Immigrants and job satisfaction: Do high performance work systems play a role?

Abstract

This study focuses on immigrants in Canada and examines their job satisfaction as compared to Canadian-born employees. The authors explore whether high performance work systems (HPWS) affect immigrant employees’ job satisfaction differently than Canadian-born, where HPWS include empowerment, motivation and skill enhancing sub-bundles of practices. The study uses Statistics Canada’s Workplace and Employee Survey, a longitudinal dataset consisting of six years of data from 1999 to 2004 and 49,344 employees. Ordered logistic regression analyses are used to explore causal relationships. The moderation results show that increases in the motivation sub-bundle and increases in overall experience of HPWS practices are related to increases in job satisfaction among immigrants when compared to Canadian-born employees. Immigrants are a segment of the workforce that is understudied in management. This study identifies HPWS bundles that positively affect immigrants’ job satisfaction, suggesting a role for managers to capitalize on job satisfaction improvements ultimately contributing to organizational success.

Keywords

Diversity high performance work systems immigrants job satisfaction motivation

With the ageing population and fewer young workers to replace retiring workers in Canada, immigration is looked to as the main driver for labour force growth (Picot, 2008). The importance of immigrant workers in the labour force can be demonstrated by comparison: for example, immigrant workers account for one in five of the workforce in Canada (Picot, 2008), one in eight in the USA (Bell et al., 2010), one in 10 in the UK (OECD, 2008) and one in 10 in Ireland (Dundon et al., 2007). The magnitude of these ratios highlights the importance of immigrants as a group of workers within the workforce. With immigrant workers already a large minority of the labour force in Canada, and expected to increase their percentage in the labour force, our study focuses on the job satisfaction of these employees.

For organizations, job satisfaction is particularly important for retention of valuable employees (Tett and Meyer, 1993). Whether hiring immigrants or Canadian-born employees, there are recruitment, orientation and training costs involved, and when a valuable employee leaves, future anticipated benefits are lost. Because immigrants are the main drivers of labour force growth in Canada, we ask the question: Are immigrants satisfied with their jobs? In our analysis, we compare immigrant employees’ job satisfaction to their Canadian-born counterparts.

To contribute to organizational performance, organizations use a variety of human resource practices, called high performance work systems (HPWS), that focus on employee empowerment, motivation and skill development (Appelbaum et al., 2000; Jiang et al., 2012; Subramony, 2009). Research shows some support for HPWS’ positive effect on job satisfaction (Kooij et al., 2013; Macky and Boxall, 2007; Mohr and Zoghi, 2008; Zatzick and Iverson, 2011), though positive findings are not persistent and the results can be mixed when human resource practices are examined individually (i.e. some practices have a significant positive effect while others have no effect or a significant negative effect) (Brown et al., 2008; Wood and de Menezes, 2011). There are no studies on the effects of HPWS on immigrants’ job satisfaction, and we ask: Do HPWS play a role in immigrants’ job satisfaction, and if yes, what would be the effect of HPWS on immigrant employees’ job satisfaction compared to Canadian-born employees?

The purpose of this study is to examine immigrants’ job satisfaction compared to Canadian-born employees, and explore the effect of HPWS on immigrant employees’ job satisfaction as compared to Canadian-born employees. A lower level of job satisfaction for immigrants would be consistent with challenges immigrants face in the labour market (Knocke, 2000). Some of the challenges faced include difficulty re-establishing their careers (Hakak et al., 2010; Salaff et al., 2002), inadequate or inappropriate jobs given their education and experience (Schellenberg and Maheux, 2007), underemployment and over-qualification (Aycan and Berry, 1996; Erdogan and Bauer, 2009; Erdogan et al., 2011; Galarneau and Morissette, 2008), and earnings generally lower than Canadian-born (Aydemir and Skuterud, 2008; Picot, 2008). We suggest that if lower levels of job satisfaction exist for immigrants relative to Canadian-born then it is important to understand whether the moderation effect of HPWS on immigrants may mitigate some of the attenuation in their job satisfaction that is related to the additional challenges they face. We adapt the Erdogan and Bauer (2011) framework to focus on what role HPWS practices play in ameliorating the negative effects associated with immigrant status on job satisfaction. In particular, Erdogan and Bauer (2011) identify work and employment practices (e.g. psychological contract, pay-for-performance and employee’s work situation) that moderate the pathway between underemployment and turnover. We use the 1999–2004 longitudinal Workplace and Employee Survey from Statistics Canada, which is representative of the Canadian workforce with some exceptions (as discussed in the Method section).

Definitions for immigrants tend to vary depending on the source and the perceptions of individuals referring to immigrants. In this study, immigrants are defined as those who were born outside of Canada and have legal status to reside and work in Canada. This definition is similar to the one used by the OECD (2008) and Dietz (2010). Our definition excludes illegal (or undocumented) migrant workers. We use the term ‘immigrant status’ to refer to legal immigrant workers (permanent residents), where the implied reference group is Canadian-born workers. In Canada, immigrant status is an important element of diversity and is appropriately becoming a part of the notion of managing employees with different demographic characteristics within the workplace (Jones et al., 2000). Further, we focus on immigrant arrival cohorts (discussed in more detail below), and take a ‘first-step’ toward understanding one of many immigrant characteristics. Our data do not allow us to conduct analysis on additional immigrant characteristics such as: immigrant category/class, ethnicity composition, country of origin, language skills and proficiency, occupation groupings, type of employment (e.g. self-employment), career profiles, human capital levels and social integration/networks, for example. The exploration of these factors and their intersection, in future research, will likely provide additional understanding of the dimensionality of factors affecting worker outcomes.

This study is important for two reasons. First, immigrants are an increasingly significant component of the workforce in organizations, yet the immigrant segment of the workforce is understudied in management research (Bell et al., 2010; Dietz, 2010), including only a few studies on immigrants’ job satisfaction. Fang et al. (2009) focus on immigrants’ job satisfaction, but they cover only immigrant professionals not the broader group of immigrants that are the focus of the current study. Kats (1983) explores the effects of underemployment on job satisfaction of male Russian immigrants to Israel. Valdivia and Flores (2012) focus on Latino/a immigrants’ job satisfaction levels in three US rural Midwest communities. Our study builds on this emerging literature and contributes to knowledge on immigrants’ job satisfaction. Second, the importance of the present study is that despite growing interest in the literature on the role of HPWS on employees’ job satisfaction (see Macky and Boxall, 2007; Wood and de Menezes, 2011), there are no studies examining the moderation effect of HPWS practices on the immigrant and job satisfaction relationship. However, studies are beginning to explore how different demographic characteristics, such as age, moderate human resource management practices and affect job satisfaction (e.g. Kooij et al., 2013). This study contributes to knowledge by looking at the causal relationship between the interaction effect of HPWS practices and immigrant status on job satisfaction using longitudinal data. This is an important topic of research because it is important to know whether or not HPWS practices can be used to mitigate some of the attenuation in job satisfaction between different groups.

Theory and hypotheses

The theoretical foundations of job satisfaction are well established in the literature. As the following discussion highlights, many of the antecedents of job satisfaction that have been identified are factors that can be enhanced by the implementation of HPWS practices. Job satisfaction is an attitude that can be best described as a person’s response to what happens to them at work (Lawler, 2005). Job characteristics that challenge an employee’s skills, allow for whole job completion, enable the employee to see the significance of their work, permit discretion and provide feedback have been identified as affecting job satisfaction (Hackman and Oldham, 1980). Work practices that specifically focus on enhancing employee involvement (i.e. job enrichment, autonomy and information exchange) have also been found to have a positive relationship with job satisfaction (Humphrey et al., 2007; Zatzick and Iverson, 2011). While some have focused on work practices (e.g. team work, autonomy and high skill) others have focused on employment practices (e.g. recruitment, selection, allocation, motivation, negotiation, training/development and layoffs/quits) and in particular motivation and rewards. For example, Lawler’s model states that job satisfaction is affected by the rewards provided by the organization and workers’ feelings about the value of those rewards for them (Lawler, 1973). Originally based on expectancy theory (Campbell et al., 1970), Lawler’s model states that individuals will put forth the greatest effort if they expect the effort to lead to performance that will be rewarded in a desirable way; and the reward, in turn, will contribute to job satisfaction (Lawler, 2000).

Empirical research testing the theoretical foundations of job satisfaction with immigrant employees is rare. A few studies have identified how: (1) immigrant professionals are less satisfied with their jobs (Fang et al., 2009), (2) male immigrant underemployment lowers job satisfaction (Kats, 1983) and (3) ethnic identity and Anglo acculturation positively affect Latino/a immigrants’ job satisfaction (Valdivia and Flores, 2012). In Canada, research on employment conditions of immigrants shows that they often experience difficulty in re-establishing their careers upon immigrating (Hakak et al., 2010; Salaff et al., 2002). For example, relative to Canadian-born workers, immigrants with education, experience and/or certification as professionals or technicians in their home country find the mechanisms to integrate into their occupation in Canada difficult to navigate in terms of upgrading or re-certifying (Ferrer et al., 2006; Salaff et al., 2002). New immigrants report finding an adequate job or appropriate job to be one of their main difficulties (Schellenberg and Maheux, 2007). They also experience underemployment; this is particularly common among skilled immigrants (Aycan and Berry, 1996; Galarneau and Morissette, 2008). In particular, among Canadian-born women and men, of those with a university degree, respectively only 13% and 12% had jobs that required lower education, whereas for recent immigrants with a university education 38% of women and 25% of men were in low-education jobs (Galarneau and Morissette, 2004). Underemployment, and the related over-qualification, of employees have been linked to lower levels of job satisfaction (Erdogan and Bauer, 2009; Erdogan et al., 2011). Generally, immigrants’ earnings are much lower than Canadian-born employees and these earnings gaps are persistent (Aydemir and Skuterud, 2008; Picot, 2008). For example, for full-time and full-year workers, immigrants’ convergence to Canadian-born earning levels has been declining over time as more recent immigrants have increasingly lower relative earnings in their initial jobs five years after arrival (i.e. falling from about 86% in 1975 to 62% in 2000, immigrant earnings over Canadian-born earnings) (Picot, 2008). Gaps in earnings raise concerns not only with regard to economic citizenship, but also with social survival; and this implies that both economic convergence with Canadian-born may be as important to immigrants as political and cultural expression (Mann, 2005). With respect to wages and other workplace rewards, we argue that these factors may be more salient for immigrants in that they are not only linked to economic value but also their perceived value within society (Mann, 2005). We argue that these working conditions and work experiences negatively affect immigrants’ job satisfaction.

Applying the job characteristics (Hackman and Oldham, 1980), reward–job satisfaction (Lawler, 1973) and underemployment models (Erdogan and Bauer, 2009; Erdogan et al., 2011) to immigrant employment, and based on the review of the above literature on immigrants’ employment conditions, we argue that immigrants in Canada might not be employed in jobs with the characteristics and rewards they believe they deserve, and thus, might be less satisfied with their jobs as compared to Canadian-born employees. Based on the above discussions, we hypothesize that:

H1: Immigrants will have lower job satisfaction than Canadian-born employees.

In his introduction to the special issue on employment discrimination against immigrants, Dietz (2010: 105) says, ‘all immigrants are alike and, more importantly, they are not’. He asks whether, in researching immigrants, all should be formed into one group or whether there should be some other meaningful sub-grouping of immigrants. In the context of our study this issue becomes important. Research shows that immigrants arriving more recently do experience a number of labour market difficulties (Schellenberg and Maheux, 2007), such as in finding employment and access to good quality jobs with good pay and benefits (Aydemir and Skuterud, 2008; Galarneau and Morissette, 2008; Picot, 2008) and promotion (Fang et al., 2009). These factors can affect the job satisfaction of recent immigrants as compared to earlier immigrants who are more established in Canada. Recent immigrants are defined as those with year of migration between 1990 and 2004 –earlier immigrant cohorts are defined in more detail below in the Method section. We argue that there will be differences between recent immigrants as compared to earlier immigrant cohorts in terms of job satisfaction and because recent immigrants tend to have poorer employment experiences we hypothesize that:

H2: Recent immigrants will have lower job satisfaction than earlier immigrants.

There is no standard definition as to what is part, or not part, of high performance work systems. Some studies include a comprehensive list of human resource practices as HPWS (see, for example, Berg, 1999; Macky and Boxall, 2007), while others focus on a subset of HPWS such as high involvement work systems (Zatzick and Iverson, 2011) and economic incentives (Wood and de Menezes, 2011). In the current study, we include both a comprehensive list of HPWS practices consisting of job rotation, quality circle, task team, self-directed workgroup, employee survey, having a suggestion programme, informing employees about workplace changes, formal performance appraisal, promotions, non-wage benefits, pay-for-performance, on-the-job training and classroom training, and subsets of HPWS practices. For the subsets, we adopt Subramony’s (2009) conceptualization of HPWS practices as comprising three main human resource management bundles: empowerment-, motivation- and skill-enhancing HPWS practices – the practices included in each bundle are discussed in detail below in the Method section. We also refer to an overall HPWS experience which includes all HPWS practices in the subsets (i.e. as an aggregate HPWS practices index). The use of these practices and bundles to measure the domain of HPWS is consistent with previous job satisfaction and HPWS studies (Wood and de Menezes, 2011; Zatzick and Iverson, 2011).

Many of the individual HPWS practices and bundles of HPWS practices have been found to have a positive relationship with job satisfaction (Appelbaum et al., 2000; Berg, 1999; Macky and Boxall, 2007; Mohr and Zoghi, 2008) though Macky and Boxall (2007) and Wood and de Menezes (2011) recommend caution in concluding that more HPWS practices will always have positive effects on employees. Empowerment-enhancing HPWS practices have been found to have a positive relationship with job satisfaction both when examined individually (Mohr and Zoghi, 2008; Wood and de Menezes, 2011) and in groups (Macky and Boxall, 2007; Mohr and Zoghi, 2008; Zatzick and Iverson, 2011). With regard to the motivation-enhancing HPWS practices, research shows that formal performance appraisals that embody aspects of motivation are positively related to job satisfaction (Brown et al., 2010). Further, performance appraisals resulting in a promotion (Clark, 1997) and an individual pay-for-performance increase (Heywood and Wei, 2006) have also been identified as having a positive relationship with job satisfaction. Non-wage employee benefits are a source of motivation for employees and research shows that non-wage benefits positively affect job satisfaction (Artz, 2010; Heywood and Wei, 2006). For the skill-enhancing HPWS practices, employer-sponsored formal training has been found to lead to increases in job satisfaction (Georgellis and Lange, 2007; Jones et al., 2009), and on-the-job training is shown to have a positive relationship with job satisfaction while formal classroom training has no significant relationship with job satisfaction (Berg, 1999).

While HPWS practices can lead to the above identified positive effects, such as increased engagement, cooperation and productivity among workers, it is fair to note that this view is not universally accepted. For instance, Godard (2004) outlined a number of possible scenarios in which HPWS practices could be harmful, or at least disconcerting to workers, in terms of extra role stress, work intensification and/or monitoring by management, for example. That said, and while accepting that possibility, the more common view, among academic researchers, is that the use of HPWS can yield benefits for employees and employers.

There are no studies that look comprehensively at HPWS and immigrants as compared to non-immigrants. Erdogan and Bauer’s (2011) propositions regarding psychological contract, pay-for-performance and improvement in an employee’s situation all suggest a positive role for HPWS practices for mitigating the negative effects of underemployment on job satisfaction. Fang et al. (2009) found a significant positive effect of job rotation on job satisfaction for immigrants with a university or professional degree, and a recent study by Chowhan et al. (2012a) found that the HPWS practice of pay-for-output significantly and substantially affects immigrants’ pay satisfaction more positively than Canadian-born employees. We adapt the Erdogan and Bauer (2011) framework to focus on the interaction between immigrant status and HPWS practices and job satisfaction. We argue that HPWS practices operate by ameliorating the negative effects associated with immigrant status, such as underemployment. For example, empowerment-enhancing practices may encourage acculturation (Aycan and Berry, 1996; Kats, 1983; Valdivia and Flores, 2012), motivation-enhancing practices may enable immigrants to feel rewarded or appreciated (Erdogan and Bauer, 2011) and skill-enhancing practices enable immigrants to keep their skills relevant and remain competitive for promotions (Erdogan and Bauer, 2011).

Thus, based on Erdogan and Bauer’s (2011) framework, the empirical findings above and the generally positive research findings on HPWS and job satisfaction for employees in general, we argue that HPWS practices experienced by immigrants will have a positive effect on their job satisfaction. This positive effect on job satisfaction will be seen when examining individual HPWS practices, sub-bundles of HPWS practices and a single overall bundle of HPWS practices experienced by immigrants. In other words, there will be a positive causal-moderation effect – HPWS practices will benefit immigrants relatively more than Canadian-born given immigrants’ relatively higher sensitivity to changes in HPWS practices (i.e. immigrant groups experiencing relatively lower levels of practices will be more sensitive to a change in the level of the practice than Canadian-born that are experiencing relatively higher levels) (Avery et al., 2010). As a result, we hypothesize that:

H3: The negative effect of immigrant status on job satisfaction will be positively moderated by the individual HPWS practices of job rotation, quality circle, task team, self-directed workgroup, employee survey, having a suggestion programme, informing employees about workplace changes, formal performance appraisal and whether it affects rewards, promotions, non-wage benefits, pay-for-performance, on-the-job training and classroom training.

H4: The negative effect of immigrant status on job satisfaction will be positively moderated by the empowerment-enhancing, motivation-enhancing, and skill-enhancing HPWS bundles.

H5: The negative effect of immigrant status on job satisfaction will be positively moderated by the overall experience of HPWS practices.

Method

Data

Our study uses Statistics Canada’s 1999–2004 Workplace and Employee Survey (WES). WES is a nationally representative sample of workplaces and their employees. The reference population is employees in Canada between the years 1999 and 2004. These data are the most current data available that enable the substantial breadth of investigation of HPWS-related variables for the Canadian-born and immigrant groups of interest; data of this nature and scale are rare in Canada and are not common internationally, and as such their value in testing theory cannot be under-appreciated. The unit of analysis in this article is the individual employee. Employees are randomly selected from workplaces that are included in the six-year longitudinal employer sample. The employee samples are refreshed every two years, so we observe the same employees in 1999 and 2000, in 2001 and 2002 and in 2003 and 2004. We pool these three separate employee samples into one dataset with effectively two time periods, where time 1 includes years 1999, 2001 and 2003 and time 2 includes years 2000, 2002 and 2004, which are the latest available linked datasets. To capture the lagged effect of factors affecting job satisfaction, we use a time 2 measure for job satisfaction matched with time 1 measures of employee and employer data. Pooled data are used because the survey content, coverage, geography and mode of collection of the survey do not change over time. Further, pooling is also appropriate when there are no extreme policy changes or external economic shocks that would make earlier employee experiences significantly different than the more recent responses, and fundamentally change the nature of the underlying relationships of interest (Thomas and Wannell, 2009).

The survey had average response rates of 87% for employers and 86% for employees, over the 1999–2004 period (for more on WES sampling and sample design, see Statistics Canada, 2007). The sample size of the panel used in the analyses is 49,344 (comprising 41,764 non-immigrants and 7580 immigrants), which is representative of an average working population of 10,150,000 employees over the reference periods (i.e. Canadian-born 8,323,000; total immigrant 1,827,000; with year of migration 1966 or earlier, 304,500; year of migration 1967–1989, 1,015,000; and year of migration 1990–2004, 507,500). The substantial sample size ensures any given cell in the analysis substantially surpasses the minimum rule of thumb of n > 30 which is sufficient for asymptotic theory (Cameron and Trivedi, 2006). The majority of the employees were: non-immigrant (82%), had full-time regular employment (76.4%), did not belong to a union (74%), married (69.5%), female (51.8%) and had a college or university degree (42.7%).

Statistics Canada made a considerable effort to ensure a high standard for data quality outcomes. With regard to non-sampling error and non-response in particular, both imputation and non-response weight adjustments were used, these steps were taken in addition to ensuring the high response rates reported above. For the employee survey the main imputation methods used were carry-over, distributional, weighted hot deck and nearest-neighbour, where the first two methods draw on the respondent’s own longitudinal and cross-sectional information and the second two methods use randomly selected donor information within stratum groupings, more detail on these methods is available in the survey documentation (Statistics Canada, 2007). With regard to the adjustments of the probability weights, Statistics Canada calibrated the final survey sample weights to known control totals for both the workplace and employee samples (using the Survey of Employment, Payroll and Hours [SEPH] for the employee benchmarks). Finally, estimates generated from the weighted data were rigorously scrutinized through multiple comparisons to other robust surveys, such as the Survey of Labour and Income Dynamics (SLID) and the Labour Force Survey (LFS), to ensure data quality.

Measures

Job satisfaction

The dependent variable is job satisfaction. Employees are asked ‘considering all aspects of this job, how satisfied are you with the job?’ and responses are coded as 4 = very satisfied, 3 = satisfied, 2 = dissatisfied and 1 = very dissatisfied. For this and other survey questions, see Statistics Canada (2004). Single-item measures of attitudinal constructs have limitations because it is either difficult or impossible to estimate the internal consistency or reliability of the construct; however, this was the only job satisfaction question in the survey (excluding a question on pay and benefit satisfaction which covers two of the nine facets of job satisfaction [Spector, 1997], and therefore not separately examined here). Some studies have demonstrated convergent validity between single-item and scale measures of job satisfaction indicating the acceptability of single-item measures (Wanous et al., 1997). Further, Rose (2005) presents employee survey results that an overall job satisfaction measure compared to a mean job-facet measure (mean of 14 job-facet items) captures the magnitude of job satisfaction outcomes and the magnitude of changes over time with a high level of similarity. However, job-facet item level differences do not always match the overall trend and as a result item level variation can be obfuscated by aggregation.

Immigrant and immigrant cohorts

The independent variables are immigrant and immigrant cohorts, and interaction variables are derived from immigrant and HPWS practice variables. The immigrant variable is based on the question ‘were you born in Canada?’ (coded 0 = yes, 1 = no), and the immigrant cohorts variables are created using, ‘if immigrant, in what year did you immigrate to Canada?’ To create the immigrant cohort variables we followed significant events in Canadian immigration policy and trends (Canada in the Making, 2011). Until 1967, preference was given to British, French, other European and American citizens in immigration, with Chinese allowed only to reunite with immediate family members. People from India, Pakistan and Sri Lanka were allowed on a quota basis. In 1967 the Points System was introduced to accept immigrants based on skills and education, and immigrants from Africa, the Caribbean, the Middle East and Asia started to enter Canada. Since the late 1980s the number of immigrants has increased and most have come from Asia, the Middle East and Africa (Canada in the Making, 2011; Statistics Canada, 2011). Using these time frames of immigration three immigrant cohort dummy variables were created: year of migration 1966 or earlier, year of migration 1967–1989 and year of migration 1990–2004. Recent immigrants refer to those with year of migration between 1990 and 2004. Canadian-born is the reference group.

As a proportion of the population the number of immigrants arriving as permanent residents (as opposed to temporary immigrants) has not varied dramatically over time remaining at under 1% of the total population. From 1940 to 1967, the average number of annual arrivals was 102,641 with an average proportion of the population at 0.65%. From 1967 to 1989, the annual arrivals averaged 142,419 with 0.62% as an average proportion of the population. Over the 1990–2004 period, the average arrivals increased in absolute terms and as a proportion of the population to 224,567 and 0.75%, respectively. Finally, for comparative purposes, the average number of arrivals (254,672) and the arrivals as a proportion of the population (0.74%) have not substantially changed over the 2005–2012 period when compared to the preceding period (Citizenship and Immigration Canada, 2012).

High performance work systems

HPWS human resource practices were included in the models as: (1) individual practices, (2) bundles and (3) an aggregate HPWS index (i.e. overall experience of HPWS practices). The inclusion of these three sets of measures permits an analysis of the effect of each individual practice on job satisfaction in addition to the sub-bundle indexes and the aggregate HPWS index. Although not the focus of this study, this makes our analysis similar to Mohr and Zoghi (2008), Wood and de Menezes (2011) and Zatzick and Iverson (2011), and permits a comparison of alternative configurational or universalist explanations, such as the synergistic properties and the magnitude of bundle effects on job satisfaction (Subramony, 2009). The aggregate HPWS index comprises all of the individual items. The practice of transforming all variables into z-scores, before combining them additively into a bundle or an aggregate HPWS index, was adopted from Appelbaum et al. (2000) and Macky and Boxall (2007). This approach enables us to account for the differing response formats of the variables used in the indexes. Following Delery (1998) and Macky and Boxall (2007), we make no assumptions about whether the items are measuring the same construct (i.e. whether the practices have high inter-correlations); rather, we are measuring the level of the bundle present when practices are experienced by employees.

The individual HPWS practices are presented here grouped by their sub-bundles category. The three sub-bundles follow Subramony’s (2009) particular composition, which is also consistent with other aggregations in the literature (Jiang et al., 2012; Wright and Boswell, 2002): empowerment-enhancing (job rotation, quality circle, task team, self-directed workgroup, employee survey, having a suggestion programme and informing employees about workplace changes), motivation-enhancing (formal performance appraisal, whether the appraisal affects rewards – such as promotions, benefits and pay – promotion in the last two years, promotion in last three or more years, non-wage benefits package and pay-for-performance) and skill-enhancing (on-the-job training and classroom training). For the empowerment-enhancing practices, following Mohr and Zoghi (2008), we code the three-point Likert format occasionally or frequently participate in these practices with a value of 1 and never participating as 0. For the motivation-enhancing measures, all the variables are coded 1 if the practice was reported by the employee and 0 otherwise, with the exception of non-wage benefits (which is an additive index indicating the number of benefits received including dental plan, employer contribution to the Registered Retirement Savings Plan [RRSP], employer contribution to stock, life-disability insurance, supplemental medical insurance plan, pension plan, group RRSP, stock purchase plan and supplements to Employment Insurance – for maternity/parental leave or layoffs). For the skill-enhancing measures only on-the-job and classroom training measures are included (with a value of 1 if training was received and 0 otherwise).

Control variables

Several variables are included as control variables because they are known to be associated with immigrants’ work experience (Aydemir and Skuterud, 2008; Galarneau and Morissette, 2008; Picot, 2008), immigrants and job satisfaction (Fang et al., 2009), job satisfaction and individual and workplace characteristics (Brown et al., 2008; Clark, 1997; Georgellis and Lange, 2007; Rose, 2005) and/or HPWS and job satisfaction (Berg, 1999; Macky and Boxall, 2007; Mohr and Zoghi, 2008; Wood and de Menezes, 2011; Zatzick and Iverson, 2011).

Employee variables include: six occupation group variables (managers, professionals, technical/trades, marketing/sales, clerical/administrative and production workers, where each binary variable equals 1 for the specific occupation group and 0 otherwise); log of hourly wage in dollars; part-time employment (= 1 if less than 30 hours per week), irregular employment (= 1 if casual on-call contract or term employment, and = 0 if regular standard employment); job tenure (= 1 if five years or less); work experience (a continuous variable, in years) and work experience sq/100 (i.e. experience squared divided by 100); post-secondary education (= 1 if college diploma, university degree or greater); qualification level; gender (= 1 if female); marital status (= 1 married); dependent child(ren) (= 1 if yes); and employment relationship covered by a collective agreement (= 1 if yes). The qualification level variable is measured by ranking the levels of education 0 to 9 (no minimum, elementary school, some secondary school, secondary school diploma, some post-secondary education, trade certificate, college diploma, university undergraduate degree, university professional accreditation and university graduate degree) and differencing the highest attained from the minimum required for the job.

The workplace control variables include: industries (resources, manufacturing, construction-utilities-transportation, education-health services, and other services, where each industry dummy = 1 if yes and = 0 otherwise), logarithm of workplace size measured using total employment at the workplace, and non-profit sector (1 = non-profit, 0 = for-profit).

Statistical analyses

Descriptive statistics and correlation analyses are conducted first. The hypotheses are tested using ordered logistic regression (OLR) analyses. This technique is appropriate given the categorical nature of the dependent variable, and is similar to the methodology used by Clark (1997) and Georgellis and Lange (2007). OLR is based on a latent regression structure, where a scale is observed as opposed to the continuous construct (in this case job satisfaction). The coefficients have a similar interpretation as other regression analyses, where the coefficient estimates represent the change in the dependent variable for a one unit change in the independent variables. The cut-points (i.e. ‘cut1’, ‘cut2’ and ‘cut3’) are similar to the constant term in standard regression analyses – they identify the thresholds for moving along the latent variable job satisfaction for the observed ordinal categories (for example, cut1 differentiates very dissatisfied from the other categories dissatisfied, satisfied and very satisfied). For all analyses, the employee level survey weight is used to estimate unbiased coefficients, and bootstrap weights are used to generate inference appropriate standard errors (Mann and Chowhan, 2011). Further, all weights were adjusted to account for the pooling of various years of data for differing employee populations (Thomas and Wannell, 2009). To account for heteroscedasticity a robust weight matrix is used for the regression analyses. Additional methodological details are available from the first author, upon request.

Three main sets of OLR analysis are conducted and presented in Models 1–7. All models include the employee and employer control variables discussed above. A baseline analysis was conducted with individual HPWS variables, bundles and the aggregate index and control variables. These analyses are not shown but are available from the first author upon request – the beta coefficients for the key variables of interest do not qualitatively differ from Models 1, 2, 3 and 4. Model 1 adds an overall measure of immigrant status with individual HPWS items included. Model 1 enables Hypothesis 1 to be tested. Models 2–4 look at the immigrant cohort effects (with Canadian-born as reference), and these models enable Hypothesis 2 to be tested. And finally, Models 5–7 look at the interaction effect of immigrant status and the individual HPWS practices (Model 5), HPWS bundles (Model 6) and the aggregate HPWS index (Model 7) on job satisfaction – testing the moderation Hypotheses 3, 4 and 5, respectively.

Results

With regard to the job satisfaction scores, the average job satisfaction score is 3.21, showing employees in Canada are satisfied with their jobs (see Table 1). Canadian-born have an average job satisfaction score of 3.23, which is 3% higher (p < .01) than the average score for immigrants (3.13). We consider this gap to be tangible and robust, although it is arguably not substantive. On the other hand, there are substantial differences across the immigrant entry cohorts. Immigrants entering Canada in 1966 or earlier have a higher average job satisfaction score (3.29) than Canadian-born and immigrant cohorts arriving later (1967–1989 and 1990–2004, 3.13 and 3.01 average scores respectively, statistically significant, p <. 01); and immigrant cohorts entering Canada in the 1967–1989 and 1990–2004 time periods have statistically significant (p < .01) lower job satisfaction than Canadian-born. We consider these gaps to be substantive as well as statistically significant – the magnitude of these proportional differences are similar to those found by Rose (2005), using the 1992–2002 British Household Panel Survey (BHPS), looking at gender differences in overall job satisfaction scores.

Table 1.

Means, standard deviations and correlations.^a

	Variable	Mean	SD	1	2	3	4	5	6	7	8	9
1	Job satisfaction	3.21	0.66
2	Canadian-born	0.82	–	0.06
3	Immigrant	0.18	–	−0.06	−1.00
4	Year of migration 1966 or earlier	0.03	–	0.02	−0.39	0.39
5	Year of migration 1967–1989	0.10	–	−0.04	−0.71	0.71	−0.06
6	Year of migration 1990–2004	0.05	–	−0.07	−0.48	0.48	−0.04	−0.07
7	Empowerment-enhancing bundle	0.00	1.00	0.06	−0.03	0.03	0.01	0.03	0.01
8	Motivation-enhancing bundle	0.00	1.00	0.07	−0.02	0.02	0.02	0.02	−0.01	0.39
9	Skill-enhancing bundle	0.00	1.00	0.06	0.03	−0.03	−0.01	−0.01	−0.02	0.29	0.26
10	HPWS aggregate index	0.00	1.00	0.09	−0.02	0.02	0.01	0.02	−0.01	0.87	0.76	0.50

Source: 1999–2004 Workplace and Employee Survey (WES), Statistics Canada.

N = 49,344.

The job satisfaction variable is a four-point scale, the immigrant status variables are binary and the index variables are continuous; correlations whose absolute values are greater than 0.02 are significant at p < .01.

Descriptive results are presented in Table 1, including means, standard deviations and correlations for the main variables of interest. As suggested by the comparison of means, more recent immigration (cohorts 1967–1989 and 1990–2004) is weakly negatively correlated with job satisfaction. Both Canadian-born and early arrival immigrants (entering in 1966 or earlier) are weakly positively correlated with job satisfaction. With regard to the HPWS bundles and the HPWS index, the means are equal to 0 and the standard deviations are equal 1; this is an artefact of the index calculation, which is discussed above. The HPWS bundles and the HPWS index are all weakly positively correlated with job satisfaction. Further, the bundles inter-correlations are below 0.70 indicating a lack of multi-collinearity, while the bundles are all moderately to strongly correlated with the aggregate HPWS index.

In Table 2, Model 1 shows full support for Hypothesis 1, that immigrants have lower job satisfaction than Canadian-born employees. In particular, immigrants are less likely to have the same odds for job satisfaction as Canadian-born employees all else being equal by a factor of 0.73 (odds ratio = OR = 0.73 = exp(−0.308) = exp(β), p < .01). In other words, holding everything else equal, immigrants have a job satisfaction odds ratio that is 27% ((1 – 0.73)/1) lower than Canadian-born employees. Although not shown in the table, the direction and magnitude of this result holds for the models with HPWS bundles and the HPWS aggregate index (these models are available from the first author upon request). Further, for the empowerment-enhancing HPWS variables, only task team is positively associated with job satisfaction (β = 0.126, p < .01) – this finding matches Mohr and Zoghi (2008) when they looked at the relationship between time 1 involvement practices and time 2 job satisfaction. For the motivation-enhancing HPWS practices, promotion in the last two years (β = 0.152, p < .01) and pay-for-performance (β = 0.089, p < .05) are positively associated with job satisfaction. For the skill-enhancing HPWS practices, only on-the-job training (β = 0.188, p < .01) is substantially and significantly positively associated with job satisfaction.

Table 2.

Effects of immigrant status and HPWS practices on job satisfaction (OLR analyses).

Variable	Model 1		Model 2		Model 3		Model 4
Variable	β	SE	β	SE	β	SE	β	SE
Immigrant (ref. Canadian-born)	−0.308***	(0.046)
Year of migration 1966 or earlier			0.0420	(0.098)	0.0336	(0.098)	0.0307	(0.099)
Year of migration 1967–1989			−0.351***	(0.065)	−0.350***	(0.064)	−0.352***	(0.064)
Year of migration 1990–2004			−0.474***	(0.081)	−0.491***	(0.083)	−0.497***	(0.083)
Employee survey	−0.0639	(0.043)	−0.0612	(0.042)
Suggestion programme	0.0226	(0.048)	0.0210	(0.048)
Informed by workplace	0.00790	(0.049)	−0.000785	(0.048)
Formal performance appraisal	0.00472	(0.044)	0.00746	(0.045)
Appraisal affects rewards	0.0578	(0.042)	0.0550	(0.042)
Promotion in the last 2 years	0.152***	(0.041)	0.147***	(0.042)
Promotion in last 3 or more years	0.0993*	(0.052)	0.0937*	(0.051)
Non-wage benefits package	0.0119	(0.0099)	0.0132	(0.0099)
Pay-for-performance	0.0890**	(0.044)	0.0841*	(0.044)
On-the-job training	0.188***	(0.037)	0.187***	(0.037)
Classroom training	0.0562	(0.040)	0.0556	(0.040)
Empowerment-enhancing bundle					0.0994***	(0.026)
Motivation-enhancing bundle					0.0759***	(0.021)
Skill-enhancing bundle					0.0742***	(0.019)
HPWS aggregate index							0.183***	(0.026)
Log (hourly wage)	0.441***	(0.059)	0.434***	(0.059)	0.409***	(0.059)	0.411***	(0.058)
Part-time employment	0.185***	(0.059)	0.183***	(0.058)	0.172***	(0.060)	0.172***	(0.060)
Irregular employment	0.116*	(0.065)	0.113*	(0.066)	0.121*	(0.065)	0.124*	(0.065)
Job tenure	−0.0388	(0.043)	−0.0335	(0.043)	−0.0279	(0.041)	−0.0209	(0.040)
Work experience	0.00961	(0.0062)	0.00849	(0.0062)	0.00793	(0.0063)	0.00770	(0.0062)
Work experience sq/100	0.00886	(0.015)	0.00856	(0.015)	0.00923	(0.015)	0.00947	(0.015)
Post-secondary education	−0.0511	(0.037)	−0.0434	(0.037)	−0.0444	(0.038)	−0.0403	(0.037)
Qualification level	−0.0190**	(0.0086)	−0.0183**	(0.0086)	−0.0176**	(0.0086)	−0.0180**	(0.0087)
Gender (female)	0.118***	(0.041)	0.116***	(0.041)	0.115***	(0.041)	0.116***	(0.042)
Marital status (married)	0.205***	(0.046)	0.210***	(0.046)	0.204***	(0.046)	0.204***	(0.046)
Dependent child(ren)	−0.0225	(0.042)	−0.0167	(0.042)	−0.0193	(0.042)	−0.0202	(0.041)
Collective agreement coverage	−0.0528	(0.045)	−0.0570	(0.045)	−0.0433	(0.042)	−0.0383	(0.042)
Log (workplace size)	−0.0720***	(0.015)	−0.0714***	(0.015)	−0.0887***	(0.015)	−0.0911***	(0.015)
Non-profit	−0.131*	(0.067)	−0.133**	(0.065)	−0.129*	(0.066)	−0.128*	(0.066)
Employee occupation	Yes		Yes		Yes		Yes
Workplace industry	Yes		Yes		Yes		Yes
Cut1	−2.298***	(0.19)	−2.333***	(0.18)	−2.770***	(0.20)	−2.767***	(0.20)
Cut2	−0.728***	(0.18)	−0.763***	(0.17)	−1.201***	(0.19)	−1.197***	(0.19)
Cut3	2.388***	(0.17)	2.357***	(0.17)	1.919***	(0.19)	1.921***	(0.19)

Pseudo R²	0.0239		0.0246		0.0244		0.0242
Wald χ²	504.46***		537.64***		482.52***		478.18***

Source: 1999–2004 Workplace and Employee Survey (WES), Statistics Canada.

significant at 10%; ** significant at 5%; *** significant at 1%.

Turning to the immigrant year of entry cohort effects, shown in Table 2 Models 2, 3 and 4, Hypothesis 2 is partially supported. Immigrants entering prior to 1966 are not significantly different from Canadian-born (β = 0.042, p >.10); however, cohorts 1967–1989 (β = −0.351, p < .01) and 1990–2004 (β = −0.474, p < .01) have increasingly lower job satisfaction levels, on average. The negative relationship ranges from approximately a factor of 0.70 (OR = 0.70 = exp(−0.351)) lower odds for 1967–1989 cohort to 0.62 (OR = 0.62 = exp(−0.474)) for 1990–2004 cohort, in Model 2. Also these findings hold once the moderation interaction of HPWS and immigrant status are added to the models (see Table 3).

Table 3.

Effects of individual HPWS practices, bundles of HPWS and the aggregate HPWS index on job satisfaction and immigrant interaction with HPWS practices and bundles (ordered logistic regression analyses).

Variable	Model 5		Model 6		Model 7
Variable	β	SE	β	SE	β	SE
Year of migration 1966 or earlier	−0.160	(0.15)	0.0198	(0.098)	0.0227	(0.099)
Year of migration 1967–1989	−0.552***	(0.16)	−0.361***	(0.065)	−0.360***	(0.065)
Year of migration 1990–2004	−0.664***	(0.14)	−0.493***	(0.085)	−0.497***	(0.084)
Job rotation * Immigrant	−0.0511	(0.12)
Quality circle * Immigrant	−0.0142	(0.13)
Task team * Immigrant	0.101	(0.14)
Self-directed workgroup * Immigrant	−0.176	(0.11)
Employee survey * Immigrant	0.145	(0.10)
Suggestion programme * Immigrant	0.0888	(0.13)
Informed by workplace * Immigrant	0.0255	(0.13)
Formal performance appraisal * Immigrant	0.0610	(0.13)
Appraisal affects rewards * Immigrant	0.0545	(0.11)
Promotion in the last 2 years * Immigrant	0.0949	(0.12)
Promotion in last 3 or more years * Immigrant	0.0472	(0.16)
Non-wage benefits package * Immigrant	0.0364	(0.022)
Pay-for-performance * Immigrant	−0.0487	(0.12)
On-the-job training * Immigrant	−0.227**	(0.098)
Classroom training * Immigrant	0.124	(0.091)
Empowerment-enhancing bundle * Immigrant			0.0808	(0.050)
Motivation-enhancing bundle * Immigrant			0.0968**	(0.048)
Skill-enhancing bundle * Immigrant			−0.0293	(0.039)
HPWS aggregate index * Immigrant					0.129***	(0.046)
Individual HPWS practices	Yes		No		No
HPWS bundles	No		Yes		No
HPWS aggregate index	No		No		Yes
Employee occupation	Yes		Yes		Yes
Workplace industry	Yes		Yes		Yes
Cut1	−2.363***	(0.18)	−2.772***	(0.19)	−2.771***	(0.19)
Cut2	−0.792***	(0.17)	−1.202***	(0.19)	−1.201***	(0.19)
Cut3	2.334***	(0.17)	1.920***	(0.19)	1.920***	(0.19)
Pseudo R²	0.0256		0.0248		0.0246
Wald χ²	578.29***		493.53***		489.20***

Source: 1999–2004 Workplace and Employee Survey (WES), Statistics Canada.

Note: The sample size of employees for each model is N = 49,344. The dependent variable is job satisfaction measured using a four-point Likert scale (4 = very satisfied, 3 = satisfied, 2 = dissatisfied and 1 = very dissatisfied). Yes and No indicate whether these variables are included in the models. For the rows in the table individual HPWS practices, HPWS bundles and HPWS aggregate index, a Yes indicates that this set of variables without the interaction is also included in the model but is omitted in the table for brevity. The individual HPWS practices include: job rotation, quality circle, task team, self-directed workgroup, employee survey, suggestion programme, informed by workplace, formal performance appraisal, appraisal affects rewards, promotion in the last two years, promotion in last three or more years, non-wage benefits package, pay-for-performance, on-the-job training and classroom training. The HPWS bundles include: empowerment-enhancing, motivation-enhancing and skill-enhancing bundles. Employee occupation variables include: managers, professionals, technical/trades, marketing/sales, clerical/administrative and production workers. The workplace industry group variables include: resources, manufacturing, construction-utilities-transportation, education-health services and other services. The model used for the estimation is an ordered logit. Bootstrap standard errors in parentheses.

significant at 10%; ** significant at 5%; *** significant at 1%.

The moderation effect of the individual HPWS practices on the immigrant status and job satisfaction relationship is tested in Table 3, Model 5. There is no support for Hypothesis 3 – that the individual practices from the empowerment-enhancing and motivation-enhancing HPWS bundles positively moderate immigrant status leading to higher job satisfaction. That is, immigrants and Canadian-born employees respond similarly to the use of these HPWS practices. Also, the on-the-job training and immigrant interaction has a negative relationship with job satisfaction (β = −0.227, p < .05), which indicates that immigrants who receive employer sponsored on-the-job training are likely to have lower job satisfaction than Canadian-born employees who receive on-the-job training. It is important to note that this negative relationship implies that the total effect of on-the-job training for immigrants is zero (i.e. on-the-job training unmoderated has a positive effect of 0.224 [included in the ‘HPWS items’ not shown in Table 3], and the moderated negative effect is −0.227, which implies that the total effect for immigrants is 0.224 + −0.227 = .003, which is not statistically significantly different from zero). For Model 6, Hypothesis 4 is partially supported; there are no moderation effects for empowerment-enhancing (β = 0.08, p >.10) and skill-enhancing (β = −0.029, p >.10) HPWS bundles on the immigrant status and job satisfaction relationship, but immigrants are significantly more likely to have higher job satisfaction than Canadian-born when they participate in a bundle of motivation-enhancing HPWS practices (β = 0.097, p < .05). Finally, Model 7 shows support for Hypothesis 5; thus, immigrant employees are more likely to have higher job satisfaction when they participate in a comprehensive bundle of HPWS practices (β = 0.129, p < .01) (i.e. higher HPWS aggregate index or a higher overall experience of HPWS practices). Although not shown in Table 3, the direction, magnitude and significance of the HPWS variables’ coefficients for Models 5, 6 and 7 are generally consistent with the results presented in Models 2, 3 and 4 (these models are available from the first author upon request).

Simple slope tests further confirm these results (see Figures 1 and 2). The slope of the relationship between the immigrant status and job satisfaction is steeper for immigrants (β = 0.03, p < .01) relative to Canadian-born (β = 0.01, p < .01) when moving from the low to high motivation-enhancing bundle conditions, where the low and high motivation-enhancing bundle outcomes are equivalent to a one standard deviation movement below and above the mean, respectively. Similarly, for the aggregate HPWS index, the slope is steeper for immigrants (β = 0.06, p < .01) relative to Canadian-born (β = 0.03, p < .01) when moving from the low to high HPWS outcome. The magnitude of these slopes is similar to the lower boundary found by Kooij et al. (2013) when they explored the moderation effects of age and HPWS bundles on job satisfaction.

Figure 1.

Interaction between HPWS bundle motivation and immigrants (and Canadian-born reference group) in relation to the dependent variable job satisfaction.

Figure 2.

Interaction between HPWS aggregate index and immigrants (and Canadian-born reference group) in relation to the dependent variable job satisfaction.

Discussion

Two major conclusions emerge from our study. First, compared to Canadian-born employees, immigrants’ job satisfaction is lower, especially for more recent arrivals. This finding is persistent even after controlling for various employee and workplace characteristics, including HPWS practices. Findings show that recent immigrants are less satisfied with their jobs than Canadian-born employees and immigrants who arrived in earlier years. Our findings add to the literature that shows recent immigrants have employment outcomes below Canadian-born employees, including difficulty in establishing their careers upon immigration to Canada (Avery et al., 2010; Salaff et al., 2002), difficulty in finding an adequate job (Schellenberg and Maheux, 2007), lower earnings than Canadian-born employees and experiencing a persistent earnings gap (Aydemir and Skuterud, 2008; Picot, 2008). Moreover, our findings are consistent with Fang et al.’s (2009) finding for immigrants who are professionals and adds to that study by showing immigrants that arrived in 1967 and after, whether professional or not, report lower job satisfaction than Canadian-born employees.

The differences across immigrant entry cohort groups suggest that there are differences between the groups that are contributing to more negative outcomes for more recent immigrants. For example, as discussed above, recent female and male immigrants with a university education were more likely to be in low-education jobs, 38% and 25%, respectively; and for non-recent immigrants both females and males did better with 24% and 18%, respectively, with a university education that were in low-education jobs (Galarneau and Morissette, 2004). A study by Ferrer et al. (2006) suggests that language and literacy difficulties are a substantial contributor to between-immigrant cohort differences (and Canadian-born differences) in earnings. It is likely that education received and job requirement education mismatching is contributing to some of these differences in earnings. In particular, this may be more salient for recent immigrants as Canadian education and experience helps improve language skills over time and contributes to earnings convergence and improved promotion opportunities (Chowhan et al., 2012b; Ferrer et al., 2006).

Second, our study shows that individual HPWS practices do not moderate the relationship between immigrant status and job satisfaction. The one exception is the negative interaction of on-the-job training and immigrant status. Georgellis and Lange (2007) found a negative relationship between any training and job satisfaction for males. Similar to Georgellis and Lange (2007), we suggest that training contributes to self-confidence and increases immigrant employees’ expectations and when rewards are not realized to the same degree for immigrants as they are for Canadian-born employees, lower job satisfaction occurs.

The positive effect of the motivation-enhancing bundle and immigrant status interaction on job satisfaction indicates the salience, for immigrant employees, of feeling valued by their workplace. When the HPWS practices of having formal performance appraisal, the appraisal affecting rewards, promotions, having a non-wage benefits package and pay-for-performance practices are formed together to create a motivation-enhancing HPWS bundle it positively affects immigrants’ job satisfaction. Thus, as proposed by Erdogan and Bauer (2011), recognition, appreciation and rewards can help to overcome feelings of relative deprivation caused by underemployment. Further, the additional importance of the motivation-enhancing HPWS bundle for immigrants, compared to Canadian-born employees, may be rooted in perceptions of equality, as suggested by Wood and de Menezes (2011). Similar to their discussion, we argue that if motivational practices can be applied in a procedurally consistent manner, this procedural justice can lead to increases in social cohesion and distributive justice, contributing to immigrants’ job satisfaction. Wages and other rewards are not only important in terms of economic equality between Canadian-born and immigrants, all else equal, but for immigrants, their previous standard of living and parity with comparators in their home country may also be relevant. Thus, a part of the immigrant experience is the struggle to maintain a comparable standard of living and preserve purchasing power parity (Mann, 2005). Finally, the positive interaction of the aggregate HPWS index with immigrant status indicates that immigrants have higher job satisfaction, relative to Canadian-born, when they are employed by workplaces that enable them to experience more HPWS practices that are a part of a comprehensive human resource management system. Thus, a comprehensive system that addresses empowerment, motivation and skills leading to an improvement in an employee’s situation suggests a positive role for a system of high performance work practices for mitigating the negative effects of immigrant status on job satisfaction. One of the main conclusions of this study is that increases in job satisfaction levels, for immigrants relative to Canadian-born, appear to be causally contingent on a comprehensive motivation-enhancing bundle and an overall HPWS index of practices – as opposed to a configurational approach that focuses on individual practices and that ignores the synergies that are realized from bundles and system approaches.

This study makes an important contribution to the literature. It examines workplace practices experienced by immigrants – immigrant status is a demographic characteristic that has received little attention in the literature (Bell et al., 2010: 177). Thus, this work contributes to identifying additional demographic characteristics of importance and helps to expand the notion of diversity (Jones et al., 2000: 378). Further, from a theoretical perspective this study identifies the importance of understanding the effects of HPWS practices on different groups of employees within organizations. In particular, we identify a causal association between HPWS practices interacting with immigrant status affecting job satisfaction. Finally, we identify individual practices and bundles of practices that contribute to reducing immigrants’ relative job satisfaction gap.

Limitations and future research directions

We would like to caution the readers of the following limiting conditions of the research. Different population sub-groups may systematically use different thresholds when assessing their satisfaction levels when responding to the same survey question (Clark, 1997; Lindeboom and van Doorslaer, 2004). Further, immigrants might have a different frame of reference than non-immigrants, which can affect job satisfaction results.

The differences across immigrant entry cohort groups and the differences between immigrants and Canadian-born employees, identified above, suggest that there are characteristics unique to these groups that lead to these differences. Future research can look at whether the frame of reference may vary across immigrant and Canadian-born groups and whether over time the frame of reference converges. The current study, by focusing on immigrant status, is taking a broad ‘first step’. Future research can explore entry cohort characteristics (such as immigrant category/class, ethnicity composition, country of origin, language skills and proficiency, occupational groupings, career profiles, human capital levels and social integration/networks, for example) to better understand which combinations of the elements of diversity may be contributing to the observed outcomes. Our data did not allow us to conduct analysis on these issues.

Practical implications

There are some practical implications for managers of our findings. Job satisfaction is important for retaining a skilled and motivated workforce, and it also has potentially important effects for maintaining and improving performance (Bowling, 2007; Judge et al., 2001). In the increasingly globalized workplace, where immigrants are viewed as a potential human resource to address labour shortages (Due Billing and Sundin, 2006), workplace diversity creates opportunities that can be used as a lever for improved business performance (Jayne and Dipboye, 2004; Mir et al., 2006). However, the diversity brought by immigrants to the workplace can challenge managers to create satisfactory jobs for their workforce within their workplace’s sub-culture; thus, job satisfaction of immigrants is an important issue to consider.

Our analysis has identified immigrant status as a key diversity characteristic contributing to gaps in job satisfaction levels between employees. That is, on average, immigrants have lower job satisfaction than Canadian-born employees, and job satisfaction levels are the lowest among recently arrived immigrants. These gaps should raise interest from all stakeholders, including practitioners. Moreover, to the extent that employers are having difficulty retaining and engaging good quality employees, it would be time well spent for managers to see if such a gap exists within their own organizations.

Given the importance of HPWS within most contemporary organizations, practitioners need to be sensitive to the possible effect of HPWS on immigrants’ job satisfaction. While individual HPWS practices and empowerment-enhancing and skill-enhancing HPWS bundles, in general, do not affect immigrants’ job satisfaction beyond the positive affect experienced by Canadian-born employees, when there are motivation-enhancing HPWS practices and when there is overall a higher level of HPWS practices experienced (i.e. comprehensive aggregate HPWS index) immigrants’ job satisfaction is higher. We suggest managers consider examining how HPWS practices can work together to positively affect immigrants’ job satisfaction and how their organizations can capitalize on this to improve immigrant employees’ job satisfaction, ultimately contributing to organizational success.

Footnotes

Acknowledgements

Access to data for this article was received through Statistics Canada Research Data Centre at McMaster University. The research and analysis are based on data produced by Statistics Canada and the opinions expressed do not represent the views of Statistics Canada.

Funding

This study was supported by a grant from the Social Sciences and Humanities Research Council of Canada.

Author biographies

James Chowhan is an Adjunct Assistant Professor with the DeGroote School of Business Management of Organizational Behaviour and Human Resources programme at McMaster University. James’s research focuses on the relationship between high performance work practices and both individual and organizational outcomes.

Isik U Zeytinoglu is Professor of Management and Industrial Relations at Human Resources and Management Area, DeGroote School of Business, McMaster University, Canada. Her research interests are flexible and non-standard employment and employer flexibility strategies.

Gordon B Cooke is Associate Professor of Industrial Relations at the Memorial University of Newfoundland. He examines the causes and effects of a range of non-standard work arrangements within Canada and beyond. His work has been published in journals such as Human Relations, Community, Work and Family and Relations Industrielles/Industrial Relations.

References

Appelbaum

Bailey

Berg

(2000) Manufacturing Advantage: Why High-Performance Work Systems Pay Off. Ithaca, NY: ILR Press.

Artz

(2010) Fringe benefits and job satisfaction. International Journal of Manpower 31: 626–644.

Avery

Tonidandel

Volpone

(2010) Overworked in America? How work hours, immigrant status, and interpersonal justice affect perceived work overload. Journal of Managerial Psychology 25: 133–147.

Aycan

Berry

(1996) Impact of employment-related experiences on immigrants’ psychological well-being and adaptation to Canada. Canadian Journal of Behavioural Science/Revue Canadienne des Sciences du Comportement 28: 240–251.

Aydemir

Skuterud

(2008) The immigrant wage differential within and across establishments. Industrial and Labor Relations Review 61: 334–352.

Bell

Kwesiga

Berry

(2010) Immigrants: The new ‘invisible men and women’ in diversity research. Journal of Managerial Psychology 25: 177–188.

Berg

(1999) The effects of high performance work practices on job satisfaction in the United States steel industry. Relations Industrielles/ Industrial Relations 54: 111–135.

Bowling

(2007) Is the job satisfaction–job performance relationship spurious? A meta-analytic examination. Journal of Vocational Behavior 71: 167–185.

Brown

Forde

Spencer

(2008) Changes in HRM and job satisfaction, 1998–2004: Evidence from the Workplace Employment Relations Survey. Human Resource Management Journal 18: 237–256.

10.

Brown

Hyatt

Benson

(2010) Consequences of the performance appraisal experience. Personnel Review 39: 375–396.

11.

Cameron

Trivedi

(2006) Microeconometrics: Methods and Applications. New York: Cambridge University Press.

12.

Campbell

Dunnette

Lawler

EE, I

(1970) Managerial Behavior, Performance and Effectiveness, Long Grove, IL: Waveland Press.

13.

Canada in the Making (2011) Specific events and topics, Immigration Acts (1866–2001). Available at: www.canadiana.ca/citm/specifique/immigration_e.html (accessed 15 July 2011).

14.

Chowhan

Zeytinoglu

Cooke

(2012a) Are immigrants’ pay and benefits satisfaction different than Canadian-born? Relations Industrielles/Industrial Relations 67: 3–24.

15.

Chowhan

Zeytinoglu

Cooke

(2012b) Skilled immigrants and promotion: Good and bad news from Canada. In: International Labour and Employment Relations Association (ILERA), 16th World Congress, Philadelphia, USA. Available at: /ilera2012.wharton.upenn.edu/RefereedPapers/ChowhanZeytinogluCooke.pdf.

16.

Citizenship and Immigration Canada (2012) Canada facts and figures: Immigration overview permanent and temporary residents, Research and Evaluation Branch of Citizenship and Immigration Canada, Document number Ci1–8/2012E-PDF. Available at: www.cic.gc.ca/english/pdf/research-stats/facts2012.pdf (accessed 14 April 2014).

17.

Clark

(1997) Job satisfaction and gender: Why are women so happy at work? Labour Economics 4: 341–372.

18.

Delery

(1998) Issues of fit in strategic human resource management: Implications for research. Human Resource Management Review 8: 289–309.

19.

Dietz

(2010) Introduction to the special issue on employment discrimination against immigrants. Journal of Managerial Psychology 25: 104–112.

20.

Due Billing

Sundin

(2006) From Managing Equity to Managing Diversity: A Critical Scandinavian Perspective on Gender and Workplace Diversity. London and Thousand Oaks, CA: Sage.

21.

Dundon

Gonzalez-Perez

M-A

McDonough

(2007) Bitten by the Celtic Tiger: Immigrant workers and industrial relations in the new ‘glocalized’ Ireland. Economic and Industrial Democracy 28: 501–522.

22.

Erdogan

Bauer

(2009) Perceived overqualification and its outcomes: The moderating role of empowerment. Journal of Applied Psychology 94: 557–565.

23.

Erdogan

Bauer

(2011) The impact of underemployment on turnover and career trajectories. In: Maynard

Feldman

(eds) Underemployment: Psychological, Economic, and Social Challenges. New York and Heidelberg: Springer, pp. 215–232.

24.

Erdogan

Bauer

Peiro

(2011) Overqualified employees: Making the best of a potentially bad situation for individuals and organizations. Industrial and Organizational Psychology – Perspectives on Science and Practice 4: 215–232.

25.

Fang

Zikic

Novicevic

(2009) Career success of immigrant professionals: Stock and flow of their career capital. International Journal of Manpower 30: 472–488.

26.

Ferrer

Green

Riddell

(2006) The effect of literacy on immigrant earnings. Journal of Human Resources 41: 380–410.

27.

Galarneau

Morissette

(2004) Immigrants: Settling for less? Perspectives on Labour and Income, Statistics Canada Catalogue No. 75−001-XIE 5: 5–16.

28.

Galarneau

Morissette

(2008) Immigrants’ education and required job skills. Perspectives on Labour and Income, Statistics Canada Catalogue No. 75−001-XIE 9: 5–16.

29.

Georgellis

Lange

(2007) Participation in continuous, on-the-job training and the impact on job satisfaction: Longitudinal evidence from the German labour market. International Journal of Human Resource Management 18: 969–985.

30.

Godard

(2004) A critical assessment of the high-performance paradigm. British Journal of Industrial Relations 42: 349–378.

31.

Hackman

Oldham

(1980) Work Redesign, Reading, MA: Addison-Wesley.

32.

Hakak

Holzinger

Zikic

(2010) Barriers and paths to success: Latin American MBAs’ views of employment in Canada. Journal of Managerial Psychology 25: 159–176.

33.

Heywood

Wei

(2006) Performance pay and job satisfaction. Journal of Industrial Relations 48: 523–540.

34.

Humphrey

Nahrgang

Morgeson

(2007) Integrating motivational, social, and contextual work design features: A meta-analytic of the summary and theoretical extension work design literature. Journal of Applied Psychology 92: 1332–1356.

35.

Jayne

MEA

Dipboye

(2004) Leveraging diversity to improve business performance: Research findings and recommendations for organizations. Human Resource Management 43: 409–424.

36.

Jiang

Lepak

(2012) How does human resource management influence organizational outcomes? A meta-analytic investigation of mediating mechanisms. Academy of Management Journal 55: 1264–1294.

37.

Jones

Pringle

Shepherd

(2000) ‘Managing diversity’ meets Aotearoa/New Zealand. Personnel Review 29: 364–380.

38.

Jones

Latreille

(2009) Training, job satisfaction, and workplace performance in Britain: Evidence from WERS 2004. Labour 23: 139–175.

39.

Judge

Thoresen

Bono

(2001) The job satisfaction–job performance relationship: A qualitative and quantitative review. Psychological Bulletin 127: 376–407.

40.

Kats

(1983) Occupational mobility of immigrants and their job satisfaction: A secondary analysis (I). International Migration 21: 345–357.

41.

Knocke

(2000) Integration or segregation? Immigrant populations facing the labour market in Sweden. Economic and Industrial Democracy 21: 361–380.

42.

Kooij

DTAM

Guest

Clinton

(2013) How the impact of HR practices on employee well-being and performance changes with age. Human Resource Management Journal 23: 18–35.

43.

Lawler

III (1973) Motivation in Work Organizations. Monterey, CA: Brooks/Cole.

44.

Lawler

III (2000) Rewarding Excellence: Pay Strategies for the New Economy. San Francisco: John Wiley and Sons.

45.

Lawler

III (2005) Motivating and Satisfying Excellent Individuals. San Francisco, CA: John Wiley and Sons.

46.

Lindeboom

van Doorslaer

(2004) Cut-point shift and index shift in self-reported health. Journal of Health Economics 23: 1083–1099.

47.

Macky

Boxall

(2007) The relationship between ‘high-performance work practices’ and employee attitudes: An investigation of additive and interaction effects. International Journal of Human Resource Management 18: 537–567.

48.

Mann

(2005) What’s a penny worth? Wages, prices, and the American working man. Ethnography 6: 315–355.

49.

Mann

Chowhan

(2011) Weighting large datasets with complex sampling designs: Choosing the appropriate variance estimation method. Journal of Modern Applied Statistical Methods 10: 110–115.

50.

Mir

Wong

(2006) Diversity: The Cultural Logic of Global Capital? London and Thousand Oaks, CA: Sage.

51.

Mohr

Zoghi

(2008) High-involvement work design and job satisfaction. Industrial and Labor Relations Review 61: 275–296.

52.

OECD (2008) International Migration Outlook SOPEMI - 2008 Edition. Paris: OECD. Available at: www.oecd.org/document/3/0,3343,en_2649_33931_41241219_1_1_1_1,00.html (accessed 1 August 2011).

53.

Picot

(2008) Immigrant economic and social outcomes in Canada: Research and data development at Statistics Canada, Canada Catalogue No. 11F0019M No. 319. Available at: www.statcan.ca/ (accessed 20 June 2011).

54.

Rose

(2005) Job satisfaction in Britain: Coping with complexity. British Journal of Industrial Relations 43: 455–467.

55.

Salaff

Greve

Ping

LXL

(2002) Paths into the economy: Structural barriers and the job hunt for skilled PRC migrants in Canada. International Journal of Human Resource Management 13: 450–464.

56.

Schellenberg

Maheux

(2007) Immigrants’ perspectives on their first four years in Canada: Highlights from three waves of the longitudinal survey of immigrants to Canada. Statistics Canada, Catalogue No. 11−008. Available at: www.statcan.gc.ca/pub/11−008-x/2007000/pdf/9627-eng.pdf (accessed 2 December 2013).

57.

Spector

(1997) Job Satisfaction: Application, Assessment, Cause and Consequences. London: Sage.

58.

Statistics Canada (2004) 2004 Workplace and Employee Survey: Questionnaire, employee questionnaire, Business and Labour Market Analysis Division and Labour Statistics Division. 4–4700–2.1: 2004−06−07 STC/LAB-075–75055. Available at: www.statcan.gc.ca/imdb-bmdi/instrument/2615_Q1_V6-eng.pdf (accessed 19 August 2013).

59.

Statistics Canada (2007) Guide to the analysis of Workplace and Employee Survey, 2004. Statistics Canada, Catalogue No. 71–221-GIE. Available at: unstats.un.org/unsd/EconStatKB/KnowledgebaseArticle10273.aspx?Keywords=Data+collection (accessed 15 July 2013).

60.

Statistics Canada (2011) Immigration and ethnocultural diversity in Canada. Statistics Canada, Catalogue No. 99−010-X2011001. Available at: www12.statcan.gc.ca/nhs-enm/2011/as-sa/99−010-x/99−010-x2011001-eng.pdf (accessed 14 April 2014).

61.

Subramony

(2009) A meta-analytic investigation of the relationship between HRM bundles and firm performance. Human Resource Management 48: 745–768.

62.

Tett

Meyer

(1993) Job satisfaction, organizational commitment, turnover intention, and turnover: Path analysis based on meta-analytic findings. Personnel Psychology 46: 259–293.

63.

Thomas

Wannell

(2009) Combining cycles of the Canadian Community Health Survey. e No. 82−003-X 20: 1–6.

64.

Valdivia

Flores

(2012) Factors affecting the job satisfaction of Latino/a immigrants in the Midwest. Journal of Career Development 39: 31–49.

65.

Wanous

Reichers

Hudy

(1997) Overall job satisfaction: How good are single-item measures? Journal of Applied Psychology 82: 247–252.

66.

Wood

de Menezes

(2011) High involvement management, high-performance work systems and well-being. International Journal of Human Resource Management 22: 1586–1610.

67.

Wright

Boswell

(2002) Desegregating HRM: A review and synthesis of micro and macro human resource management research. Journal of Management 28: 247–276.

68.

Zatzick

Iverson

(2011) Putting employee involvement in context: A cross-level model examining job satisfaction and absenteeism in high-involvement work systems. The International Journal of Human Resource Management 22: 3462–3476.