Spilling over

Literature review

Focus of current study

The job demands–resources model provides a theoretical underpinning for understanding how different workplace variables may be associated with strain-based conflict. This model divides the workplace into demands and resources (Bakker and Demerouti, 2007; Demerouti and Bakker, 2011; Demerouti et al., 2001; Hall et al., 2010; Hu et al., 2011). Workplace demands place a strain on workers, and over time can wear down the person, resulting in negative outcomes (Hall et al., 2010; Schaufeli and Taris, 2014). As noted by Mauno et al. (2006: 212), workplace ‘demands are physical, psychological, social, or organizational features of the job requiring physical and/or psychological costs (i.e., strain).’ Policing is a unique occupation. It requires shift work, dealing at times with unwilling individuals, and being at risk of harm (Ortega et al., 2007). Officers face many situations that can place psychological and emotional stress on them, which may influence the quality of their home life (Youngcourt and Huffman, 2005). Workplace demands are likely to add to the strain of police work, and it is doubtful that these strains stay at work when officers go home. Strains from work likely follow an officer home and cause strain-based conflict. Possible workplace demands on police officers are time-based conflict, behavior-based conflict, role conflict, role ambiguity, role overload, role underload, and perceived dangerousness of the job.

Workplace resources help workers be successful at their jobs, make work more pleasant, and often are linked to positive outcomes (Hu et al., 2011). As pointed out by Mauno et al. (2006: 212), workplace ‘resources are physical, psychological, social features of the job, which in turn, are functional in (1) achieving work goals, (2) reducing job demands and physical and/or psychological costs associated with them, and stimulating personal growth and development.’ Not only do workplace resources provide positive outcomes for people, they can also buffer people from the negative effects of workplace demands (Bakker and Demerouti, 2007; Bakker et al., 2005; Mauno et al., 2006). Furthermore, a lack of workplace resources can result in frustration and other negative psychological feelings (Mauno et al., 2006). In addition, Hall et al. (2010) contended that people who lack resources at work are at greater risk of conflict in their home domain because work and home are interlinked. Work resources help people be more effective at their jobs. Missing or low workplace resources can hamper the ability of officers to do their work, which, in turn, can increase psychological strain. Frustration arising from a lack of resources is likely to follow officers home, increasing the chances of strain-based conflict in their personal lives. Possible workplace resources of police officers are quality supervision, training, job autonomy, and job variety.

As previously mentioned, time-based conflict occurs when the amount of time work requires and/or the scheduling of work shifts interfere with home life (Mitchel et al., 2011). Both overtime and amount of time worked have been found to result in higher work–family conflict (Britt and Dawson, 2005; Fu and Shaffer, 2001). Police and the military often work for very structured organizations that at times require long hours and the need for around-the-clock coverage. In fact, police organizations are sometimes described as being paramilitary organizations (Ortega et al., 2007). Among military personnel, the number of hours worked has been shown to be positively associated with increased work–family conflict (Britt and Dawson, 2005). Unlike many other occupations, police agencies operate 24 hours a day, every day of the year, including holidays. In addition, posts cannot always be vacated, so when an officer is off sick or on leave, another officer may be required to stay past his or her shift or to come in on a day off to fill the open position. Unexpected emergencies or problems may also arise, requiring officers to stay past their normal shift or to come in during their time off from work (Burke, 1988; Kinman et al., 2012; Mikkelsen and Burke, 2004). In addition, shift work and being on call, which are required for officers, can result in conflict at home (Youngcourt and Huffman, 2005). Time-based conflict was, therefore, hypothesized to have a positive association with strain-based conflict among Indian officers.

Behavior-based conflict occurs when work roles are incompatible with home roles and result in conflict (Brough and O’Driscoll, 2005; Netermeyer et al., 1996). Behavior-based conflict is not often examined (Kinman et al., 2012; Mauno et al., 2006; Youngcourt and Huffman, 2005). As noted by Dierdorff and Ellington (2008) and Kinman et al. (2012), this form of domain conflict may apply more to some occupations than others, particularly occupations that involve interpersonal dependence on others and dealing with interpersonal conflict, such as are found in law enforcement. Policing requires work roles that are not typically found in other occupations. For example, officers who are firm, dominant, detached, suspicious, questioning, and, at times, aggressive at work may be utilizing roles that serve them well when dealing with uncooperative suspects while on duty, but these same roles will probably work less well when interacting with family and friends, which can cause strain at home (Mikkelsen and Burke, 2004; Ortega et al., 2007). In addition, officers must rely on their fellow officers, resulting in a high level of interdependence, and in behavior roles not found in other occupations (Dierdorff and Ellington, 2008; Kinman et al., 2012). In addition, officers must adapt roles to deal with conflict with others while on the job. The work roles needed to be a successful officer are not those typically needed to be successful in home life. As noted by Youngcourt and Huffman (2005: 140), ‘an individual’s role as a police officer may conflict with that individual’s role as a family member.’ Behavior-based conflict was, therefore, hypothesized to have a positive association with the strain dimension of work–family conflict.

Role conflict can occur when role behaviors, duties, and directions are inconsistent with one another (Lambert, Hogan, Paoline and Clarke, 2005; Rizzo et al., 1970). Role conflict is a possibility among officers who work in bureaucratic organizations with conflicting rules and directions. Also, the expectations and direction of supervisors can differ in police agencies, which may result in role conflict. In addition, the nature of law enforcement entails both enforcement and service; this dichotomy in and of itself can result in role conflict for officers. Correctional staff also work in a bureaucratic, paramilitary structure in which conflicting orders and conflicting expectations can occur. Among staff at a private U.S. prison, role conflict was observed to result in greater work-on-family conflict (Lambert, Minor et al., 2015). Role conflict can result in psychological frustration, which for public safety workers is unlikely to end when the shift ends (Howard et al., 2004). This psychological strain from work may spill over, negatively affecting home life, including conflict and strain with family and friends (Demerouti and Bakker, 2011). Role conflict was hypothesized to have a positive association with strain-based conflict for Indian police officers.

Role ambiguity occurs when there is a lack of clarity about carrying out job duties and responsibilities (Lambert, Hogan, Paoline and Clarke, 2005; Rizzo et al., 1970). Role ambiguity can hamper the ability of a person to do his or her job, which may result in frustration and anger (Michel et al., 2011; Schaufeli and Taris, 2014). These negative feelings can follow the person home, resulting in greater strain and conflict. Role ambiguity is a possibility in policing, where officers have to deal with a wide array of fluid and complex situations. Corrections, like law enforcement, is a complex, bureaucratic paramilitary organization in which role ambiguity can arise. Among correctional staff at a U.S. public prison, role ambiguity resulted in greater work on family conflict (Lambert and Hogan, 2006). Role ambiguity was hypothesized to have a positive association with strain-based conflict among Indian police officers.

Role overload occurs when a worker is required to handle too many tasks or is not provided with the necessary resources or equipment to accomplish the assigned duties (Lambert, Hogan, Paoline and Clarke, 2005). Police officers have to deal with a wide array of job tasks. In addition, officers are often called upon to deal with complex situations that require a multitude of responses and actions in quick succession for the matter to be successfully resolved. Also, officers may not always have the necessary equipment or resources at hand to deal effectively with all situations they encounter on the job. This can be a trying experience, which may result in frustration and psychological stress. Feeling overwhelmed at work can be a trying experience that is unlikely to dissipate immediately at the end of a work shift, but rather to follow a person home, leading to strain (Michel et al., 2011). Among correctional officers working at a high-security U.S. prison, role overload was related to increased domain conflict (Lambert, Minor et al., 2015). As such, role overload was hypothesized to have a positive association with strain-based work–family conflict for Indian police officers.

Whereas role overload requires a person to deal with too many work tasks and duties in the given time frame, role underload is not being given enough work (Lambert, Hogan, Paoline and Clarke, 2005; Melamed et al., 1995). Sales (1970: 593) saw role underload as a ‘condition in which the person is faced with obligations which, taken as a set, require him to do considerably less than he is able to do in the time available.’ Role underload can be a stressor because it can result in boredom or a lack of mental stimulation (Sales, 1970; Triplett et al., 1996). In policing, some jobs require too much work, whereas others require too little; both can result in strain for officers (Lambert, Hogan, Paoline and Clarke, 2005). This negative strain can result in negative outcomes for officers (Melamed et al., 1995). One such negative outcome is that the strain from being underwhelmed at work may follow an officer home, resulting in greater strain-based conflict. As such, role underload was postulated to be positively related to strain-based conflict for Indian police officers.

The dangerousness of the job is a perception on the part of an employee that the job is threatening to the employee’s safety (Cullen et al., 1985; Lambert, Hogan, Paoline and Clarke, 2005). Police work can be a dangerous job requiring interaction with people who may sometimes be violent (Youngcourt and Huffman, 2005). Being a police officer can involve dangerous work, which may be emotionally trying (Kinman et al., 2012). Dealing with potentially violent individuals can place a strain on officers (Boles et al., 2001), who face the possibility of being injured or even killed. Perceptions that the job carries risks of harm can psychologically wear down a person, and this can spill over, increasing the chances of strain-based work-on-family conflict. Working in correction facilities is often viewed as being just as dangerous as law enforcement. Lambert and Hogan (2006) found that perceived dangerousness of the job was positively associated with work-on-family conflict among staff at a U.S. government-run maximum security prison. Likewise, Lambert, Minor et al. (2015) reported that perception of the job being dangerous was positively linked to greater conflict at home for correctional officers working at a private U.S. prison. Perceived dangerousness of the job was hypothesized to be positively associated with strain-based conflict among Indian police officers.

As a workplace resource, perceptions of quality supervision can make work more pleasant and aid a person in being more successful in his or her job (Kossek et al., 2011). Supervisors are the first line of management, and they provide guidance, direction, control, structure, and support to their subordinates (Brough and Williams, 2007). Quality supervision can also be a resource for officers, helping them to avoid or deal more effectively with work stressors, which should reduce the level of stress at the end of the workday (Goh et al., 2015; Lambert et al., 2009). Additionally, good supervision is often a pleasant experience, as well as helping an officer be more successful at work (Lambert, Qureshi et al., 2015). Being less stressed from work and having a more positive state of mind result in improvements in the quality of an officer’s personal life. Conversely, a perception of poor supervision can be a stressor, leading to conflict that follows an officer home (Elby et al., 2005; Goh et al., 2015). Poor supervision can be frustrating in and of itself, as well as hampering the an officer’s ability to be successful at the job, which may result in strain that detracts from home experiences (Lambert et al., 2015). Perception of quality supervision was predicted to have a negative effect on strain-based conflict for Indian officers.

Proper training can provide workers with the skills and confidence needed to do their jobs effectively, as well as providing skills, experiences, and knowledge to deal with potentially demanding and stressful situations (Griffin, 2001; Lambert et al., 2009). Quality training can also help officers be more successful at their jobs, which allows them to feel good about themselves and their work. Perceptions of quality work training should result in a more positive work experience, which should follow officers home, resulting in less conflict. By contrast, poor training or a lack of training can result in individuals who feel that they are ill-prepared to deal with job demands. Furthermore, poor training can hamper the ability to do the job correctly and to avoid possible problems (Lambert et al., 2009). Uncertainty and frustration from these negative experiences can follow a person home, leading to further conflict and problems. Perception of quality training was postulated to have a negative association with Indian police officer strain-based conflict.

Job autonomy is the perceived degree of control people feel they have over their job, such as how the job is done or the order in which tasks are undertaken (Schaufeli and Taris, 2014). It can provide a sense of control for workers, as well as allowing them to feel trusted and valued by the organization. Job autonomy can also allow employees flexibility in dealing with the demands and stressors of the job (Lambert and Hogan, 2010; Mauno et al., 2006). Job autonomy should help buffer officers against workplace demands, and allow them to feel valued and trusted by the organization. In turn, this should result in positive feelings, which lead to lower levels of strain following an officer home (Lambert, Qureshi et al., 2015). Conversely, perceptions of low job autonomy may result in frustration and strain for employees who may feel untrusted and confined in their jobs (Schaufeli and Taris, 2014). This strain can follow an officer home, resulting in conflict outside work. Job autonomy was hypothesized to have a negative association with the stain-based conflict among Indian police officers.

Job variety is the degree of variation within a job; some jobs have a high degree of variety, whereas others are more repetitive (Lambert and Hogan, 2010). Job variety allows employees the chance to experience new things, develop and utilize new skills and knowledge, and grow as individuals. This can be an enriching experience. Positive feelings from work may spill over to home (Lambert and Hogan, 2006; Lambert et al., 2015). A lack of job variety can be a stressor, because repetitive jobs are boring and monotonous, which, over time, can psychologically wear down a person, and these unpleasant work experiences can follow an individual home (Schaufeli and Taris, 2014). As such, job variety was hypothesized to have a negative effect.

The demographic characteristics of age, gender, post location, rank, district, educational level, tenure in current position, and marital status were included in the current study. This was done for several reasons. Research suggests that work–family experiences and perceptions of workplace factors may vary by demographic characteristics and many past studies on work–family conflict have included demographic characteristics more as control than explanatory variables (Britt and Dawson, 2005; Dierdorff and Ellington, 2008; Janzen et al., 2007; Lambert and Hogan, 2006; Lambert, Qureshi et al., 2015; Mauno et al. 2006; Rathi and Barath, 2013). For example, women, younger workers, married/partnered individuals, those holding lower or line ranks in the organization, and those with greater tenure have been reported to have higher levels of strain and work–family conflict (Britt and Dawson, 2005; Janzen et al., 2007; Lambert and Hogan, 2006; Lambert, Minor et al., 2015; Ortega et al., 2007; Vinokur et al., 1999). Another reason for including these demographic variables is to explore how much of an effect they have on strain-based conflict among Indian police officers. Demographic characteristics are typically not in the control of police administrators. The other variables in this study, however, can be changed by administrators with varying degrees of difficulty. The amount of variance accounted for by the demographic variables was predicted to be much lower than that accounted for by workplace variables for Indian police officers and it was also predicted that their significant effects would become non-significant once the workplace variables were taken into account.

In this current exploratory study, the direct effects of the aforementioned variables were tested to determine which workplace factors are linked to strain-based conflict among Indian officers. As previously indicated, there has been very limited research on the predictors of strain-based conflict in general, and among Indian police officers in particular. These results are needed in order for police administrators, especially those in India, to understand how workplace factors may contribute either positively or negatively to strain-based conflict. Furthermore, this information is critical if efforts are to be undertaken to reduce strain-based conflict among officers. In addition, our results should help scholars conduct further studies in this area. Knowing which variables are significant predictors and which are not, will allow for more trimmed models to be tested, including interaction effects or differences according to different subgroups, such as women and men. It is important to note that the job demands–resources model does not specify specific workplace demand or resource variables that result in negative outcomes, such as strain-based conflict (Bakker and Demerouti, 2007; Demerouti et al., 2001). A wide array of workplace demands and resources may be linked to strain-based conflict. Exploring how various workplace factors may or may not be associated with this form of domain conflict is, therefore, important, particularly in understudied groups such as Indian police officers.

Method

Variables

Predicted variable

The predicted variable was strain-based conflict, which was measured using an index comprised of six items based on Carlson et al. (2000). The items are presented in Appendix 1, as well as the response options for the items used to measure the predicted variable. Finally, the responses to the strain-based conflict items were summed together to form an additive index.

Predictor variables

The predictor variables were summed indexes of time-based work–family conflict, behavior-based work–family conflict, role conflict, role ambiguity, role overload, role underload, perceived dangerousness of the job, perceptions of supervision, views on training, job autonomy, and job variety. Time-based conflict and behavior-based conflict were each measured using five items derived from Carlson et al. (2000). The items for role conflict, role ambiguity (four items), role underload (two items), and role overload (two items) were adapted from Triplett et al. (1996). Perceived dangerousness of the job was measured using four items from Cullen et al. (1985). Perceptions of fair and supportive supervision were measured using six items adapted from Teas (1981). Views of training were measured with two items from Lambert et al. (2009). Finally, job autonomy and job variety were both measured using two items, and these items were from Curry et al. (1986). All the items used to create the predictor indexes are presented in Appendix 1, together with the response options for the items.

As previously indicated, the demographic characteristics of age, gender, post location, rank, district, educational level, tenure in current position, and marital status included more as control variables than explanatory variables in this study. See Table 1 for how these personal characteristic variables were coded.

OPEN IN VIEWER

Table 1.

Descriptive statistics of study variables.

Variable	Description	Min	Max	Med	Mean	SD
Age	Age in years	21	57	34	36.53	9.46
Gender	88% male (coded 1) 12% female (coded 0)	0	1	1	0.88	0.33
Posting location	Posting Location 50% police station (coded 1) 50% other location (coded 0)	0	1	0.50	0.50	0.50
Rank	69% line officer (coded 1) 31% other position (coded 0)	0	1	1	0.69	0.46
District	50% Rohtak District (coded 1) 50% Sonipat District (coded 0)	0	1	0.50	0.50	0.50
Education level	42% college degree (coded 1) 58% no college degree (coded 0)	0	1	0	0.42	0.49
Tenure	Tenure in years for current position	0	30	11.20	12.44	3.05
Marital status	85% currently married (coded 1) 15% not married (coded 0)	0	1	1	0.85	0.36
Time-based	5 item additive index, α = .65	5	25	19	18.49	3.72
Behavior-based	5 item additive index, α = .60	5	25	15	15.32	3.72
Role conflict	4 item additive index, α = .62	4	20	12	11.89	3.18
Role ambiguity	2 item additive index	2	10	5	5.26	1.73
Role overload	2 item additive index	2	10	4	4.77	1.75
Role underload	2 item additive index	2	10	4	4.77	1.75
Dangerousness	4 item additive index, α = .67	4	20	15	14.44	3.33
Supervision	6 item additive index, α = .70	6	30	18	17.89	4.54
Training	2 item additive index	2	10	7	7.04	1.98
Job autonomy	2 item additive index	2	10	6	5.34	1.87
Job variety	2 item additive index	3	10	8	7.94	1.41
Strain-based	6 item additive index, α = .75	6	30	18	18.06	4.95

Note. Min, minimum value; Max, maximum value; Med, median value; Mean, mean value; Tenure, tenure in position; Time-based, time-based work–family conflict; Behavior-based, behavior-based work–family conflict; Dangerousness, perceived dangerousness of the job; Supervision, perceptions of supervision; Training, views of training; Strain-based, strain-based work–family conflict; α is the internal reliability value. The total number of participants was 827.

Results

The descriptive statistics for the variables used in the study are reported in Table 1. There appeared to be significant variation in the dependent and independent variables (i.e., none were constants). The data conformed approximately with a normal distribution based on skewness and kurtosis statistics. Likewise, the median and mean values for the variables are similar, also suggesting a normal distribution. Cronbach’s alpha, a measure of the internal reliability for indexes with three or more items, is presented in Table 1. No Cronbach’s alpha values were calculated for indexes with only two items (Eisinga et al., 2013). For each latent index variable, the items were entered into a factor analysis, and the items loaded on a single factor, indicating the unidimensionality of the items and convergent validity (Gorsuch, 1983). The factor scores for each item are reported in Appendix 1.

A correlation matrix is presented in Table 2. Post location, district, time-based conflict, behavior-based conflict, role conflict, role ambiguity, role overload, dangerousness, supervision, training, and job autonomy all had statistically significant correlations with strain-based conflict. Officers at station locations and those in the Rohtak district reported greater strain on average. Both time-based and behavior-based conflict were correlated with higher levels of strain-based conflict. Similarly, role conflict, role ambiguity, role overload, and dangerousness were also positively correlated, which means that increases in each were associated with greater strain-based conflict. Conversely, higher levels of supervision, training, and job autonomy were associated with lower forms of conflict at home. The remaining predictor variables had non-significant correlations with the strain-based variable.

OPEN IN VIEWER

Table 2.

Correlation table of study variables.

Variable	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19
1	1.00
2	.20**	1.00
3	.05	.02	1.00
4	–.68**	–.11**	–.09**	1.00
5	–.05	–.06	.32**	.10**	1.00
6	–.32**	–.17**	–.20**	.18**	–.15**	1.00
7	.22**	.01	.00	–.14**	.06	–.11**	1.00
8	.37**	.05	–.02	–.17**	–.07*	.00	13**	1.00
9	–.11**	.04	.11**	.05	.07	.04	–.03	.00	1.00
10	–.13**	.07*	.02	.05	–.01	.05	.01	–.09**	.36**	1.00
11	–.11**	.08*	.06	.05	.14**	.02	–.04	–.08*	.24**	.43**	1.00
12	–.09*	.10**	–.03	.04	.04	.02	–.01	–.09**	.21**	.38**	.35**	1.00
13	–.11**	.03	.13**	.09*	.13**	.03	.02	–.04	.45**	.26**	.18**	.05	1.00
14	.04	.02	.00	–.03	–.02	–.06	–.04	–.03	–.19**	.14**	.22**	.07	–.21**	1.00
15	–.07*	.03	.23**	.05	.19**	–.01	–.06	–.08*	.41**	.22**	.17**	.02	.36**	–.07	1.00
16	.11**	–.08*	–.05	–.15**	–.19**	.01	.03	.05	–.29**	–.16**	–.33**	–.29**	–.21**	.04	–.21**	1.00
17	.07	.02	.06	–.07*	–.01	–.01	–.02	–.01	–.05	–.08*	–.19**	–.32**	.01	–.12**	.15**	.18**	1.00
18	.07*	–.04	–.09*	–.04	–.11**	.05	.00	.06	–.41**	–.08*	–.14**	–.20**	–.25**	.22**	–.20**	.36**	.13**	1.00
19	–.01	.01	.04	–.01	–.05	.01	–.04	.00	.15**	.10**	–.10**	–.13**	.08*	.11**	.13**	.06	.15**	–.07	1.00
20	–.06	.06	.11**	.05	.11**	–.04	.01	.05	.64**	.40**	.38**	.29**	.38**	–.03	.26**	–.29**	–.23**	–.37**	–.01

Note. Var, variable; 1, age; 2, gender; 3, post location; 4, rank; 5, district; 6, educational level; 7, tenure in position; 8, marital status; 9, time-based conflict; 10, behavior-based conflict; 11, role conflict; 12, role ambiguity; 13, role overload; 14, role underload; 15, perceived dangerousness of the job; 16, perceptions of supervision; 17, views of training; 18, job autonomy; 19, job variety; 20, strain-based conflict. See Table 1 for the descriptive statistics of the study variables. The number of participants was 827.

*P ≤ .05; **P ≤ .01.

Four multivariate regression models were estimated and the results are reported in Table 3. Multicollinearity occurs when two or more variables share too large an overlap in variance, which can affect regression results. Multicollinearity is seen as a problem when variance inflation factor (VIF) scores exceed 5 (Tabachnick and Fidell, 1996). Because the highest VIF score was 2.40, multicollinearity was not a problem.

OPEN IN VIEWER

Table 3.

Ordinary least squares regression results on predicting strain-based work–family conflict among Indian police officers.

Predictor variables	Model 1		Model 2		Model 3		Model 4
	B	β	B	β	B	β	B	β
Demographic characteristics
Age	–0.07	–.14**					0.01	.02
Gender	1.22	.08*					–0.32	–.02
Post location	0.87	.09*					0.16	.02
Rank	–0.11	–.01					0.23	.02
District	0.78	.08*					–0.07	–.01
Educational level	–0.40	–.04					–0.66	–.07*
Tenure in position	0.01	.01					0.00	.00
Marital status	1.50	.10*					0.88	.05
Workplace demands
Time-based conflict			0.70	.53**			0.62	.46**
Behavior-based conflict			0.10	.07*			0.17	.13**
Role conflict			0.28	.18**			0.24	.15**
Role ambiguity			0.24	.09**			0.12	.04
Role overload			0.08	.06			0.31	.11**
Role underload			0.03	.01			0.06	.02
Dangerousness of job			–0.03	–.02			–0.02	–.01
Workplace resources
Perceptions of supervision					–0.18	–.16**	–0.01	–.01
Views of training					–0.39	–.16**	–0.28	–.11**
Job autonomy					–0.80	–.31**	–0.36	–.14**
Job variety					0.01	.01	–0.18	–.05
R-squared	.04**			.48**		.20**		.53**
Adjusted R-squared	.03**			.47**		.19**		.52**

Note. B, unstandardized regression coefficient; β, standardized regression coefficient. See Table 1 for the descriptive statistics for the variables. The number of participants was 827.

*P ≤ .05; **P ≤ .01.

For Model 1, the R-squared value was .04, which meant that the eight demographic variables only explained ∼4% of the variance observed in the strain-based conflict measure. Age, gender, post location, district, and marital status were all statistically significant predictors. Older officers reported less strain-based conflict. Men in general reported higher strain at home. Likewise, officers working at a station house and in the Rohtak district reported increased domain conflict. Married officers tended, on average, to report more strain conflict at home. Rank, educational level, and tenure in the position all had non-significant effects.

For Model 2, the predictor variables accounted for ∼48% of the variance in the strain-based conflict index. Time-based conflict, behavior-based conflict, role conflict, and role ambiguity were significant predictors. Increases in each of these demands were associated with higher levels of reported strain-based conflict at home. Role overload, role underload, and perceived dangerousness of the job were not significant predictors.

For Model 3, the workplace resource variables accounted for ∼20% of the variance of strain-based conflict. Perceptions of supervision, views of training, and job autonomy were significant predictors. Increases in each of these variables were associated with reductions in the level of strain-based conflict reported. Job variety had a non-significant effect.

For Model 4, the full model, the R-squared value for the regression equation was .53, which means that the predictor variables explained ∼53% of the observed variation in the strain-based index. Among the personal characteristics, only educational level was a statistically significant predictor; those with a college degree generally reported lower strain conflict. Among the demand variables, time-based conflict, behavior-based conflict, role conflict, and role overload each had positive associations with the predictor variable, which means increases in each were associated with greater conflict at home. Role ambiguity, role underload, and perceived dangerousness of the job had non-significant relationships with the predictor variable. Among the resource variables, views of training and job autonomy had significant associations, both negative, which means increases in each were associated with lower strain-based conflict. Finally, perceptions of supervision and job variety had non-significant effects.

Discussion and conclusion

The balance between work and home is often difficult to maintain, and this can especially be true for police officers due to the nature of their work and work conditions. When conflict in one domain spills over into the other, work–family conflict occurs. One type of work–family conflict is strain-based conflict, where strain from work causes conflict at home and affects the quality of life for officers. With strain-based work–family conflict, there is an increased likelihood of negative consequences for officers, their families, and their work performance. Prior research has found that work–family conflict, especially strain-based conflict, can impact the ability of officers to succeed in the workplace, such as higher rates of absenteeism, job turnover, decreased work performance, and diminished organizational commitment (Burke, 1988, 1994; Elby et al., 2005; He et al., 2002; Martinussen et al., 2007). Work–family conflict can be an issue for police officers (Burke, 1988; Hall et al., 2010; He et al., 2002; Mikkelsen and Burke, 2004). Officers suffering from strain-based work conflict can sustain mental and physical disabilities. Finally, work–family conflict may also impact the officers’ family relationships through reduced marital satisfaction. As such, reduction of this conflict can significantly benefit officers, their families, and their police agencies.

This exploratory study examined the possible predictors of strain-based work–family conflict for police officers in in two districts in India. When only demographic variables were used (see Model 1 in Table 3), several were significant; however, as predicted, the amount of variance explained by strain-based conflict was very low. This indicates that demographic variables may not have much of a direct role in shaping domain conflict problems at home. In the full model (see Model 4 in Table 3), educational level was the only demographic characteristic to have a significant association with the predicted variable. Based on the standardized regression coefficient (i.e., β), it had the smallest effect. The vast majority of personal characteristics lacked a non-significant association with strain-based conflict. The full multivariate model supports the job demands–resources model, which holds that workplace variables are more important in shaping the level of strain felt by officers. Moreover, this finding is good news for police administrators because it is easier to change workplace variables than demographic characteristics. Educational level had a negative relationship with strain-based conflict, which is an interesting finding. Those with a college degree might be paid more, allowing them greater financial resources to deal with the conflict that spills over at home. A college education might provide officers with the experiences and skills needed to deal more effectively with workplace demands, reducing their negative effects at home. More educated officers might just have better jobs, which result in lower strain-based conflict. Also, the finding might be spurious or the result of random chance. Additional research is needed to confirm the negative relationship between educational level and strain-based work–family conflict and to determine the reasons for it.

In the full model, some demand variables were positive predictors. Specifically, time-based work–family conflict was positively associated with strain-based work–family conflict. The standardized regression coefficient for this variable indicates that time-based conflict is by far the strongest predictor of strain-based conflict among the surveyed Indian police officers, a finding consistent with those reported by Borum and Philpot (1993) and Violanti and Aron (1995) among Western police officers. Police work is performed seven days a week and on public holidays, and officers are often required to work at times that conflict with family schedules. This is even more pronounced in the study site, because Indian officers are required to be available 24 hours a day during the scheduled working week, which is not found among law enforcement agencies in Western nations. Working such extreme hours leads to conflict at home for Indian police officers. The current findings suggest finding a more manageable work schedule for Indian police officers to reduce their conflict at home.

Similarly in the full model, behavior-based conflict was also positively associated with strain-based work–family conflict, this is not surprising because the task environment has consistently been found to be a source of stress in police agencies in other nations (Brooks and Piquero, 1998; Crank and Caldero, 1991; Violanti and Aron, 1995). Police work requires officers to interact with citizens in situations that are often adversarial, authoritarian, and antagonistic. There is an occupational need to suppress emotions (Morash et al., 2006). Officers can develop a working personality that conflicts with the demeanor appropriate for home life and interactions with family members (Skolnick, 1966). These work behavioral roles are not conductive to productive and healthy interactions with others at home. Officers in our study indicated difficulty in shedding their police personality when they return home. The current findings suggest the need to intervene and reduce the impact of behavior-based conflict causing conflict in the home lives of officers. Police administrators should consider implementing workshops that explain behavior-based conflict and how to reduce its effects at home. Research is needed to explore how to best reduce the effects of work roles that spill over and cause conflict at home.

Role conflict and role overload were also found in Model 4 to be significant positive predictors of strain-based conflict, and both originate from within the police agency, because both are considered to be internal organizational stressors (Haarr and Morash, 1999; Jaramillo et al., 2005). Conflicting supervisor directives and excessive work demands likely create stress, and the current findings suggest that this stress follows Indian police officers home, causing conflict. Police administrators need to be aware that these two stressors are not limited to workplace harm; they spill over, causing problems at home. A negative home life could spill back, resulting in decreased work performance and increased turnover (Lambert, Hogan, Paoline and Baker, 2005). As such, these stressors need to be reduced. Police administrators should hold focus groups with officers to learn what is causing role conflict and role overload, so that these can be addressed. In addition, supervisors need to be trained to be aware of how to avoid contributing to these stressors, as well as recognizing their occurrence so that interventions can quickly be put into place to reduce their negative effects. For example, there could be greater coordination between different supervisors, so that line officers do not feel that one supervisor’s orders conflict with those of another supervisor. As such, feedback from officers about when supervisors contradict one another would probably be helpful. Studies are needed to examine the effectiveness of initiatives to stop role conflict and role overload from occurring and, when they do occur, to reduce their harmful effects on officers. Further research is needed among police in other nations to determine whether the effects of the two role stressor variables are universal across a wide array of police agencies in different cultures.

The workplace resource variables for views on training and job autonomy were also significant predictors of strain-based work–family conflict in the full model. Both had negative associations, which means that they appear to reduce home conflict. Conversely, if they are missing or low, strain at home is more likely to occur. Positive views of training probably mean that employees feel that they have the skills needed to be effective in their job. Feeling confident at work is likely to follow a person home, helping create a more pleasant experience. A perception of poor training probably results in feeling ill-prepared at work, resulting in strain, which follows the person home, decreasing home life satisfaction. Likewise, perceptions of job autonomy appear to reduce strain at home for Indian police officers. Allowing people some control over their jobs probably results in positive feelings at work that follow the person home. There is no reason why positive work experiences should remain at work. A lack of autonomy can be a confining and frustrating experience both at work and at home. When a person walks through the proverbial exit from work, it is unreasonable to expect work experiences or feelings to remain behind. They are part of the person and will follow the individual home and continue to affect him or her. Indian police administrators need to be aware that different workplace variables appear to play a role in strain-based conflict for officers. Efforts need to be undertaken to enhance these two variables. Feedback from officers concerning training and job autonomy may be one avenue to increasing positive perceptions of the two variables. Officers should be asked how to improve training realistically and allow greater job autonomy. In addition, supervisors need to be made aware of the importance of these workplace variables. Supervisors are often the linchpin between line staff and administrative staff. Research is needed to explore whether the effects of these variables are limited to the surveyed Indian police officers or apply to officers in different police agencies.

Based on the block entry of the predictor variables (i.e., Models 1–3 in Table 3), demand variables explained the greatest amount of variance of strain-based conflict. Workplace resource variables explained the second most variance. As previously indicated, demographic variables explained the least amount of variance. This suggests that demand variables contribute most to strain-based conflict among Indian police officers. As such, efforts are needed to reduce the demands faced by officers. Role ambiguity was a significant predictor in Model 2, but not in Model 4, suggesting that resources may help reduce the workplace ambiguity faced by officers in this study. Interestingly, in Model 3, perception of supervision was a significant predictor, but this was not the case in Model 4. This suggests that once demands are taken into account, this resource is not as important as training and job autonomy. It may be that supervision can help reduce demands, but does not play a direct role in strain-based conflict once other variables are taken into account. In Models 2 and 4, role underload and perceived dangerousness of the job were not significant predictors. This suggests that these workplace demands may not play a role in shaping strain-based conflict in Indian officers. It might be that role underload is not a great negative demand on officers compared with role overload. Officers are more likely to be become stressed when dealing with too many tasks, than bored dealing with too few. As for perceived dangerousness, officers may have expected the job to be dangerous and, as such, are not as strained by this demand. Similarly, job variety was not a significant predictor in either Model 3 or Model 4. Again, this resource may not play a direct role in reducing strain-based conflict among the Indian police officers in this study.

Overall, the current study provides support for the job demands–resources model among Indian police officers. As previously indicated, six variables, four representing demands and two representing resources, were significant predictors of strain-based conflict. The current findings suggest that both demands and resources play a role in affecting Indian officers. It is also important to note that the effects of specific workplace variables varied. Five workplace variables (role ambiguity, role underload, perceived dangerousness of the job, supervision, and job variety) had non-significant associations with the dependent variable in the multivariate analysis. It might be that these variables are not important in predicting strain-based conflict when the other workplace variables are taken into account. The non-significant results might have been due to chance. Additional studies are needed to confirm the results. Even if these workplace variables are found to be non-significant predictors of strain-based conflict in subsequent studies, this does not imply that there are not important workplace variables. There may important variables shaping job satisfaction, job involvement, and organizational commitment of police officers.

As with many others, the current study had limitations. It was a single study of officers in two districts in the State of Haryana, India. The findings may be situational and contextual, varying across different police districts and agencies. Studies among officers in other districts and police agencies are needed to determine whether the findings are replicated elsewhere. The use of cross-sectional surveys does not allow for causality to be demonstrated. Longitudinal studies are needed to demonstrate the causal relationships of the predictor variables and the predicted variable empirically. The current study was exploratory and only examined the direct effects of variables on strain-based conflict, which limits the findings. Future research is needed to explore the indirect effects of workplace demands and workplace resources on strain-based conflict. This would require reducing the number of demand and resource variables used. Future research in this area may use the current results to select salient variables and determine how they interact in their effects on domain spillover for officers. In addition, future studies should explore whether workplace resources moderate the effects of work–family conflict, including strain-based conflict on possible outcomes for Indian police officers, such as job satisfaction, organizational commitment, well-being, etc. (Mauno et al., 2006).

Although the current study labelled behavior-based conflict as a workplace demand, this may not be entirely accurate based on how this construct was measured. As previously indicated, behavior-based conflict occurs when work and home roles are incompatible with one another (Brough and O’Driscoll, 2005; Netermeyer et al., 1996). This form of domain conflict is the result of role expectations in one domain causing problems in the roles needed for the other domain. As Allen et al. (2012: 18) noted that ‘behavior-based conflict occurs when behaviors exhibited in one role are incompatible or incongruent with behavior patterns needed in the other role.’ This suggests that role problems are bidirectional between work and home. In the current study, two items asked about work roles causing problems at home and one item asked about home roles resulting in conflict at work. Future research may wish to break behavior-based conflict into two variables, one measuring work roles causing problems at home and the other measuring home roles causing problems at work. This should provide a more detailed picture of how this form of work–family conflict affects police officers. At the very least, the behavior-based conflict variable in the current study may not be a workplace demand per se, but rather demand overall, which is broader than just work.

Several of the index variables had lower than expected values for internal reliability. This might be a result of cultural differences with Indian police officers. While all were above the .60 rule-of-thumb cut-off level, several had alpha values below .70. Whereas the factor analysis results indicated unidimensionality and convergent validity, the Cronbach’s alpha values need to be improved in future research. Future studies should explore adding new items to the indexes to increase the alpha values, as well as determine whether cultural differences exist in the perceptions of these concepts. There is a need to explore how other workplace variables, such as instrumental communication, formalization, organizational justice, input into decision-making, and social support are associated with strain-based conflict. Additional research on work–family conflict among police officers is needed, including possible antecedents and the consequences of the different forms and dimensions of work–family conflict. Studies need to examine the effectiveness of various interventions to reduce work–family conflict or to deal with it. There is a need for far more research in this and other areas not only among Indian officers, but also police in other nations.

In closing, officers are the most important resource for any police agency. Officers face many different job demands, and the effects of these demands can spill over to their home life, resulting in strain-based conflict. Based on our data from a survey of officers from the Sonipat and Rohtak districts in Haryana, India, the job demands of time-based conflict, behavior-based conflict, role conflict, and role overload had positive associations with the strain dimension of work–family conflict. In addition, the workplace resources of views of training and job autonomy had negative associations with strain-based conflict. The findings indicate that Indian police administrators should take note of the associations between selected workplace factors and strain-based conflict. At the very least, the authors hope that this study will encourage more research on how workplace factors are associated with strain-based conflict for police officers not only in India, but also in other nations. With evidence-based practices, India should be able to reduce strain-based conflict among police officers. In the end, too much is at stake to ignore how workplace factors may be linked with strain-based conflict not only among Indian officers, but also those across the globe. Finally, additional studies across nations will provide salient information of which workplace factors are contextual on their effects and which ones are more universal in their effects.