺.37
ⴱ
0.17 — — .36
ⴱ
.18 — —
Level of safety issues in Job 1: High
a
0.16 0.15 — — .08 .16 — —
Level of safety issues in Job 2: Low
a
——⫺.35
ⴱ
.16 — — .51
ⴱ
.16
Level of safety issues in Job 2: High
a
— — .43
ⴱ
.19 — — .28 .19
Transformational leadership
Job 1 .19
ⴱ
.07 ⫺.11 .07 ⫺.03 .08 ⫺.02 .08
Job 2 .02 .06 .43
ⴱ
.07 .09 .07 ⫺.06 .07
Note. log ⫽log transformation; Job 1 ⫽primary job; Job 2 ⫽secondary job. N⫽159.
a
Versus reference category a bit/moderate amount.
ⴱ
p⬍.05.
286 INNESS, TURNER, BARLING, AND STRIDE
effect of transformational leadership in the secondary
job to employee safety performance in the primary
job; Model 5). Finally, we removed all cross-job
effects (i.e., so that transformational leadership in
each job predicted only safety compliance and safety
participation in the same job; Model 6). The one
model that offered a significantly weaker fit than the
full model was Model 3, in which the paths from
transformational leadership to safety participation
were removed, confirming the evidence from the full
model path coefficients that the within-job transfor-
mational leadership–safety participation relationships
are an essential component to the best-fitting model.
The sequence of models did not indicate that re-
taining either pair of cross-job effects (Models 4 and
5) resulted in a significantly better fit to the data than
the entirely context-specific model (Model 6); thus,
the latter was the most parsimonious model, in which
transformational leadership in a given job is related
to employee safety performance in that job only.
Discussion
The goal of the present study was twofold. First,
we investigated the predictive relationships between
transformational leadership and employee safety per-
formance (i.e., safety compliance and safety partici-
pation). Second, we compared a context-specific and
a context-spillover model of the relationships be-
tween transformational leadership and employee
safety performance. To this end, we used a sample of
moonlighters, each of whom works two jobs, creating
a robust design that sidesteps the effects of between-
persons differences when testing between jobs.
Hypothesis 1b was fully supported: Within-job
transformational leadership was associated with
within-job safety participation in the primary and
secondary jobs. This is consistent with the notion that
transformational leadership serves to motivate supe-
rior employee contextual performance (Conger &
Kanungo, 1988). This finding also suggests that
transformational leadership does not need to have a
safety-specific focus to motivate safety participation
in employees. Generalized transformational leader-
ship is an ongoing leadership style and can be used
by supervisors to achieve a number of interpersonal
and organizational goals, including encouraging em-
ployees to take extra measures to make the work
environment safe.
Some control factors were also found to be related
to safety participation, highlighting additional under-
pinnings of safety participation. Most notably, in
both jobs, when safety was of greater concern, em-
ployees engaged in higher levels of safety participa-
tion, probably as a result of the salience of risk on the
job. In addition, in the context of the primary job but
not the secondary job, conscientiousness and the
number of hours employees spent on the job showed
a positive relationship with safety participation. This
suggests that some people may be more likely than
others to engage in safety-related efforts that go
beyond the call of duty, and that employees are more
likely to engage in safety participation when the
opportunity to do so is greater. As neither of these
factors influenced safety participation in the second-
ary job where, by definition, employees work fewer
hours per week, this may suggest that the decision to
engage in safety participation may require particular
motivational influences, such as transformational
leadership.
In contrast, Hypothesis 1a was not supported:
Transformational leadership was not related to safety
compliance in either the primary or secondary job.
We expected that, given that transformational leaders
exhibit supportiveness and general concern for em-
ployees’ well-being, experiencing transformational
leadership would render employees more attentive to
their own well-being, with one way of doing so being
complying with safety procedures. However, the ex-
perience of transformational leadership did not affect
whether employees followed work safety rules. One
possible explanation for this finding may have to do
with the nature of transformational leadership. In
particular, higher levels of transformational leader-
ship may indirectly give employees greater latitude to
use their discretion in deciding whether to comply
with existing organizational policies such as safety
procedures, resulting in variability in individual
safety compliance. This may suggest that the exertion
of formal control through rewards and punishments
to gain employee compliance, activities more closely
associated with transactional leadership (Kelloway et
al., 2006), may be more appropriate. Future research
should examine whether the full range of leadership,
including transactional leadership, exerts a different
effect on safety outcomes.
Our findings also suggest that certain control fac-
tors consistently affected safety compliance in the
primary and secondary jobs. Conscientiousness (pos-
itively) and negative affectivity (negatively) affected
safety compliance in both jobs, suggesting that some
people are more likely to comply with safety proce-
dures than are others. This might also suggest that
when safety compliance is important in a particular
work setting, employee selection may be a valuable
consideration (Tetrick, Perrewe´, & Griffin, in press).
287TRANSFORMATIONAL LEADERSHIP AND SAFETY
In addition, across both jobs, employees were more
likely to comply with safety procedures when safety
was of low concern or high concern than when it was
of moderate concern. Although seemingly counterin-
tuitive, the higher levels of compliance for low as
opposed to moderate concern may be explained in
several ways. When safety concern is low, it may be
easier for employees to comply with safety rules, or
safety-compliant employees may perceive that they
have greater perceived control over the risks of the
work environment (Rebitzer, 1995).
Our second goal was to compare context-specific
and context-spillover models of the relationships be-
tween transformational leadership and employee
safety performance, and in so doing, we implicitly
addressed how transformational leadership exerts ef-
fects on employee safety performance, whether it is
by motivation, socialization, or both. The context-
specific model was supported, with potentially inter-
esting implications. From a practical perspective, be-
cause the benefits of transformational leadership did
not spillover beyond the job in which that leadership
was experienced, this suggests that it is important to
provide transformational leadership in all jobs where
safety performance, and in particular, safety partici-
pation is desired, even when an employee has expe-
rienced a previous or concurrent work environment
in which transformational leadership has been expe-
rienced. Conceptually, this suggests that there are
motivational underpinnings of transformational lead-
ership that arise in relation to a specific leader, per-
haps as a form of reciprocity for a particular leader’s
extra efforts on behalf of the employee, as a result of
liking or a sense of oneness with the leader (Brown &
Keeping, 2005) or perceiving that the leader has a
compelling vision (Conger & Kanungo, 1998). This
motivational explanation seems particularly likely
given the finding that safety participation (i.e., con-
textual performance) but not safety compliance (i.e.,
task performance) is influenced by transformational
leadership. The notion that employee motivation is
critical to safety participation has been suggested in a
recent meta-analysis (Christian et al., 2009), indicat-
ing that motivation was a proximal mediator of the
impact of workplace experiences on employee safety.
As with all research, the present study has some
limitations. First, all data were obtained from the
participants, raising possible questions about single-
source bias. Nevertheless, the fit of measurement
models that collapsed focal constructs within-person
that would be suggestive of single-source bias was
poor, with the measurement model that recognized
both within-person and between-jobs measurement
yielding a comparatively superior fit. In addition to
the research design, conscientiousness and negative
affectivity were controlled in the measurement and
structural models, which further assuages concern
about systematic response biases inflating focal rela-
tionships (Podsakoff, MacKenzie, Lee, & Podsakoff,
2003). Second, all analyses in the present study were
based on cross-sectional data, and, as such, causal
inferences about the role of transformational leader-
ship causing employee safety performance are pre-
mature. A third issue that may limit the generaliz-
ability of the model is that this study was conducted
with moonlighters. However, we argue that there is
little conceptual or empirical reason to expect that
differences between moonlighters and nonmoonlight-
ers threaten conclusions from the results obtained. In
a study that directly compared moonlighters with
other workers, moonlighters shared substantial simi-
larities with full-time employees (Sinclair, Martin, &
Michel, 1999). Although moonlighters constituted
the participants in the present study, we believe that
the context-specific effects are applicable to individ-
uals in a wide range of work contexts.
Overall, the present study holds potentially impor-
tant implications for the role of transformational
leadership in enhancing employee safety perfor-
mance. Transformational leadership is most strongly
related to safety participation, suggesting a motiva-
tional influence. Individual differences are most
strongly related to safety compliance, suggesting that
disposition and experience with the job are important.
The finding that the impact of transformational lead-
ership is job-specific highlights the importance of the
contextual nature of transformational leadership on
employee safety performance. These results are
strengthened by the design of the present study,
which controlled for within-person differences.
References
Barling, J., Kelloway, E. K., & Iverson, R. (2003). Acci-
dental outcomes: Attitudinal consequences of workplace
injuries. Journal of Occupational Health Psychology, 8,
74 – 85.
Barling, J., Loughlin, C. A., & Kelloway, E. K. (2002).
Development and test of a model linking safety-specific
transformational leadership and occupational safety.
Journal of Applied Psychology, 87, 488 – 496.
Barling, J., Weber, T., & Kelloway, E. K. (1996). Effects of
transformational leadership training and attitudinal and
fiscal outcomes: A field experiment. Journal of Applied
Psychology, 81, 827– 832.
Bass, B. M., & Avolio, B. J. (1995). The Multifactor Lead-
ership Questionnaire 5-X. Palo Alto, CA: Consulting
Psychologists Press.
288 INNESS, TURNER, BARLING, AND STRIDE
Bass, B. M., & Riggio, R. E. (2006). Transformational
leadership (2nd ed.). Mahwah, NJ: Erlbaum.
Borman, W. C., & Motowidlo, S. J. (1997). Task perfor-
mance and contextual performance: The meaning for
personnel selection research. Human Performance, 10,
99 –109.
Brown, D. J., & Keeping, L. M. (2005). Elaborating the
construct of transformational leadership: The role of af-
fect. The Leadership Quarterly, 16, 245–272.
Christian, M. S., Bradley, J. C., Wallace, J. C., & Burke,
M. J. (2009). Workplace safety: A meta-analysis of the
roles of person and situation factors. Journal of Applied
Psychology, 94, 1103–1127.
Clarke, S., & Ward, K. (2006). The role of leader influence
tactics and safety climate in engaging employees’ safety
participation. Risk Analysis, 26, 1175–1185.
Cohen, A. (1977). Factors in successful occupational safety
programs. Journal of Safety Research, 9, 168 –178.
Conchie, S. M., & Donald, I. J. (2009). The moderating role
of safety-specific trust on the relation between safety-
specific leadership and safety citizenship behaviors.
Journal of Occupational Health Psychology, 14, 137–
147.
Conger, J. A. (1989). The charismatic leader: Behind the
mystique of exceptional leadership. San Francisco: Jos-
sey-Bass.
Conger, J. A., & Kanungo, R. N. (1998). Charismatic
leadership in organizations. Thousand Oaks, CA: Sage.
Costa, P. T., & McCrae, R. R. (1992). The five-factor model
of personality and its relevance to personality disorders.
Journal of Personality Disorders, 6, 343–359.
Fornell, C., & Larcker, D. (1981). Structural equation mod-
els with unobservable variables and measurement error.
Journal of Marketing Research, 18, 39 –50.
Fortunato, V. J., & Goldblatt, A. M. (2002). Construct
validation of Revised Strain-Free Negative Affectivity
Scale. Educational and Psychological Measurement, 62,
45– 63.
Griffin, M. A., & Neal, A. (2000). Perceptions of safety at
work: A framework for linking safety climate to safety
performance, knowledge, and motivation. Journal of Oc-
cupational Health Psychology, 5, 347–358.
Ha¨ma¨la¨inen, P., Takala, J., & Saarela, K. L. (2006). Global
estimates of occupational accidents. Safety Science, 44,
137–156.
Hogan, J., & Ones, D. S. (1997). Conscientiousness and
integrity at work. In R. Hogan & J. A. Johnson (Eds.),
Handbook of personality psychology (pp. 849– 870). San
Diego, CA: Academic Press.
Howell, J. M., & Avolio, B. J. (1993). Transformational
leadership, transactional leadership, locus of control, and
support for innovation: Key predictors of consolidated
business-unit performance. Journal of Applied Psychol-
ogy, 78, 891–902.
Hu, L., & Bentler, P. M. (1998). Fit indices in covariance
structure modeling: Sensitivity to underparameterized
model misspecification. Psychological Methods, 3, 424 –
453.
Iverson, R. D., & Erwin, P. J. (1997). Predicting occupa-
tional injury: The role of affectivity. Journal of Occupa-
tional and Organizational Psychology, 70, 113–128.
Jablin, F. M. (2001). Organizational entry, assimilation and
disengagement/exit. In F. M. Jablin & L. L. Putnam
(Eds.), The new handbook of organizational communica-
tion: Advances in theory, research, and methods (pp.
732– 818) Thousand Oaks, CA: Sage.
Kaplan, S., Bradley, J. C., Luchman, J. N., & Haynes, D.
(2009). On the role of positive and negative affectivity in
job performance: A meta-analytic investigation. Journal
of Applied Psychology, 94, 162–176.
Kelloway, E. K. (1998). Using LISREL for structural equa-
tion modeling: A researcher’s guide. Thousand Oaks,
CA: Sage.
Kelloway, E. K., Mullen, J., & Francis, L. (2006). Divergent
effects of transformational and passive leadership on
employee safety. Journal of Occupational Health Psy-
chology, 11, 76 – 86.
Koh, W. L., Steers, R. M., & Terborg, J. R. (1995). The
effects of transformation leadership on teacher attitudes
and student performance in Singapore. Journal of Orga-
nizational Behavior, 16, 319 –333.
Lieberman, S. (1956). The effects of changes in roles on the
attitudes of role occupants. Human Relations, 9, 385–
402.
Maclagan, P. (2007). Hierarchical control or individuals’
moral autonomy? Addressing a fundamental tension in
the management of business ethics. Business Ethics: A
European Review, 16, 48 – 61.
Mischel, W., & Shoda, Y. (1999). Integrating dispositions
and processing dynamics within a unified theory of per-
sonality: The cognitive–affective personality system. In
L. A. Pervin & O. John (Eds.), Handbook of personality:
Theory and research (2nd ed., pp. 197–218). New York:
Guilford Press.
Mullen, J. E., & Kelloway, E. K. (2009). Safety leadership:
A longitudinal study of the effects of transformational
leadership on safety outcomes. Journal of Occupational
and Organizational Psychology, 82, 253–272.
Neal, A., & Griffin, M. A. (2006). A study of the lagged
relationships among safety climate, safety motivation,
safety behavior, and accidents at the individual and group
levels. Journal of Applied Psychology, 91, 946 –953.
Neal, A., Griffin, M. A., & Hart, P. M. (2000). The impact
of organizational climate on safety climate and individual
behavior. Safety Science, 34, 99 –109.
Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., & Podsa-
koff, N. P. (2003). Common method biases in behavioral
research: A critical review of the literature and recom-
mended remedies. Journal of Applied Psychology, 88,
879 –903.
Rebitzer, J. B. (1995). Job safety and contract workers in the
petrochemical industry. Industrial Relations, 34, 40 –57.
Ritter, B. A., & Lord, R. G. (2007). The impact of previous
leaders on the evaluation of new leaders: An alternative
to prototype matching. Journal of Applied Psychology,
92, 1683–1695.
Roberts, K., & Markel, K. S. (2001). Claiming in the name
of fairness: Organizational justice and the decision to file
for workplace injury compensation. Journal of Occupa-
tional Health Psychology, 6, 332–347.
Ross, L., & Nisbett, R. E. (1991). The person and the
situation: Perspectives of social psychology. McGraw-
Hill Series in Social Psychology. New York: McGraw-
Hill.
Salancik, G. R., & Pfeffer, J. (1978). A social information
processing approach to job attitudes and task design.
Administrative Science Quarterly, 23, 224 –253.
Sinclair, R. R., Martin, J. E., & Michel, R. P. (1999).
289TRANSFORMATIONAL LEADERSHIP AND SAFETY
Full-time and part-time subgroup differences in job atti-
tudes and demographic characteristics. Journal of Voca-
tional Behavior, 55, 337–357.
Study Response Project. (n.d.). Retrieved from http://
studyresponse.syr.edu/studyresponse/index.htm
Tetrick, L., Perrewe´, P., & Griffin, M. A. (in press). Work
stress, safety, and personnel selection. In J. L. Farr (Ed.),
Handbook of employee selection. New York: Taylor &
Francis.
U.S. Bureau of Labor Statistics. (2006). Number of jobs
held, labor market activity, and earnings growth among
the youngest baby boomers: Results from a longitudinal
survey. Retrieved from http://www.bls.gov/news.release/
pdf/nlsoy.pdf
U.S. Census Bureau. (2006). Accessed from http://www
.census.gov/Press-Release/www/2006/ cb06ff-12.pdf
U.S. Department of Labor. (2006). Multiple jobholders by
selected demographic and economic characteristics. Re-
trieved from http://www.bls.gov/cps/cpsaat36.pdf
Wallace, C., & Chen, G. (2006). A multilevel integration of
personality, climate, self-regulation, and performance.
Personnel Psychology, 59, 529 –557.
Watson, D., & Clark, L. A. (1984). Negative affectivity:
The disposition to experience aversive emotional states.
Psychological Bulletin, 96, 465– 490.
Witt, L. A., Burke, L. A., Barrick, M. A., & Mount, M. K.
(2002). The interactive effects of conscientiousness and
agreeableness on job performance. Journal of Applied
Psychology, 87, 164 –169.
Zohar, D. (1980). Safety climate in industrial organizations:
Theoretical and applied implications. Journal of Applied
Psychology, 65, 96 –102.
Zohar, D. (2002). The effects of leadership dimensions,
safety climate, and assigned priorities on minor injuries
in work groups. Journal of Organizational Behavior, 23,
75–92.
Received May 17, 2009
Revision received January 29, 2010
Accepted February 2, 2010 y
290 INNESS, TURNER, BARLING, AND STRIDE
A preview of this full-text is provided by American Psychological Association.
Content available from Journal of Occupational Health Psychology
This content is subject to copyright. Terms and conditions apply.
