Does Demographic Dissimilarity Matter for Perceived Inclusion? Evidence From Public Sector Employees

Abstract

This study examines the relationship between individual dissimilarity and perceptions of organizational inclusion. Data from a national survey of public agencies conducted in Florida and Texas show that gender dissimilarity is negatively associated with perceptions of inclusion and the negative relationship is more acute for men than for women. In contrast, tenure dissimilarity is positively related to perceptions of inclusion and this positive association is more acute for those with longer tenure than for those with shorter tenure. These results suggest that the effect of dissimilarity on the perception of inclusion depends on both the observability of individual-level characteristics and the status of the demographic group. In particular, dissimilarity along characteristics that are easily observable (such as gender) is more likely to influence perceptions of inclusion and dissimilarity is more influential for higher status groups (such as men or long-tenured employees).

Keywords

demographic dissimilarity gender dissimilarity tenure dissimilarity organizational inclusion hierarchical linear modeling (HLM)

Introduction

Past studies on diversity in the public sector have generally focused on issues such as discrimination, affirmative action, reverse discrimination, and valuing and managing diversity (Kellough & Naff, 2004; Naff & Kellough, 2003; Pitts, 2006, 2009; Riccucci, 2002; Wise & Tschirhart, 2000). This body of research has revealed several important issues related to employee diversity and policies to address changing demographics in the workplace. Concurrently, a recent stream of research in psychology and business examines ways in which diverse populations can be integrated in an organization (Mor Barak, 2000; Mor Barak & Cherin, 1998; Pless & Maak, 2004; Roberson, 2006; Shore et al., 2011). The concept of inclusion—defined as “the degree to which individuals feel part of critical organizational processes” (Mor Barak & Cherin, 1998, p. 48)—is thus gaining popularity among business communities. In recent years, public management literature has also explored the concept of inclusion examining the relationship between diversity or representativeness and inclusion (Andrews & Ashworth, 2014; Selden, 2006; Shore et al., 2011).

Previous organizational diversity literature has explored the effect of group demographic composition on organizational performance, job satisfaction, or turnover intention (Choi, 2009, 2013; Choi & Rainey, 2010; Pitts, 2006, 2009). A drawback of this organizational approach, however, is that it assumes that all individuals of the same group will behave the same way regardless of the social context (Cunningham, 2007). In response to this argument, studies have recognized that attention to the impact of diversity at the individual level is also important. O’Reilly, Caldwell, and Barnett (1989) defined individual dissimilarity as individual-level demographic differences with others. Individuals use demographic characteristics, such as sex and age, to categorize others as in-group or out-group, with in-group members viewed more favorably than out-group (Tsui & Gutek, 1999). Moreover, existing studies, focused primarily on private sector firms, have found that individual dissimilarity is negatively related to organizational inclusion (Pelled, Ledford, & Mohrman, 1999; Tsui, Egan, & O’Reilly, 1992).

To the best of our knowledge, this study is the first to analyze the association between individual dissimilarities and organizational inclusion in the public sector. Although Andrews and Ashworth (2014) examined the relationship between representativeness and inclusion using civil service organizations within the United Kingdom, they do not consider individual-level demographic characteristics. Our study extends previous research by examining the relationship between multiple individual dissimilarity variables and employee inclusion in public organizations. Our study builds on this work by creating four individual-level measures of dissimilarity focused on gender, age, tenure, and education level. We use these measures to examine how differences in demographic characteristics influence the perception of inclusion at the individual level while accounting for differences in agency type and size. We use a hierarchical linear modeling (HLM) approach to examine both individual- and organizational-level factors. Next, we present the theoretical foundations that informed our study—self-categorization theory and organizational inclusion—and the hypotheses derived from existing literature. We then describe our data and analytic method followed by a discussion of our findings and concluding remarks.

Theoretical Background

Self-Categorization Theory

According to self-categorization theory, individuals recognize themselves through their group membership (Hogg & Terry, 2000). People categorize themselves and others into social groups as in-group or out-group based on characteristics such as gender, age, or organizational membership (Tajfel, 1982). People prefer homogeneous groups (Messick & Mackie, 1989; Schneider, 1987) and regard members of an out-group less favorably than members of an in-group (Kramer, 1991). In addition, belonging to one’s in-group results in a high level of self-esteem and a positive self-identity (Kramer, 1991). In contrast, interacting with those who are considered out-group may result in anxiety. As Stephan and Stephan (1985) suggested, “people who regard themselves as superior experience anxiety concerning interaction with others who are regarded as inferior” (p. 163).

Furthermore, the perception of similarity or dissimilarity can affect the attitude of an individual in a psychological group. For example, Hoffman and Hurst (1990) found that individuals have the highest satisfaction in an organization comprised of members with the same gender. This indicates that individuals prefer homogeneity over heterogeneity in a group, and job satisfaction increases in a homogeneous group. Similarly, if individuals categorize themselves by age, job satisfaction is the highest when people work with a similar-age group (McCain, O’Reilly, & Pfeffer, 1983). Ultimately, individual preferences for interacting with homogeneous groups and more favorable evaluations of in-group members conflict with increasing diversity in the workplace.

In addition to examining how dissimilarity affects individuals, existing research has also found individual’s demographic dissimilarity to be negatively associated with organizational outcomes such as turnover, communication, and social integration. Here, dissimilarity can be defined as “the extent of which an employee’s demographic profile differs from the demographic profiles of others in his/her work unit” (O’Reilly et al., 1989, p. 21). Jackson et al. (1991) revealed that if top managers were dissimilar to their teammates in terms of age, education level, or industry experience, they were more likely to leave. Other studies suggest that individual age dissimilarity has a negative relationship with communication (Zenger & Lawrence, 1989) and social integration (O’Reilly et al., 1989).

Organizational Inclusion

Many studies on diversity have focused on the effects of gender or racial diversity on organizational performance (Choi & Rainey, 2010; Frink et al., 2003) and job satisfaction (Choi, 2013; Pitts, 2009). However, recent studies have begun to take an interest in the degree to which women and non-White workers have been accepted and treated as insiders by others in an organization (Andrews & Ashworth, 2014; Selden, 2006). The effect of such inclusion on organizational performance has also been recently examined in the public sector (Sabharwal, 2014). Inclusion has been conceptualized as the extent to which people in an organization feel engaged in the practices most crucial to the organization (Mor Barak & Cherin, 1998). More specifically, inclusion can be described as “access to information, connectedness to co-workers, and the ability to participate in and influence the decision making process” (Mor Barak, 2000, p. 341).

Pelled et al. (1999) examined the relationship between individual dissimilarity and three inclusion indicators—decision-making influence, which is the influence that an employee has over decisions that affect the work that he or she does; access to sensitive work information, indicating the degree to which an employee is kept well-informed about the company objectives and plans; and job security, which is the likelihood that an employee will retain his or her job. Similar to Tsui et al. (1992), Pelled et al. (1999) found gender dissimilarity to be negatively associated with inclusion, but tenure and education-level dissimilarity to be positively associated. Because individuals tend to categorize themselves by noticeable physical characteristics (Fiske & Taylor, 1991), it follows that age and gender dissimilarity are more likely to have a negative association with the perception of inclusion. Dissimilarity in education level or tenure, however, plays a different role. Education level and tenure are less visible than gender and age in organizations. Education level and tenure also have a task-related relationship with an organization, whereas gender and age may be considered less relevant to the objectives of an organization (Zenger & Lawrence, 1989).

Based on self-categorization theory and empirical research on organizational inclusion, we predict that individual dissimilarity in gender and age will have a negative relationship with organizational inclusion (i.e., decision-making influence, information access, and job security), whereas education level and tenure dissimilarity will have a positive association with inclusion.

Hypothesis 1a: Individual gender dissimilarity will be negatively associated with perceptions of organizational inclusion.

Hypothesis 1b: Individual age dissimilarity will be negatively associated with perceptions of organizational inclusion.

Hypothesis 1c: Individual education-level dissimilarity will be positively associated with perceptions of organizational inclusion.

Hypothesis 1d: Individual tenure dissimilarity will be positively associated with perceptions of organizational inclusion.

Although demographic characteristics influence the perception of inclusion, these effects influence demographic groups in different ways. Previous studies on diversity have examined the response of individuals in groups such as women, employees with a low-education level, and short-tenure groups in the presence of the majority groups (Chatman & Spataro, 2005; Pelled et al., 1999; Stephan & Stephan, 1985; Tsui et al., 1992; Zenger & Lawrence, 1989). Traditionally, men have occupied positions of power and have dominated the workplace (Acker, 2012; Eagly & Carli, 2007). Consequently, men feel anxious and make less investment in organizations where they are a minority (Chattopadhyay, 1999; Chattopadhyay, Finn, & Ashkanasy, 2010; Chattopadhyay, George, & Lawrence, 2004; Heilman, 1994; Konrad & Gutek, 1987; Pelled et al., 1999). In such circumstances, men hold negative feelings about the workplace, which leads to reduced organizational inclusion—decision-making influence, information access, and job security (O’Farrell & Harlan, 1982; Wharton & Baron, 1987).

O’Farrell and Harlan (1982) found that men in predominantly male jobs treated female employees with hostility, while Schreiber (1979) reported that women in predominantly female jobs did not treat male employees with hostility. Previous studies reported that men in mixed or balanced gender workplaces show lower job satisfaction and self-esteem when compared with men in a male-dominated workplace (Wharton & Baron, 1987). Other studies suggested that women in a gender balanced workplace and in a predominantly male workplace have higher job satisfaction than women in a female-dominated workplace (Wharton & Baron, 1991). Andrews and Ashworth (2014) also found that the association between gender representation and perceived inclusiveness varies by gender. Accordingly, we hypothesize that the effect of gender dissimilarity on perceived organizational inclusion is different for men and women.

Hypothesis 2a: The negative relationship between gender dissimilarity and perceived organizational inclusion will be greater for males than for females.

As education level and tenure are task-related and connected to the objectives of the workplace (Zenger & Lawrence, 1989), a different set of predictions is warranted for education level and tenure dissimilarity. Because employees with more education and experience are apt to possess relevant knowledge and information, education-level dissimilarity and tenure dissimilarity may enhance decision-making capacity. Individuals with longer tenure and a higher education level may have greater access to information and more job security because they receive information through diverse sources. Thus, both education and tenure dissimilarity lead to enhanced access to information. Furthermore, dissimilar employees’ unique knowledge and experience make them requisite workers in the workplace. Previous empirical studies have supported the argument that a positive relationship between education-level dissimilarity and inclusion is pronounced in organizations with a higher education-level group (Pelled et al., 1999; Tsui et al., 1992). They also found the positive association between tenure dissimilarity and inclusion to be pronounced for those with a longer tenure group. Based on these observations, we hypothesize that the effects of education-level dissimilarity and tenure dissimilarity on inclusion will be different by education level and tenure.

Hypothesis 2b: The positive relationship between education-level dissimilarity and perceived organizational inclusion will be greater for the high-education group than for the low-education group.

Hypothesis 2c: The positive relationship between tenure dissimilarity and perceived organizational inclusion will be greater for the long-tenure group than for the short-tenure group.

In addition to individual-level factors that influence the perception of inclusion, organizational-level factors such as agency type and agency size matter, and hence are included as controls in our study. The tenets of social role theory would suggest that females are interested in issues related to women, children, and caretaking (Sabharwal, 2015). Following this assumption, large numbers of women would then be attracted to work in agencies such as the Department of Education, Department of Veteran Affairs, and Department of Housing and Development. According to Lowi (1985), such agencies are considered redistributive agencies or agencies whose “rules or the rules for which they are responsible affect society on a larger scale than any others” (p. 93). A large number of women work in redistributive agencies because jobs in these agencies require “soft skills” and emotional labor (Guy & Newman, 2004; Meier, Mastracci, & Wilson, 2006; Ryan & Haslam, 2007; Stivers, 1993). Distributive, regulatory, and constituent agencies on the contrary are male dominated and are more likely to experience glass ceilings and gender and pay disparities (Dolan, 2004; Kelly & Newman, 2001; Kerr, Miller, & Reid, 2002; Newman, 1994; Reid, Miller, & Kerr, 2004; Sneed, 2007). Kelly and Newman (2001) found that distributive agencies comprise a lower percentage of mid- and upper-level female employees than male employees. Furthermore, regulatory agencies have an equal percentage of mid- and upper-level female and male employees. Dolan (2004) found that slightly larger proportions of male employees work for constituent agencies.

Data and Method

Sample

Data for this study are taken from a national survey conducted on issues of human resources in Florida and Texas in the summer 2011. The survey participants are employed at various state agencies in health and welfare, environment, transportation, personnel, and education. This is part of a larger study that included Washington, Utah, and Oregon. However, the latter states were not included in this study as they did not choose to identify their departments. This was crucial to our study as we include organizational-level data that classifies departments into four categories based on Lowi’s (1985) classification that includes distributive, redistributive, regulatory, and constituent agencies. Agency type information was only available for Texas and Florida; these two states had a combined N of 455 (196 for Texas and 259 for Florida). The response rate for Florida was 32% and for Texas 24.4%. Respondents consist of managerial and non-managerial positions that include low-, mid-, and senior-level employees (non-managerial); supervisors; and lower managers.

Data were collected using Qualtrics, a software for online surveys. In line with the Institutional Review Board (IRB) requirements, respondents were informed about the voluntary nature of the study and assured complete anonymity. In Florida, permission to conduct the survey was sought from senior agency directors while an open records request for emails was filed in Texas.¹

The average number of years respondents worked in government was 20.8 years. The sample was split by gender with females constituting 50.3% of the responses. The majority of the respondents (80%) were 45 years and older. A large majority of the sample holds a Bachelor’s degree or beyond (87%). Furthermore, the majority of the respondents were in supervisory positions (87%). Half of the respondents worked in distributive type of agencies, while one fourth were employed at redistributive type agencies and the remaining one fourth at regulatory (21%) and constituent agencies (5%).

Measures

Dependent variables

The following indicators have previously been used by scholars to measure organizational inclusion: decision-making influence, information access, and job security (Ledford & Mohrman, 1993; Pelled et al., 1999). Decision-making influence is measured by an average of four questionnaire items (α = .7622) pertaining to influence on decisions related to improving productivity and the work environment. Responses to each question are coded on a Likert-type scale ranging from 1 to 7 with 1 being strongly agree and 7 being strongly disagree. The four questions are as follows: (a) We empower (ask) employees to make important decisions, (b) We frequently develop innovative programs, (c) We frequently re-engineer or re-design our work processes, and (d) Staff are required to pursue continuing professional development.

Information access is measured as the average of five questionnaire items (α = .6078) pertaining to clarity in work assignments and tools for communication and improvement. Responses to each question are coded on a Likert-type scale ranging from 1 to 7 with 1 being strongly agree and 7 being strongly disagree. The five questions are as follows: (a) I achieve job goals, targets, and deadlines set in my job; (b) My work responsibilities are clear and specific; (c) We use advanced information technology applications; (d) We use performance measurement in our program management; and (e) We regularly use strategic planning.

Last, this study uses two questionnaire items (α = .5419) to measure job security. Responses to each question are coded on a Likert-type scale ranging from 1 to 7 with 1 being strongly agree and 7 being strongly disagree. The two questions related to job security are as follows: (a) My job security is satisfactory and (b) I expect to be working here for many years. Appendix explains questionnaire items for scales with more than two items.

Independent variables

Following the procedures used by Jackson et al. (1991), Pelled et al. (1999), and Wagner, Pfeffer, and O’Reilly (1984), this study used the Euclidean distance to measure individual dissimilarity from the group. The Individual dissimilarity score is calculated as follows where n is the number of group members, S_i is the ith individual’s value on the attribute, and S_j is the jth member’s value on the attribute. (For gender, the value for $S_{i} - S_{j}$ was 0 if the two unit members were of the same gender and 1 if they were of different gender.)

D_{i} = {[\sum_{j = 1}^{n} \frac{{(S_{i} - S_{j})}^{2}}{(n - 1)}]}^{1 / 2} .

The individual dissimilarity score represents the distance between an individual and all others in the agency who responded to the survey with higher scores indicating more dissimilarity. We calculate measures for gender dissimilarity, age dissimilarity, and education-level dissimilarity to examine the effects of individual dissimilarity on organizational inclusion.

Control variables

Following existing work, this study contains several demographic control variables to distinguish dissimilarity effects and demographic effects (Pelled et al., 1999; Tsui et al., 1992; Tsui & O’Reilly, 1989). At the individual level, we include gender, age, education level, supervisor, and tenure as control variables. Female is a dichotomous variable recorded as “0” for male respondents and “1” for female respondents. Age is an ordinal variable, which categorizes a respondents’ age into under 35, 35 to 44, 45 to 54, or more than 54. Education level is an ordinal variable which categorizes a respondent’s highest degree into associates or less, bachelor’s degree, master’s degree, or PhD/JD. Supervisor is a dichotomous variable recorded as “1” for supervisor respondents and “0” for other respondents. Tenure is measured by the number of years for which a respondent has worked in government instead of years worked in the agency or other sectors.

At the organizational level, we control for organization size and agency type. Organization size is measured by the number of employees in each organization. To measure agency type, we use Lowi’s (1985) typology, classifying agencies into four types: distributive, redistributive, regulatory, and constituent. This classification has been widely used by scholars who study the gendered nature of institutions (Guy & Newman, 2004; Kelly & Newman, 2001; Kerr et al., 2002; Naff, 2001; Riccucci, 2009; Sabharwal, 2015; Saidel & Loscocco, 2005; Sneed, 2007). Distributive agencies used in this study are the Department of Transportation and the Department of State Health Services. Redistributive agencies are the Department of Children and Families, the Department of Education, the Department of Elder Affairs, and the Texas Education Agency. Regulatory agencies are the Environment Protection Agency and the Texas Commission on Environmental Quality. Constituent agencies include the Governor’s Office, State Auditor’s Office, Department of Management Services, and the Texas Work Commission. We define redistributive agency type as the reference group and is not included in the regression model.

Analytic Method

We used hierarchical linear modeling (HLM) to examine the relationship between individual dissimilarity and organizational inclusion. HLM is appropriate in analyzing the effect of individual dissimilarity because we assume that organizational inclusion is associated with both individual- and organizational-level factors. A multilevel model is appropriate with data on individuals (Level 1) collected through surveys that are clustered within organizations (Level 2). HLM reduces problems associated with the underestimation of standard errors for parameter estimates and heterogeneity (Steenbergen & Jones, 2002). Accordingly, this study examines whether individual- and organizational-level variables are associated with our three measures of organizational inclusion—decision-making influence, information access, and job security. The HLM equation is as follows:

\begin{array}{l} Level 1 : Y_{i j} = β_{0 j} + β_{1 j} E m p l o y e e^{’} s p o s i t i o n + β_{2 j} F e m a l e + β_{3 j} A g e + β_{4 j} D e g r e e \\ + β_{5 j} T e n u r e + β_{6 j} G e n d e r d i s s i m i l a r i t y + β_{7 j} A g e d i s s i m i l a r i t y \\ + β_{8 j} E d u c a t i o n l e v e l d i s s i m a r i t y + β_{9 j} T e n u r e d i s s i m i l a r i t y + r_{i j} . \end{array}

\begin{array}{l} Level 2 : β_{0 j} = r_{00} + r_{01} O r g a n i z a t i o n s i z e + r_{02} C o n s t i t u e n t + r_{03} D i s t r i b u t i v e \\ + r_{04} R e g u l a t o r y + u_{a j} . \end{array}

Results

Table 1 reports the descripitive statistics for the sample. Table 2 presents the bivariate correlations among the variables included in our model. The largest correlation is .66, and the mean and median correlation magnitudes (absolute value) are .12 and .06. The results of the variance inflation factor (VIF) test indicate that all VIFs are below 5 suggesting no serious concern for multicollinearity.

Table 1.

Descriptive Statistics.

	M	SD	Minimum	Maximum	Unit
1. Decision-making influence	3.18	1.22	1	7	Factor score
2. Information access	2.48	0.79	1	6	Factor score
3. Job security	3.04	1.10	1	6	Factor score
4. Organization size	87.39	46.89	1	146	Number of employees
5. Constituent	0.05	0.21	0	1	Constituent = 1, others = 0
6. Distributive	0.50	0.50	0	1	Distribution = 1, others = 0
7. Regulatory	0.21	0.41	0	1	Regulatory = 1, others = 0
8. Supervisor	0.87	0.34	0	1	Supervisor =1, other = 0
9. Female	0.50	0.50	0	1	Female = 1, male = 0
10. Age	3.19	0.81	1	4	Under 35 = 1, 35-44 = 2, 45-54 = 3, more than 54 = 4
11. Education level	2.41	0.84	1	4	Associates or less = 1, bachelor’s degree = 2, master’s degree = 3, PhD/JD= 4
12. Tenure	20.82	9.44	1	45	Years in government
13. Gender dissimilarity	0.43	0.23	0	1.27	Index
14. Age dissimilarity	0.85	0.44	0	3.21	Index
15. Education-level dissimilarity	0.88	0.46	0	2.69	Index
16. Tenure dissimilarity	20.82	9.44	1	45	Index

Table 2.

Correlation of Variables.

	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17
1. Decision making	1
2. Information access	.73	1
3. Job security	.34	.29	1
4. Organization size	.16	.15	−.10	1
5. Constituent	−.03	−.05	−.07	−.36	1
6. Distributive	−.05	−.04	−.18	.62	−.22	1
7. Regulatory	.15	.11	.11	.05	−.11	−.51	1
8. Redistributive	−.07	−.03	.14	−.60	−.13	−.57	−.29	1
9. Supervisor	.01	−.05	.08	.20	−.14	−.01	.14	−.05	1
10. Female	.04	.03	−.05	.12	−.04	.06	−.18	.12	−.04	1
11. Age	.02	.01	.03	.01	.03	−.01	−.05	.04	.02	−.02	1
12. Education level	.16	.06	−.04	.10	−.01	.03	−.05	.01	−.01	−.02	−.06	1
13. Tenure	−.07	−.14	.10	−.03	−.03	−.04	.12	−.04	.11	−.05	.35	−.20	1
14. Gender dissimilarity	.03	.06	.02	.17	−.12	.06	.15	−.14	.08	.14	.03	−.10	.08	1
15. Age dissimilarity	.10	.09	.05	.05	−.02	−.11	.27	−.12	.05	−.05	.01	−.05	.08	.58	1
16. Education dissimilarity	.10	.09	.03	.24	−.04	.12	.16	−.27	.04	.01	.05	.13	.07	.60	.60	1
17. Tenure dissimilarity	.06	.01	.10	.15	−.04	.08	.05	−.12	.13	.05	.08	.01	.07	.5	.67	1	1

Note. All correlations above the absolute value .05 (bold-faced value) are significant at p < .05 for a two-tailed test.

The HLM used to test Hypotheses 1a and 1b is summarized in Table 3. Hypothesis 1a states that individual gender dissimilarity will be negatively associated with our three measures of organizational inclusion—decision-making influence, information access, and job security. Model 1 shows that gender dissimilarity has a significant negative relationship with decision-making influence (β = −.785, p < .1). This result partly supports Hypothesis 1a: the greater the individual’s gender distance from others in the agency, the lower the perception of organizational inclusion. The results neither support Hypothesis 1b, which expected a negative association between age dissimilarity and organizational inclusion, nor did the study find support for Hypothesis 1c, which expected a positive association between education-level dissimilarity and organizational inclusion. However, we found a significant positive relationship between tenure dissimilarity and job security in Model 3 (β = .042, p < .1), supporting Hypothesis 1d. Model 3 shows that constituent and distributive agencies have significant negative relationships with job security (constituent: β = −0.638, p < .1; distributive: β = −.743, p < .05). These results suggest that employees in redistributive agencies, which are typically female dominated, tend to have lower job security compared with employees in male-dominated agencies such as constituent and distributive agencies.

Table 3.

HLM Analysis as Perception of Organizational Inclusion.

	Decision-making influence	Information access	Job security
	Model 1	Model 2	Model 3
Organizational-level factors
Organization size	0.005 (0.003)	0.004 (0.003)	0.001 (0.003)
Constituent	0.187 (0.362)	−0.012 (0.268)	−0.638* (0.384)
Distributive	−0.233 (0.295)	−0.281 (0.233)	−0.743** (0.311)
Regulatory	0.326 (0.325)	0.056 (0.269)	−0.167 (0.347)
Individual-level factors
Supervisor	−0.110 (0.192)	−0.186 (0.128)	0.054 (0.192)
Female	0.087 (0.127)	−0.033 (0.084)	−0.002 (0.127)
Age	0.123 (0.085)	0.052 (0.056)	−0.034 (0.082)
Education level	0.181** (0.076)	0.007 (0.050)	−0.052 (0.075)
Tenure	−0.012 (0.007)	−0.009* (0.005)	0.012 (0.007)
Individual dissimilarity variables
Gender dissimilarity	−0.785* (0.437)	−0.169 (0.291)	−0.618 (0.453)
Age dissimilarity	0.280 (0.259)	−0.007 (0.174)	0.069 (0.249)
Education-level dissimilarity	0.262 (0.240)	0.237 (0.158)	0.157 (0.238)
Tenure dissimilarity	0.001 (0.023)	−0.013 (0.015)	0.042* (0.234)
Constant	2.639*** (0.352)	2.447*** (0.246)	2.904*** (0.355)
Number of observations	406	408	336
Number of groups	13	13	11
Wald χ²	22.06**	13.79*	20.97*
Pseudo R²	.08	.19	.10

Note. The coefficient and standard error (in parentheses) are reported. Pseudo R² was calculated based on Rabe-Hesketh and Skrondal (2008, p. 103). HLM = hierarchical linear modeling.

p < .1. **p < .05. ***p < .01.

Our results are consistent with the findings of Tsui et al. (1992) and Pelled et al. (1999) who argued that gender dissimilarity is negatively related to inclusion, whereas tenure dissimilarity has a positive association with inclusion. In agreement with this work, our results imply that differences in gender among individuals in an organization reduce perceived organizational inclusion (in case of decision making). Because gender is easily noticeable, dissimilarity among members yields a negative association with inclusion.

This study also conducts sub-sample analyses to examine the diverse effects of individual dissimilarity on organizational inclusion. A sub-group analysis is used to analyze the different effects of dissimilarity by group because the correlation between the interaction terms and dissimilarity variables is too high.² The high-education group consists of respondents with a master’s degree or higher whereas the low-education group is composed of individuals with a bachelor’s degree or lower. For tenure, we categorized the sub-sample into long-tenure and short-tenure based on median tenure period.

Table 4 presents the results of the sub-sample analyses in this study. Consistent with Hypothesis 2a, gender dissimilarity maintains a significant negative association with job security (β = −2.587, p < .01) in the male sub-sample but had a non-significant association with job security (β = −0.684, ns) in the female sub-sample. Education dissimilarity does not show a significant positive association with decision-making influence, information access, or job security in the high-education sub-sample. The results do not support Hypothesis 2b. In the long-tenure sub-sample, tenure dissimilarity has a significant positive relationship with decision-making influence (β = .093, p < .1), which supports Hypothesis 2c.

Table 4.

Sub-Sample Analysis.

	Decision-making influence	Information access	Job security
Gender dissimilarity effects
Male sub-sample	β_G1 = −1.124	β_G1 = −.455	β_G1 = −2.587***
Female sub-sample	β_G2 = .052	β_G2 = .187	β_G2 = −.684
Education-level dissimilarity effects
High-education sub-sample	β_E2 = .429	β_E2 = .561	β_E2 = .012
Low-education sub-sample	β_E1 = .405	β_E1 = .008	β_E1 = −.096
Tenure dissimilarity effects
Long-tenure sub-sample	β_T2 = .093*	β_T2 = .007	β_T2 = −.017
Short-tenure sub-sample	β_T1 = .013	β_T2 = −.008	β_T2 = .016

Note. The coefficient and standard error (in parentheses) are reported. The reported p values are based on two-tailed tests.

p < .1. **p < .05. ***p < .01.

Our sub-sample analyses show that a negative association between gender dissimilarity and job security is stronger in the male group sample than in the female group sample. This result is in line with that of Pelled et al. (1999). We found that being different in gender has more negative effects on organizational inclusion for males than females. Furthermore, this study supports the argument that the positive association between tenure dissimilarity and inclusion (capacity to influence decision making) is more pronounced for those with longer tenures.

Conclusion

The perspectives on the relationship between individual demographic dissimilarity and organizational inclusion fall into two camps. According to the first perspective, being different from others within the organization is negatively associated with organizational inclusion because individuals feel uncomfortable and face barriers to effective communication when they interact with others of different demographic backgrounds. The second perspective argues that difference is positively associated with organizational inclusion because diverse backgrounds bring more perspectives and create knowledge among individuals.

Consistent with the first perspective, our results show that gender dissimilarity is negatively related to perceptions of organizational inclusion. This is also consistent with self-categorization theory, which argues that people categorize themselves into in- and out-groups based on characteristics such as gender and age (Hogg & Terry, 2000; Tajfel, 1982). According to this theory, employees are most satisfied when they are with members belonging to similar demographic composition (Messick & Mackie, 1989; Schneider, 1987). Although gender dissimilarity reduces perceptions of inclusive behaviors, we found the results to hold true only for decision-making capacity. This suggests that in gender dissimilar groups, employees feel less empowered in making decisions. Inclusion is an umbrella term used for various factors that can cause employees to feel disconnected from an organization. Thus, it is important that future research parse out the concept for enhancing our understanding of inclusion in the workplace.

We also find that dissimilarity can lead to positive perceptions about inclusion. Consistent with the second perspective, our results show that tenure dissimilarity is positively associated with parts of inclusion. These results imply that gender dissimilarity acts as a disadvantage for organizational inclusion, whereas tenure dissimilarity is advantageous to inclusion. Because gender is characteristic easily noticed, gender dissimilarity among members yields a negative association with inclusion. By contrast, tenure is high in task-relevance. Co-workers tend to value the opinions of those with a longer tenure in the organization. Accordingly, tenure dissimilarity (those who have been employed for a longer time) has a positive relationship with the decision-making element of inclusion.

Our sub-sample analyses suggest that individual gender dissimilarity is negatively associated with perceived organizational inclusion and that this negative effect is more acute for men than for women. We also find that individual tenure dissimilarity is positively associated with perceived organizational inclusion and that this positive relationship is more acute for those with longer tenure than for those with shorter tenure. Dissimilarity is more likely to be influential for more observable characteristics (such as gender). The effect of dissimilarity is also more pronounced for the perception of organizational inclusion at the individual level for demographic groups that are typically in the majority or hold higher status (such as men). The effects of dissimilarity in the less observable characteristics (such as tenure) are more acute for high-status groups (such as long-tenured employees).

Although our results are consistent with studies done in private sector, our findings raise additional insights for public sector organizations. In particular, serving an increasingly diverse workforce and citizenry will require organizations that are not only descriptively representative but are also inclusive in their practices within the organization. Even if organizations are proactively making efforts to be inclusive, these efforts can be perceived differently across demographic groups. Although this study focused on gender, education, and tenure, future studies can use other demographic variables such as racial categories, sexual orientation, gender identity, veteran’s status, disability status, political orientation, religious affiliation, and so on to examine the effects of dissimilarity on organizational inclusion.

Given the move from diversity to inclusion, this study has some important implications. Through our analyses, we find that men feel less secure in an environment where there is greater gender dissimilarity. Although the results might not be particularly shocking, the results do raise questions about how gender roles continue to affect places of work. We are not suggesting that organizations segregate by clustering groups of people by gender to create an inclusive environment. In fact, this is quite contrary to what we are suggesting. Training is one way to create an awareness of structural policies and procedures that might not lead to discriminatory practices (Avery, McKay, & Wilson, 2008). If not already providing gender/diversity trainings, organizations should offer these to create a heightened awareness among its members about issues of equity and providing information about how decisions are made (i.e., in a fair and just manner).

Beyond structural changes, organizations need to address the climate for diversity and inclusion. Conversations about gender diversity and inclusion are not easy, especially in environments where males continue to dominate high-status positions. Similar to Foldy and Buckley’s (2014) approach to addressing issues of race at work, we argue that the first step is to recognize and be aware of the climate of the organization (especially in- and out-groups). Second, differences across groups need to be addressed, and third, a safe space has to be created for members of the organizations to talk about dissimilarities and the perceptions about feeling included or excluded. This approach needs to go beyond a structural shift to acknowledging dissimilarities and providing avenues for starting a conversation about issues that concern individuals and work groups.

However, the results of the study need to be interpreted with caution, as the sample is limited to public sector employees in states of Florida and Texas. We would urge future research to expand data that would include additional states as well as allow for cross-agency comparisons. Our data are also limited wherein we use the number of years for which a respondent has worked in government as a measure of tenure rather than the years worked in the agency. Furthermore, we use agency-level data and not individual work unit due to sample size and anonymity concerns. Future research can examine dissimilarity within work groups. As our data are limited by employees in the government sector, we do not know whether there is a bias against work experience in other sectors or deference to those with more work experience in general. Furthermore, as there is a significant relationship between tenure dissimilarity and supervisory status, future studies can compare dissimilarity effects for supervisors versus non-supervisors to try and gain additional insights. Studies can also focus on examining supervisor–supervisee relationship. This will add to the understanding of trust and power relationship between in- and out-groups. In addition, future studies can expand the analysis by including the supervisor similarity effects or variables such as men with shorter or longer tenure and women with shorter or longer tenure.

The study also utilizes cross-sectional data, which is typical of studies in organizational settings. However, future researchers can use data longitudinal in nature to both verify the findings of this study and capture change overtime. Although we use the definition of inclusion employed by Pelled et al. (1999), future research can use other variables such as empowerment, equity, justice, collaborative work arrangements, and conflict resolution processes to measure organizational inclusion.

Footnotes

Appendix

Pelled, Ledford, and Mohrman (1999)	Questionnaires	Cronbach’s α
Decision-making influence		α = .76
Influence over decision about ways to improve quality of work environment	We empower (ask) employees to make important decisions
	We frequently develop innovative programs
Influence over decisions about ways to improve quality of product or service	We frequently re-engineer or re-design our work processes
	Staff are required to pursue continuing professional development
Information access		α = .61
Employee is well-informed about plant’s goals	I achieve job goals, targets, and deadlines set in my job
Employee is well-informed about plant’s goals	My work responsibilities are clear and specific
Employee is well-informed about new technologies	We use advanced information technology applications
Employee is well-informed about business plans	We use performance measurement in our program management
Employee is well-informed about business plans	We regularly use strategic planning
Job security		α = .54
Employee is unlikely to be laid off	I expect to be working here for many years
Employee has high degree of job security	My job security is satisfactory

Acknowledgements

The authors acknowledge Jonathan West, Professor, University of Miami for collecting the Florida data that are part of the ROPPA 33(2) symposium on “Human Resource Management in the Asia-Pacific Region.”

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Kwang Bin Bae is an assistant professor in the Department of Public Administration at North Carolina Central University. His primary interests are human resource management, performance management, and nonprofit finance. His articles have appeared in the American Review of Public Administration, Public Personnel Management, Local Government Studies, KEDI Journal of Education Policy, and International Review of Public Administration.

Meghna Sabharwal is an associate professor in the public and nonprofit management program at the University of Texas at Dallas. Her research interests are focused on public human resources management. She is widely published in public administration journals. She also has two book publications: Public Personnel Administration (6th Ed.) and Public Administration in South Asia: India, Bangladesh, and Pakistan.

Amy E. Smith is an associate professor in the McCormack Graduate School of Policy and Global Studies at the University of Massachusetts Boston. Her current research focuses on gender diversity in leadership in public organizations, career paths in public service, and issues of work-life balance. She also conducts research on teaching and mentoring in graduate education.

Evan Berman is professor of Public Management and Director of Internationalisation at Victoria University of Wellington (New Zealand). He is recipient of the 2016 Award for Outstanding Scholarship in HRM from SPALR for contributions throughout a career.

References

Acker

(2012). Gendered organizations and intersectionality: Problems and possibilities. Equality, Diversity and Inclusion: An International Journal, 31, 214-224.

Andrews

Ashworth

(2014). Representation and inclusion in public organizations: An empirical evaluation of the UK civil service. Public Administration Review, 75, 279-288.

Avery

D. R.

McKay

P. F.

Wilson

D. C.

(2008). What are the odds? How demographic similarity affects the prevalence of perceived employment discrimination. Journal of Applied Psychology, 93(2), 235.

Chatman

J. A.

Spataro

S. E.

(2005). Using self-categorization theory to understand relational demography-based variations in people’s responsiveness to organizational culture. Academy of Management Journal, 48, 321-331.

Chattopadhyay

(1999). Beyond direct and symmetrical effects: The influence of demographic dissimilarity on organizational citizenship behavior. Academy of Management Journal, 42, 273-287.

Chattopadhyay

Finn

Ashkanasy

N. M.

(2010). Affective responses to professional dissimilarity: A matter of status. Academy of Management Journal, 53, 808-826.

Chattopadhyay

George

Lawrence

S. A.

(2004). Why does dissimilarity matter? Exploring self-categorization, self-enhancement, and uncertainty reduction. Journal of Applied Psychology, 89, 892-900.

Choi

(2009). Diversity in the US federal government: Diversity management and employee turnover in federal agencies. Journal of Public Administration Research and Theory, 19, 603-630.

Choi

(2013). Demographic diversity of managers and employee job satisfaction empirical analysis of the federal case. Review of Public Personnel Administration, 33, 275-298.

10.

Choi

Rainey

H. G.

(2010). Managing diversity in U.S. federal agencies: Effects of diversity and diversity management on employee perceptions of organizational performance. Public Administration Review, 70, 109-121.

11.

Cunningham

G. B.

(2007). Perceptions as reality: The influence of actual and perceived demographic dissimilarity. Journal of Business and Psychology, 22, 79-89.

12.

Dolan

(2004). Gender equity: Illusion or reality for women in the federal executive service? Public Administration Review, 64, 299-308.

13.

Eagly

A. H.

Carli

L. L.

(2007, September). Women and the labyrinth of leadership. Harvard Business Review, 85(9), 62-71.

14.

Fiske

S. T.

Taylor

S. E.

(1991). Social cognition (2nd ed.). New York, NY: McGraw-Hill.

15.

Foldy

E. G.

Buckley

T. R.

(2014). The color bind: Talking (and not talking) about race at work. New York, NY: Russell Sage.

16.

Frink

D. D.

Robinson

R. K.

Reithel

Arthur

M. M.

Ammeter

A. P.

Ferris

G. R.

. . . Morrisette

H. S.

(2003). Gender demography and organization performance: A two-study investigation with convergence. Group & Organization Management, 28, 127-147.

17.

Guy

M. E.

Newman

M. A.

(2004). Women’s jobs, men’s jobs: Sex segregation and emotional labor. Public Administration Review, 64, 289-298.

18.

Heilman

M. E.

(1994). Affirmative action: Some unintended consequences for working women. Research in Organizational Behavior, 16, 125-169.

19.

Hoffman

Hurst

(1990). Gender stereotypes: Perception or rationalization? Journal of Personality and Social Psychology, 58, 197-208.

20.

Hogg

M. A.

Terry

D. I.

(2000). Social identity and self-categorization processes in organizational contexts. Academy of Management Review, 25, 121-140.

21.

Jackson

S. E.

Brett

J. F.

Sessa

V. I.

Cooper

D. M.

Julin

J. A.

Peyronnin

(1991). Some differences make a difference: Individual dissimilarity and group heterogeneity as correlates of recruitment, promotions, and turnover. Journal of Applied Psychology, 76, 675-689.

22.

Kellough

J. E.

Naff

K. C.

(2004). Responding to a wake-up call: An examination of federal agency diversity management programs. Administration & Society, 36, 62-90.

23.

Kelly

R. M.

Newman

(2001). The gendered bureaucracy: Agency mission, equality of opportunity, and representative bureaucracies. Women & Politics, 22(3), 1-33.

24.

Kerr

Miller

Reid

(2002). Sex-based occupational segregation in US state bureaucracies, 1987-97. Public Administration Review, 62, 412-423.

25.

Konrad

A. M.

Gutek

B. A.

(1987). Theory and research on group composition: Applications to the status of women and ethnic minorities. In Oskamp Spacapan

(Eds.), Interpersonal processes (pp. 85-121). Newbury Park, CA: Sage

26.

Kramer

R. M.

(1991). Intergroup relations and organizational dilemmas: The role of categorization processes. Research in Organizational Behavior, 13, 191-228.

27.

Ledford

G. E.

Mohrman

S. A.

(1993). Self-design for high involvement: A large-scale organizational change. Human Relations, 46, 143-173.

28.

Lowi

T. J.

(1985). The state in politics: The relation between policy and administration. In Noll

R. G.

(Ed.), Regulatory policy and the social sciences (pp. 67-105). Berkeley: University of California Press.

29.

McCain

B. E.

O’Reilly

C. A.

III Pfeffer

(1983). The Effects of departmental demography on turnover. Academy of Management Journal, 26, 626-641.

30.

Meier

K. J.

Mastracci

S. H.

Wilson

(2006). Gender and emotional labor in public organizations: An empirical examination of the link to performance. Public Administration Review, 66, 899-909.

31.

Messick

D. M.

Mackie

D. M.

(1989). Intergroup relations. Annual Review of Psychology, 40, 45-81.

32.

Mor Barak

M. E

. (2000). The inclusive workplace: An ecosystems approach to diversity management. Social Work, 45, 339-353.

33.

Mor Barak

M. E.

Cherin

D. A

. (1998). A tool to expand organizational understanding of workforce diversity: Exploring a measure of inclusion-exclusion. Administration in Social Work, 22(1), 47-64.

34.

Naff

K. C.

(2001). To look like America: Dismantling barriers for women and minorities in government. Boulder, CO: Westview Press.

35.

Naff

K. C.

Kellough

J. E.

(2003). Ensuring employment equity: Are federal diversity programs making a difference? International Journal of Public Administration, 26, 1307-1336.

36.

Newman

M. A.

(1994). Women in management: Problem or opportunity? Public Productivity & Management Review, 17, 405-410.

37.

O’Farrell

Harlan

S. L.

(1982). Craft workers and clerks: The effect of male co-worker hostility on women’s satisfaction with non-traditional jobs. Social Problems, 29, 252-265.

38.

O’Reilly

C. A.

III Caldwell

D. F.

Barnett

W. P.

(1989). Work group demography, social integration, and turnover. Administrative Science Quarterly, 34(1), 21-37.

39.

Pelled

L. H.

Ledford

G. E.

Mohrman

S. A.

(1999). Demographic dissimilarity and workplace inclusion. Journal of Management Studies, 36, 1013-1031.

40.

Pitts

(2006). Modeling the impact of diversity management. Review of Public Personnel Administration, 26, 245-268.

41.

Pitts

(2009). Diversity management, job satisfaction, and performance: Evidence from US federal agencies. Public Administration Review, 69, 328-338.

42.

Pless

Maak

(2004). Building an inclusive diversity culture: Principles, processes and practice. Journal of Business Ethics, 54, 129-147.

43.

Rabe-Hesketh

Skrondal

(2008). Multilevel and longitudinal modeling using Stata. College Station, TX: Stata Press.

44.

Reid

Miller

Kerr

(2004). Sex-based glass ceilings in US state-level bureaucracies, 1987-1997. Administration & Society, 36, 377-405.

45.

Riccucci

N. M.

(2002). Managing diversity in public sector workforces. Boulder, CO: Westview Press.

46.

Riccucci

N. M.

(2009). The pursuit of social equity in the federal government: A road less traveled? Public Administration Review, 69, 373-382.

47.

Roberson

Q. M.

(2006). Disentangling the meanings of diversity and inclusion in organizations. Group & Organization Management, 31, 212-236.

48.

Ryan

M. K.

Haslam

S. A.

(2007). The glass cliff: Exploring the dynamics surrounding the appointment of women to precarious leadership positions. Academy of Management Review, 32, 549-572.

49.

Sabharwal

(2014). Moving beyond diversity management: Organizational inclusion to further performance. Public Personnel Management, 43, 197-217.

50.

Sabharwal

(2015). From glass ceiling to glass cliff: Women in senior executive service. Journal of Public Administration Research and Theory, 25, 399-426.

51.

Saidel

J. R.

Loscocco

(2005). Agency leaders, gendered institutions, and representative bureaucracy. Public Administration Review, 65, 158-170.

52.

Schneider

(1987). The people make the place. Personnel Psychology, 40, 437-453.

53.

Schreiber

C. T.

(1979). Changing places: Men and women in transitional occupations. Cambridge, MA: The MIT Press.

54.

Selden

S. C.

(2006). A solution in search of a problem? Discrimination, affirmative action, and the new public service. Public Administration Review, 66, 911-923.

55.

Shore

L. M.

Randel

A. E.

Chung

B. G.

Dean

M. A.

Ehrhart

K. H.

Singh

(2011). Inclusion and diversity in work groups: A review and model for future research. Journal of Management, 37, 1262-1289.

56.

Sneed

B. G.

(2007). Glass walls in state bureaucracies: Examining the difference departmental function can make. Public Administration Review, 67, 880-891.

57.

Steenbergen

M. R.

Jones

B. S.

(2002). Modeling multilevel data structures. American Journal of Political Science, 46, 218-237.

58.

Stephan

W. G.

Stephan

C. W.

(1985). Intergroup anxiety. Journal of Social Issues, 41(3), 157-175.

59.

Stivers

(1993). Rationality and romanticism in constitutional argument. Public Administration Review, 53, 254-257.

60.

Tajfel

(1982). Social identity and intergroup relations. Cambridge, UK: Cambridge University Press.

61.

Tsui

A. S.

Egan

O’Reilly

C. A.

III . (1992). Being different: Relational demography and organizational attachment. Administrative Science Quarterly, 37, 549-579.

62.

Tsui

A. S.

Gutek

B. A.

(1999). Demographic differences in organizations: Current research and future directions. Lanham, MD: Lexington Books.

63.

Tsui

A. S.

O’Reilly

C. A.

III . (1989). Beyond simple demographic effects: The importance of relational demography in superior-subordinate dyads. Academy of Management Journal, 32, 402-423.

64.

Wagner

W. G.

Pfeffer

O’Reilly

C. A.

III . (1984). Organizational demography and turnover in top-management group. Administrative Science Quarterly, 29, 74-92.

65.

Wharton

A. S.

Baron

J. N.

(1987). So happy together? The impact of gender segregation on men at work. American Sociological Review, 52, 574-587.

66.

Wharton

A. S.

Baron

J. N.

(1991). Satisfaction? The Sociological Quarterly, 32, 365-387.

67.

Wise

L. R.

Tschirhart

(2000). Examining empirical evidence on diversity effects: How useful is diversity research for public-sector managers? Public Administration Review, 60, 386-394.

68.

Zenger

T. R.

Lawrence

B. S.

(1989). Organizational demography: The differential effects of age and tenure distributions on technical communication. Academy of Management Journal, 32, 353-376.