Pluralistic Ignorance and the Flexibility Bias: Understanding and Mitigating Flextime and Flexplace Bias at Work

Abstract

Workers who request flexibility are routinely stigmatized. The authors experimentally tested and confirmed the hypothesis that individuals believe others view flexworkers less positively than they do. This suggests flexibility bias stems, in part, from pluralistic ignorance. The authors also found that flexplace requesters were stigmatized significantly more than flextime requesters. Given this finding, they recommend research distinguish between different types of flexwork. In a second study, they assessed whether exposure to information suggesting organizational leaders engage in flexible work reduced bias. They found that when the majority of high-status employees work flexibly, bias against flextime (but not flexplace) workers was attenuated.

Keywords

flextime flexplace flexibility stigma pluralistic ignorance

Flexible work is an umbrella term used to describe any one of a spectrum of work arrangements that alters the time or place that work gets done on a regular basis. For example, flextime entails adjusting the beginning and ending of the workday such that an employee works the same number of hours but at less traditional times, and telecommuting (or flexplace) allows workers to work a portion of their normally scheduled hours from a remote location.

Given extremely high levels of work–family conflict in the United States (Gornick & Meyers, 2005), many workers report they prefer flexible work (or flexwork) to more traditional workplace structures (Finn & Donovan, 2013). Scholars, too, have advocated for more flexible work policies (e.g., Correll, Kelly, O’Connor, & Williams, 2014; Singley & Hynes, 2005; Williams, 2010) and many companies have responded, rolling out flexwork programs (Galinsky, Bond, Sakai, Kim, & Giuntoli, 2008). Yet, few employees actually take advantage of flexible work opportunities (Blair-Loy & Wharton, 2002; Blair-Loy, Wharton, & Goodstein, 2011; Galinsky et al., 2008). This may, in part, be due to a flexibility bias in which those who seek flexible work are discredited, devalued, and stigmatized (Blair-Loy & Wharton, 2004; Cohen & Single, 2001; Glass, 2004; Wharton, Chivers, & Blair-Loy, 2008; Williams, Blair-Loy, & Berdahl, 2013). Workers, aware of this bias, may engage in bias avoidance by forgoing flexwork opportunities (Blair-Loy & Wharton, 2002; Drago, Black, & Wooden, 2005; Jacobs & Gerson, 2004).

We consider one mechanism through which discrimination against flexworkers might be perpetuated. We consider whether situations in which workers request accommodations for flexible work are contexts of pluralistic ignorance. Pluralistic ignorance entails public support but private rejection—or public rejection but private support—of a perceived norm thought to be endorsed by others (Katz & Allport, 1931). Disapproval for those who take advantage of flexible work, we argue, may be a perceived norm that workers assume most others support even though they themselves, privately, do not. In situations of pluralistic ignorance, some people actively enforce norms they mistakenly believe to be widely held to avoid public censure, despite privately rejecting those norms (Willer, Kuwabara, & Macy, 2009). Thus, we assert that pluralistic ignorance may lead individuals to stigmatize workers who take advantage of flexible work because they believe the majority of their peers view flexible work pejoratively. We test this assertion with an experimental study that compares personal biases with the biases individuals assume others have. We find evidence in support of the pluralistic ignorance hypothesis suggesting that flexibility bias may be fostered by individuals’ erroneous perception that others feel more negatively about flexible work than they do.

If flexibility bias is perpetuated by a misperception of the social norm, then altering perceptions of the norm could reduce discrimination against flexworkers. In a second experimental study, we test the hypothesis that the use of flexibility by leadership will reduce flexibility bias. We find that when individuals are told that more than half of the company’s senior managers engage in flexible work, bias against workers who request flextime is reduced.

An additional goal of this research is to examine differences in perceptions of various types of flexwork. Flexible work comes in a variety of forms. For example, workers may compress their workweek by working fewer days but longer hours per day or job share by splitting the responsibilities of a position and working only part of the week. We consider the two most common types of flexible work: flextime and flexplace (WorldatWork, 2013). Previous flexibility bias research often fails to examine differences in the consequences of different types of flexible work, yet we found significant differences by accommodation.

A final goal of this research is to identify how much support is needed among high-level managers to affect subjects’ opinions of workplace flexibility. We examine whether the majority of leadership needs to express support or whether a smaller but substantial minority will reduce or eliminate flexibility bias.

In the following, we delve more deeply into the phenomenon of pluralistic ignorance and consider its relevance to the workplace. Next, we examine flexibility bias in the workplace and describe why it might be susceptible to pluralistic ignorance. We then derive our major hypotheses that disapproval of flexible work involves pluralistic ignorance and that changing perceived approval of such work will reduce bias. Finally, we describe two experiments designed to test these hypotheses.

Pluralistic Ignorance

Pluralistic ignorance entails situations where an individual holds an opinion, but mistakenly believes that others hold the opposite opinion (Allport, 1924; Prentice & Miller, 1996). As a result, individuals often behave in ways that are incongruent with their personal opinions and congruent, instead, with what they mistakenly believe to be the norm (Prentice & Miller, 1993). In a demonstration of this phenomenon, Willer et al. (2009) showed that anticipated conformity pressures caused study participants to comply with a norm they privately rejected but (mistakenly) assumed others supported. Once having complied, these participants were then more likely to publically criticize a deviate in an apparent effort to signal the sincerity of their own compliance.

Researchers have examined pluralistic ignorance across a variety of domains. For example, research by Katz and Allport (1931) and O’Gorman (1975) found that individuals often supported racial integration but mistakenly believed that others did not. More recently, Prentice and Miller (1993) conducted a series of studies showing that pluralistic ignorance underlies college student perceptions and responses to binge drinking. The authors asked students how comfortable they were with the drinking habits of their peers and how comfortable they believed the average college student was with these practices. They found that students perceived other students to be much more comfortable with campus drinking habits than they reported themselves to be.

Attitudinal contexts characterized by pluralistic ignorance are especially problematic because shared cognitive misperceptions influence behavior. For example, bystanders fail to intervene in emergency situations, and students fail to ask questions in the classroom, because they mistakenly assume that others believe no intervention or clarification is necessary (Latane & Darley, 1970; Miller & McFarland, 1987). Similarly, Prentice and Miller (1993) found that over time, college men’s private attitudes toward drinking tended to shift toward the (mis)perceived norm of greater acceptance.

Group identification lies at the heart of pluralistic ignorance (Prentice & Miller, 1996). Whereas individuals often act out of a desire to be good group members, they often interpret similarly motivated behavior in others to be reflective of personal beliefs. At work, the desire to be seen as a good group member may be especially high. Not only do people derive part of their identity from workplace organizations, the social and economic consequences of being perceived as a poor group member are particularly steep. Thus, it is possible that pluralistic ignorance may be especially pervasive in the workplace. Indeed, organizational scholars have documented widespread pluralistic ignorance among workers. For example, individuals believe they, as compared with their colleagues, are more professional and ethical (Sallot, Cameron, & Weaver Lariscy, 1998), more receptive to interpersonal feedback (Isenberg, 1980), and more burned out (Halbesleben & Buckley, 2004). We apply the construct of pluralistic ignorance to another organizational context: flexible work. To assert that flexibility bias is perpetuated by pluralistic ignorance, however, we must first understand why workers might personally be more accepting of flexible work yet assume that others disapprove.

Ideal Workers and Workplace Flexibility

U.S. firms are defined in part by what Blair-Loy (2003) calls the “schema of work devotion” in which the expectation is that work is the central focus of life. Thus, current organizational climates typically privilege “ideal workers.” Ideal workers work full-time and long hours. They are committed to their work and employer, available for overtime on short notice, and have few commitments outside of work. These expectations are based on a traditional, gendered division of labor (Acker, 1990; Williams, 2000). Historically, men have been more able to comply with these norms because women have been responsible for the domestic realm (Davies & Frink, 2014). Yet, women’s participation in paid labor is now widespread (Cohen & Bianchi, 1999; Toossi, 2012). In the United States, both parents are in the labor force for 70% of families with children (Kornbluh, 2003). Yet, women still perform the lion’s share of housework and childcare and are more likely to take time off to attend to family needs (Becker & Moen, 1999; Raley, Bianchi, & Wang, 2012). Thus, ideal-worker norms are norms that men, but not women, are better able to approximate (Acker, 1990; Williams, 2000).

In light of this inequity, scholars have advocated for organizations to adopt flexwork policies (e.g., Williams, 2000). While such policies have the potential to help women better manage the demands of work and home, research suggests that both men and women desire flexible work (Gerson, 2010). Moreover, men, women, and organizations benefit from flexible work. Flexwork is associated with reduced work–family conflict (Bond, Galinsky, Kim, & Brownfield, 2005; Kelly, Moen, & Tranby, 2011), improved physical and mental health (Galinsky, Bond, & Hill, 2004; Matos & Galinsky, 2011; Moen, Kelly, & Hill, 2011), fewer unplanned absences (Bond et al., 2005; Hammer, Bauer, & Grandey, 2003), reduced employee turnover (Galinsky et al., 2004; Matos & Galinsky, 2011; Moen et al., 2011), reduced turnover intentions (Kelly et al., 2011), increased engagement and job satisfaction (Bond et al., 2005; Galinsky et al. 2004; Matos & Galinsky, 2011), and increased productivity (Bloom, Liang, Roberts, & Ying, 2013).

Despite these well-documented benefits, employees who seek flexible work endure a flexibility stigma or bias (Blair-Loy & Wharton, 2004; Cohen & Single, 2001; Glass, 2004; Rogier & Padgett, 2004; Wharton et al., 2008; Williams, 2010; Williams et al., 2013). These workers suffer wage penalties (Blair-Loy & Wharton, 2004; Budig & England, 2001; Glass, 2004) and lower performance evaluations (Wharton et al., 2008). Moreover, flexibility bias may reduce the likelihood of promotion (Cohen & Single, 2001; Judiesch & Lyness, 1999). Workers, fearing these negative repercussions (Blair-Loy & Wharton, 2002; Jacobs 2004), often forgo flexible work opportunities (Blair-Loy & Wharton, 2002; Drago et al., 2005; Frye & Breaugh, 2004; Jacobs & Gerson, 2004).

If workers desire workplace flexibility (Crowley & Weiner, 2010; Gerson, 2010), and if even managers view flexwork as beneficial (Sweet & James, 2013), this stigmatization is perplexing. We suggest pluralistic ignorance may underlie this seemingly irrational bias. As a nontraditional structure of work designed to accommodate workers’ nonwork concerns, flexible work violates ideal-worker norms. As a result, workers may assume that others disapprove of flexible work while they themselves are privately more accepting. Thus, workers may (mis)perceive a negative social consensus about flexible work, while the true underlying social consensus is more positive. Our first hypothesis is that individuals’ opinions of flexible work will be more accepting than their estimates of how others feel about flexible work. Study 1 describes an experimental test of this hypothesis.

Flextime and Flexplace

While there is significant empirical evidence that flexible work positively affects workers and organizations, and that those who seek flexible work are penalized, there has been little direct comparison of the consequences associated with different types of flexible work. Scholars have examined the consequences of part-time work (e.g., Epstein, Seron, Oglensky, & Saute, 1999; Vandello, Hettinger, Bosson, & Siddiqi, 2013), compressed workweeks (e.g., Amendola, Weisburd, Hamilton, Jones, & Slipka, 2011), and schedule flexibility (e.g., Carlson, Grzywacz, & Kacmar, 2010). Yet, we do not know whether flexwork penalties differ depending on the accommodation requested. We explore the stigmatization of two types of flexible work: flextime and telecommuting. Flextime permits employees to adjust their hours at the beginning or end of the day provided an agreed period of each day is spent at work. Telecommuting, which we refer to as “flexplace,” involves working from a satellite location (typically from home) some or all of the time. Because ideal workers work full-time and long hours, neither of which can be directly observed when one telecommutes, we expect that workers who request flexplace accommodations will be more stigmatized than flextime requesters. We make no hypotheses regarding differences in flextime or flexplace in their susceptibility to pluralistic ignorance.

Implications for Mitigating Flexibility Bias

If personal opinions of flexible work are more accepting than estimates of others’ opinions, it follows that countering this misperception may reduce flexibility bias. Previous research finds that exposing pluralistic ignorance eliminates cognitive misperceptions (Schanck, 1934) and some of the negative consequences associated with pluralistic ignorance (Schroeder & Prentice, 1998). For example, Schroeder and Prentice (1998) examined the effect of educating students about pluralistic ignorance on their drinking behavior. Freshman participated in one of two types of discussion groups: a peer-oriented discussion with data dispelling their drinking misperceptions or an individual-oriented discussion focusing on decision making and drinking. Four to six months later, students in the former condition reported drinking significantly less than students in the latter condition.

Similarly, we hypothesize we might reduce or eliminate flexibility bias by providing evidence that positive opinions of flexible work are widespread. Specifically, we sought to identify whether, and how much, support was needed among high-level managers to affect opinions of workplace flexibility. Support by high-level managers provides workers with a particularly clear indication of the actual normative climate of the workplace.

Study 1

Study 1 experimentally tests whether perceptions of workplace flexibility involve pluralistic ignorance. This study also sought to detect differences in the perceptions of workers who request flextime and flexplace arrangements.

Sample

In sum, 422 participants were recruited to participate in a 2 × 2 between-subject experiment for pay ($2.00) via Amazon.com’s Mechanical Turk (MTurk). Given the sample, the findings are not statistically generalizable to larger populations. However, previous research finds that MTurk respondents are more representative than in-person convenience samples typically used in experimental social science research and only modestly less representative than national probability samples (Berinsky, Huber, & Lenz, 2012).

Of the 422 participants recruited, we excluded 13 because they failed the manipulation check,¹ resulting in a sample of 199 women and 210 men residing in the United States. The mean age was 32.5 (SD = 11.2) and 83% identified as White (n = 340).

Procedure and Measures

In line with previous research (Rudman & Mescher, 2013), participants learned that a fictitious company (Hatin Corporation) has shared their human resources (HR) transcripts for research purposes. The participant was then shown a transcript and told it was an actual conversation between an HR representative and an employee (K. D.)² at Hatin. Participants then read one of two scenarios: (a) an employee request to start and leave work early (but keep the same number of hours) on Mondays, Wednesdays, and Fridays (flextime) or (b) an employee request to work from home (but keep the same number of hours) on Tuesdays and Thursdays (flexplace).³ Participants were then asked to either (a) provide their own opinion or (b) provide the opinion they believed the average American to have regarding the employee and the request.

If flexibility bias is driven, in part, by pluralistic ignorance, then participants asked for their personal opinions should indicate more support for the request than participants asked how supportive others would be. Thus, our first dependent measure, a support scale (α = .92), consists of the average across responses to three questions: how reasonable respondents found the request, how deserving they felt the employee was of having the request granted, and how likely they would be to grant the request (or the average American opinion on these three measures). Next, we assessed opinions regarding the employee’s personal characteristics. If participants believe that others are unsupportive of flexwork, they may think others hold this opinion because of assumptions about the kind of people who make these requests (e.g., disrespectable, disagreeable, and uncommitted). We asked “How much do you (think the average American would) respect K. D.?” and “How admirable do you (think the average American would) find K. D.?” These items were combined to form a respect scale (α = .89). We also asked how likeable participants found the employee, and how committed, dependable, and dedicated they found the employee. These last three items formed a committed scale (α = .96). Previous research suggests that workplace biases may be linked to organizational rewards (Correll, Benard, & Paik, 2007). Thus, we also asked participants how likely they would be (or thought others would be) to recommend the employee for a promotion and how much merit raise money they would (or thought others would) allocate to the employee. Respondents were asked to select an amount of money, in $1,000 increments, between $0 and $10,000. Responses were recoded with $0 = 1 and $10,000 = 11. To be consistent with previous pluralistic ignorance studies, all other response categories ranged from 1 to 11, with 1 denoting very unfavorable opinions and 11 denoting very favorable opinions. Participants then completed a demographic questionnaire.

Results

Table 1 provides the means and standard deviations for the dependent variables, along with corresponding independent sample t tests (equal variances assumed) to compare means.⁴ Because no significant gender differences emerged within either condition, the data for men and women were pooled. The first two columns represent flextime and compare participant ratings of a person who has requested flexible work with perceptions of how others rate a person who has requested flexible work. As predicted, across all six dependent variables, participants who were asked to give their impressions expressed more favorable opinions than participants who were asked to indicate the impressions of the average American. Participants believed that others would be significantly less supportive of the request, t(208) = −1.93, p < .05, and would find the employee less respectable, t(208) = −1.54, p < .10, less likeable, t(208) = −1.36, p < .10, and less committed, t(208) = −3.02, p < .01. They also believed others would be less likely to recommend the employee for a promotion, t(208) = −4.71, p < .01, and would recommend lower merit raises for the employee, t(208) = −2.19, p < .05, compared with participants who gave their own evaluations.

Table 1.

Study 1: Means (and Standard Deviations) for Employee Evaluation Variables by Type of Flexibility Requested and Type of Opinion Provided.

Flextime		Flexplace
Personal opinion	Estimate of others’ opinion	Personal opinion	Estimate of others’ opinion
Support	8.67** (1.76)	8.20 (1.79)	6.98** (2.43)	6.39 (2.26)
Respect	7.55* (1.95)	7.16 (1.69)	6.28 (2.13)	6.04 (2.16)
Likeable	7.81* (2.00)	7.44 (1.99)	6.62 (2.18)	6.39 (2.21)
Committed	8.21*** (1.83)	7.44 (1.99)	6.76*** (2.40)	5.74 (2.30)
Promotion	7.85*** (1.95)	6.50 (2.22)	6.12*** (2.50)	5.02 (2.44)
Merit raise	4.54** (1.68)	4.05 (1.55)	3.49** (1.63)	3.10 (1.52)
Observations	102	107	101	99

Note. Asterisks indicate a significant difference between personal opinions and estimates of others’ opinions within each type of flexible work.

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

The results for the flexplace vignette appear in the last two columns of Table 1. Four of the six employee evaluation variables differed significantly in the predicted direction between the personal and average opinion conditions. Participants who were asked about the opinions of others stated that others would be less supportive of the request, t(199) = −1.78, p < .05; deem the employee less committed, t(199) = −3.02, p < .01; be less likely to recommend the employee for a promotion, t(199) = −3.17, p < .01; and recommend lower merit raises, t(199) = −1.72, p < .05, than participants who were asked to give personal opinions.

We also ran multivariate linear regression models estimating the effects of the type of opinion provided, the type of accommodation requested, and their interaction on each of the dependent variables (Table 2). For four of the six dependent variables, the main effect of personal opinion is statistically significant and in the predicted direction. That is, net of the type of accommodation requested, those who were asked to provide their own opinions, compared with those who were asked to provide the opinions of others, were more supportive of the request (p < .10), found the employee to be more committed (p < .01), were more likely to recommend the employee for a promotion (p < .01), and recommended larger merit raises (p < .05). Thus, the results largely support our hypothesis that people assess flexwork employees more positively than they believe others will.

Table 2.

Study 1: Estimated Regression Coefficients for the Effects of Providing One’s Personal Opinion (Independent Variable), Employee Requesting a Flexplace Arrangement (Independent Variable), and Their Interaction on Employee Evaluation Variables (Dependent Variables).

Support	Respect	Likeable	Committed	Promotion	Merit raise
Personal	0.474*	0.390	0.374	0.804***	1.358***	0.492**
opinion	(0.287)	(0.275)	(0.290)	(0.297)	(0.316)	(0.221)
Flexplace	−1.809***	−1.118***	−1.045***	−1.664***	−1.475***	−0.946***
(0.289)	(0.277)	(0.292)	(0.299)	(0.318)	(0.223)
Interaction	0.116	−0.148	−0.145	0.204	−0.249	−0.108
(0.411)	(0.393)	(0.414)	(0.425)	(0.452)	(0.316)
Constant	8.196***	7.159***	7.439***	7.408***	6.495***	4.047***
(0.201)	(0.192)	(0.202)	(0.208)	(0.221)	(0.154)
R-squared	0.163	0.088	0.071	0.150	0.163	0.105

Note. Robust SEs in parentheses. N = 409 participants. See text for variable descriptions.

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

A secondary goal of Study 1 was to explore differences in perceptions of workers who request flextime as opposed to flexplace. In Table 2, the main effect of flexplace is negative and highly significant (p < .01) across all six models. That is, compared with flextime, evaluators are less supportive of flexplace requests. They also find employees who ask for flexplace accommodations to be less respectable, less likeable, less committed, less deserving of a promotion, and worthy of lower merit raises than employees who ask for flextime accommodations.

The extent of pluralistic ignorance does not appear to depend on flexwork as all interactions between the type of opinion and the type of flexwork are insignificant.

Study 2

Study 2 tests the hypothesis that when individuals are presented with information that a high percentage of the firm’s leadership engages in flexible work, flexwork bias will decrease. A related goal was to determine how widespread flexible work must be in order to reduce stigma. We examined whether flexwork bias was reduced or eliminated if subjects were informed that the majority of high-status workers supported it or if a smaller but substantial minority of high-status workers supported it.

Sample

Three hundred and seventeen people residing in the United States participated in the experiment for pay ($2.00) via Mechanical Turk. Eight participants were removed from the sample because they failed the manipulation check resulting in a final sample of 309 (152 men and 157 women). Of these, 250 (81%) self-identified as White.

Procedure

Similar to Study 1, participants read an HR transcript. Again, the employee requested to come in early and leave early (but keep the same number of hours) on Mondays, Wednesdays, and Fridays (flextime) or requested to work from home (but keep the same number of hours) on Tuesdays and Thursdays (flexplace). Within each transcript, we included a manipulation designed to test the hypothesis that under conditions of widespread social consensus for flexible work, flexibility bias would be reduced. In response to the request, respondents read that the request would be decided by a committee that includes the employee’s supervisor and two HR representatives. In Condition 1, the participant then read, “Twenty to 25% of our senior managers—both men and women—work flexibly.” In Condition 2, the participant read that “more than 50%” of senior managers work flexibly. Participants in the control condition were given no information regarding the extent of flexible work.

Respondents were then asked manipulation check questions,⁵ and completed the same measures (except for one additional question regarding competence) used in Study 1. The scales were reliable, ranging from .84 to .97.

Study 1 found that employees who request flexplace accommodations experience significantly more stigma than those who request flextime accommodations. People’s stronger biases against flexplace may be more resistant to change. Given this possibility, we analyzed the results separately by accommodation type.

Flextime Results

We first present results for the flextime condition. Columns 1 to 3 of Table 3 show the means and standard deviations for the employee evaluation variables for the respondents exposed to the flextime vignette (n = 155) by condition, and Table 4 displays regression results for this condition.

Table 3.

Study 2: Means (and Standard Deviations) for Employee Evaluation Variables by Type of Flexibility Requested and Information Provided.

Flextime			Flexplace
No info	Substantial minority	Majority	No info	Substantial minority	Majority
Support	8.65	8.98	9.22	7.92	7.82	8.56
	(1.23)	(1.63)	(1.28)	(1.73)	(1.98)	(1.48)
Respect	7.81	7.83	8.51	7.68	7.42	8.07
(1.21)	(1.83)	(1.60)	(2.06)	(1.91)	(1.61)
Likeable	8.24	8.16	8.71	7.94	7.88	8.24
(1.44)	(1.80)	(1.53)	(1.85)	(1.97)	(1.68)
Committed	8.20	8.14	8.95	7.81	7.97	8.07
(1.70)	(1.97)	(1.86)	(2.03)	(1.96)	(1.89)
Competent	8.27	8.33	8.80	8.21	8.29	8.56
(1.61)	(2.02)	(1.44)	(1.86)	(1.63)	(1.78)
Promotion	7.31	7.43	8.41	7.11	7.43	7.29
(1.75)	(2.11)	(1.44)	(2.24)	(2.11)	(1.82)
Merit raise	3.88	4.51	4.33	4.09	4.00	4.42
(1.41)	(1.91)	(1.65)	(1.61)	(1.39)	(1.45)
N	51	50	54	54	51	49

Table 4.

Study 2: Estimated Regression Coefficients for the Effect of Information Provided Regarding Flexible Work by Leadership (Independent Variable) on Employee Evaluation Variables (Dependent Variables), Flextime.

Support	Respect	Committed	Likable	Competent	Recommend	Merit raise
Critical mass	0.300	0.0263	−0.0293	−0.0753	0.0655	0.126	0.618*
(0.277)	(0.309)	(0.365)	(0.317)	(0.338)	(0.369)	(0.332)
Majority	0.630**	0.770**	0.847**	0.598*	0.596*	1.205***	0.377
(0.272)	(0.303)	(0.358)	(0.311)	(0.332)	(0.362)	(0.325)
Constant	8.654***	7.814***	8.203***	8.235***	8.275***	7.314***	3.882***
(0.195)	(0.217)	(0.257)	(0.223)	(0.238)	(0.260)	(0.233)
R-squared	0.034	0.052	0.049	0.036	0.025	0.081	0.023

Note. Robust SEs in parentheses. N = 155 participants. See text for variable descriptions.

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

Table 4 presents estimated regression coefficients for the effect of extent of flexibility support on each of the dependent variables. Two dummy variables measure the extent of support—substantial minority or majority. The omitted category is the condition in which no information regarding the extent to which leaders engage in flexwork was given. By estimating the effects of each level of support, we were able to evaluate whether the use of flexible work by more than half of the leadership would reduce flexibility bias or if a substantial minority (i.e., 20–25%) would suffice.

In the flextime condition, majority use is significant or marginally significant in the predicted direction for six of the seven dependent variables. Participants who were told that more than half of the leadership worked flexibly were more supportive of the request (p < .05), respected the employee more (p < .05), found the employee more committed (p < .05), and were likely to recommend the employee for a promotion (p < .01) compared with participants who received no information about management involvement in flexible work. These same participants found the employee to be marginally more likeable and competent as well (p < .10). Thus, there is strong support for our hypotheses that exposure to information indicating widespread support for flexible work reduces flexibility bias.

While majority use reduced flexibility bias, substantial minority use—that is, working in an environment in which 20% to 25% of the senior managers work flexibly—did not. Substantial minority use did lead to a marginally significant increase in recommended merit raise (p < .10). However, no other significant differences were found for any of the remaining questions. Evidently, the stigma of requesting flextime accommodations is so abiding that even when 20% to 25% of the company’s senior managers work flexibly, flex stigma remains.

Flexplace Results

Columns 4 to 6 of Table 3 show the dependent variable means and standard deviations for respondents exposed to the flexplace vignette, and Table 5 presents regression analyses for those in this condition. The regression analyses reveal that, even in an environment in which the majority of senior managers work flexibly, significant penalties for flexplace requests remain. When subjects were told that 20% to 25% of senior managers work flexibly, and even when respondents were told that more than 50% of senior managers work flexibly, respondent opinions remained unchanged across all dependent measures as compared with those in the control condition. Thus, it appears that requesting to work from home results in such severe stigma that it cannot be attenuated by widespread organizational support as we have operationalized it in this study.

Table 5.

Support	Respect	Committed	Likable	Competent	Recommend	Merit raise
Critical	−0.115	−0.282	0.111	−0.0806	0.0719	0.283	−0.0556
mass	(0.342)	(0.367)	(0.385)	(0.357)	(0.342)	(0.401)	(0.294)
Majority	0.565	0.347	0.288	0.323	0.308	0.138	0.332
(0.345)	(0.371)	(0.385)	(0.361)	(0.346)	(0.405)	(0.297)
Constant	7.938***	7.704***	7.855***	7.963***	8.222***	7.148***	4.056***
(0.238)	(0.256)	(0.267)	(0.249)	(0.239)	(0.280)	(0.205)
R-squared	0.028	0.018	0.004	0.009	0.006	0.003	0.013

Note. Robust SEs in parentheses. N = 154 participants. See text for variable descriptions.

***

p < .01.

Discussion and Conclusion

Although many U.S. employers offer some form of flexwork, relatively few workers actually take advantage of them. Moreover, those who do are often penalized (e.g., Blair-Loy & Wharton, 2004; Cohen & Single, 2001; Glass, 2004; Wharton et al., 2008). To better understand this bias, we conducted an experiment examining the relationship between people’s own attitudes and their estimates of others’ attitudes toward workers who ask for flextime or flexplace accommodations. For workers who request either type of accommodation, we find evidence of pluralistic ignorance: Participants believe themselves to be more accommodating of flexible work requests and less discriminatory toward flexwork requesters than they believe others to be. This is important since situations of pluralistic ignorance lead people to actively enforce norms they privately do not endorse (Willer et al., 2009). In other words, people may stigmatize flexworkers despite having relatively positive opinions about them. Moreover, situations of pluralistic ignorance are associated with attitude shifts in the direction of mistakenly believed norms (Prentice & Miller, 1993). Thus, even if individuals hold no biases regarding flexwork when they enter an organization, personal opinions may be swayed such that employees become biased over time.

An alternative explanation for our findings is that participants may have been hesitant to admit their personal biases against flexible workers, but have no such apprehensions about the estimates of others. While this is a valid concern, our participants were anonymous, identifiable only by an alphanumeric ID. Research finds that people will report more socially undesirable opinions when answering questionnaires anonymously than when their identity is known (Booth-Kewley, Edwards, & Rosenfeld, 1992; Gordon, 1987). Additionally, some participants did report substantial personal bias against flexworkers. This suggests at least some participants felt comfortable disclosing personal biases.

We also find that individuals do not view all types of flexible work similarly. To our knowledge, this is the first study to distinguish between flextime and flexplace bias. We found that participants rated employees who made flexplace requests more negatively than employees who made flextime requests. While we suspect this is because ideal-worker norms cannot be directly observed when one works away from the office, multiple mechanisms may be at play. For example, it may be that technological advancements such as smartphones and virtual private networks have only recently allowed for flexplace accommodations and that people do not know how they feel about them yet. Indeed, there was larger variation in the flexplace conditions compared with the flextime conditions. Or it may be more difficult for workers to arrange meetings and collaborate with flexplace employees. Participants may imagine that this would be irritating and thus express more hostility toward these workers. It is also possible that flexplace accommodations, such as working from home, are gendered in a way that flextime is not. Because women have traditionally been linked to the home, working from home may signal femininity. If this is the case, both men and women who make flexplace requests may be seen as more feminine and susceptible to gender discrimination as opposed to flexplace discrimination. Future research should investigate variation of bias by type of request and gender. Our findings also suggest the need for future research that distinguishes between different types of flexible work.

In Study 2, we examined whether the use of flexibility by leaders influenced judgments of workers. We found that when the majority of senior managers engaged in flexible work, bias against flextime workers—but not flexplace workers—was attenuated. A smaller percentage of use, between 20% and 25%, was insufficient to reduce bias. Our experimental design did not enable us to determine why engagement in flexible work by leadership reduces flexibility bias. We argue that majority use is indicative of widespread social consensus, although it is possible that this context provides counter stereotypical evidence that companies can function when employees work flexibly. Alternatively, the scenario may have led respondents to believe that the firm in our study was an exceptional workplace where flexibility is the norm. Future research should explore why engagement in flexible work by leadership reduces flexibility bias. Additionally, a fruitful line of inquiry would be to determine exactly what percentage of leadership engagement is needed to have this effect.

Because bias against flexplace workers remained even when 50% of high-level managers telecommuted, we speculate that respondents are aware of normative ideal-worker characteristics such as working long hours and full-time. These norms are still observable in flextime, but not flexplace, contexts. Thus, we believe respondents stigmatized flexplace requesters to such an extent, even when the majority of managers worked flexibly, that it was insufficient to overcome these biases. However, we did not collect information regarding how respondents interpreted the manipulations. It is plausible that respondents presumed managers engaged in schedule flexibility (flextime) but not telecommuting (flexplace), a reasonable assumption if senior managers need to be present to supervise employees. If this was the case, exposure to information indicating senior managers work flexibly would have little effect on flexplace requests.

Our findings have important implications for both individuals and organizations. For individuals, engaging in flexible work is associated with reduced work–family conflict (Bond, Thompson, Galinsky, & Prottas, 2002; Kelly et al., 2011) and improved physical and mental health (Galinsky et al., 2004; Matos & Galinsky, 2011). For organizations, flexwork is associated with less employee turnover (Matos & Galinsky, 2011; Moen et al., 2011), more engaged and satisfied employees (Matos & Galinsky, 2011), fewer unplanned absences (Bond et al., 2005), and productivity and profit boosts (Harter, Schmidt, & Hayes, 2002). Moreover, the implications of our research for gender equality are clear. Because the current climate within most organizations privileges the ideal worker, and men are better able to approximate ideal-worker norms, finding ways to allow women to work flexibly without penalty has the potential to level the playing field.

Previous research finds that the flexibility stigma differs by race (Coltrane, Miller, DeHaan, & Stewart, 2013), class (Dodson, 2013), and organizational status (Brescoll, Glass, & Sedlovskaya, 2013). Thus, exposing pluralistic ignorance may reduce not only flexibility bias but also other forms of workplace inequality. An important area for future research is to examine how various social statuses interact with both flextime and flexplace bias, and the extent to which they are amenable to change when cognitive misperceptions are exposed.

Our findings suggest that the flextime stigma may be attenuated by publicizing that many people only disapprove of flexible work because they, inaccurately, believe others do. The information about managers’ usage shows that, if a company is serious about eliminating the flexibility stigma, a pilot program in which a majority of high-level managers—very publicly—work flexibly may be an effective strategy. This may seem daunting but perhaps not impossible for companies truly convinced of the business benefits associated with workplace flexibility. Unfortunately, reducing bias against flexplace employees may prove more difficult.

Footnotes

Acknowledgments

The authors thank Shelley Correll, Julie Kmec, Lindsey Trimble O’Connor, and three anonymous reviewers for comments on drafts of this article.

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

Christin L. Munsch is an assistant professor of sociology at Furman University. She received her PhD from Cornell University in 2012 and then completed a postdoctoral research fellowship at the Michelle R. Clayman Institute for Gender Research at Stanford University. Her current research focuses on the social psychological underpinnings of gender inequality in work and in families.

Cecilia L. Ridgeway is Lucie Stern professor of social sciences at Stanford University. Her research addresses the contribution of social hierarchies in everyday interaction to social inequality, especially gender inequality. Her recent book is Framed By Gender: How Gender Inequality Persists in the Modern World (Oxford, 2011).

Joan C. Williams is a distinguished professor of law, UC Hastings Foundation Chair, and the Founding Director of the Center for WorkLife Law at UC Hastings College of the Law. She is the author or coauthor of more than 70 academic articles and several books, the most recent being What Works for Women at Work: Four Patterns Every Woman Should Know (forthcoming NYU Press, 2014) with her daughter, Rachel Dempsey.

References

Acker

(1990) Hierarchies, jobs, bodies: A theory of gendered organizations. Gender & Society 4(2): 139158.

Allport

F. H.

(1924) Social psychology, Boston, MA: Houghton Mifflin.

Amendola

K. L.

Weisburd

Hamilton

E. E.

Jones

Slipka

(2011) An experimental study of compressed work schedules in policing: Advantages and disadvantages of various shift lengths. Journal of Experimental Criminology 7(4): 407442.

Becker

P. E.

Moen

(1999) Scaling back: Dual-earner couples’ work-family strategies. Journal of Marriage and the Family 61(4): 9951007.

Berinsky

A. J.

Huber

G. A.

Lenz

G. S.

(2012) Evaluating online labor markets for experimental research: Amazon.com’s Mechanical Turk. Political Analysis 20(3): 351368.

Blair-Loy

(2003) Competing devotions: Career and family among women executives, Cambridge, MA: Harvard University Press.

Blair-Loy

Wharton

A. S.

(2002) Employees’ use of work-family policies and the workplace social context. Social Forces 80(3): 813845.

Blair-Loy

Wharton

A. S.

(2004) Organizational commitment and constraints on work-family policy use: Corporate flexibility policies in a global firm. Sociological Perspectives 47(3): 243267.

Blair-Loy

Wharton

A. S.

Goodstein

(2011) Exploring the relationship between mission statements and work-life practices in organizations. Organization Studies 32(3): 427450.

10.

Bloom

Liang

Roberts

Ying

Z. J.

(2013) Does working from home work? Evidence from a Chinese experiment. (No. w18871). Unpublished paper, National Bureau of Economic Research, Cambridge, MA. Retrieved from http://www.stanford.edu/∼nbloom/w18871.pdf.

11.

Bond

J. T.

Galinsky

Kim

S. S.

Brownfield

(2005) 2005 national study of employers, New York, NY: Families and Work Institute.Retrieved from http://www.familiesandwork.org/site/research/reports/2005nse.pdf.

12.

Bond

J. T.

Thompson

Galinsky

Prottas

(2002) The 2002 national study of the changing workforce, New York, NY: Families and Work Institute.

13.

Booth-Kewley

Edwards

J. E.

Rosenfeld

(1992) Impression management, social desirability, and computer administration of attitude questionnaires: Does the computer make a difference? Journal of Applied Psychology 77(4): 562.

14.

Brescoll

V. L.

Glass

Sedlovskaya

(2013) Ask and ye shall receive? The dynamics of employer-provided flexible work options and the need for public policy. Journal of Social Issues 69(2): 367388.

15.

Budig

M. J.

England

(2001) The wage penalty for motherhood. American Sociological Review 66(2): 204225.

16.

Carlson

D. S.

Grzywacz

J. G.

Kacmar

K. M.

(2010) The relationship of schedule flexibility and outcomes via the work-family interface. Journal of Managerial Psychology 25(4): 330355.

17.

Cohen

J. R.

Single

L. E.

(2001) An examination of the perceived impact of flexible work arrangements on professional opportunities in public accounting. Journal of Business Ethics 32(4): 317328.

18.

Cohen

P. N.

Bianchi

S. M.

(1999) Marriage, children, and women’s employment: What do we know? Monthly Labor Review 122: 22.

19.

Coltrane

Miller

E. C.

DeHaan

Stewart

(2013) Fathers and the flexibility stigma. Journal of Social Issues 69(2): 279302.

20.

Correll

S. J.

Benard

Paik

(2007) Getting a job: Is there a motherhood penalty? American Journal of Sociology 112(5): 12971339.

21.

Correll

S. J.

Kelly

E. L.

O’Connor

L. T.

Williams

J. C.

(2014) Redesigning, redefining work. Work and Occupations 41: 317.

22.

Crowley

J. E.

Weiner

M. D.

(2010) What mothers want: Workplace flexibility in the twenty-first century. Grant report on the survey component of funded research, New York, NY: Alfred P. Sloan Foundation.Retrieved from http://ejb.rutgers.edu/faculty/crowley/WhatMothersWant.pdf.

23.

Davies

A. R.

Frink

B. D.

(2014) The origins of the ideal worker: The separation of work and home in the United States from the market revolution to 1950. Work and Occupations 41: 1839.

24.

Dodson

(2013) Stereotyping low-wage mothers who have work and family conflicts. Journal of Social Issues 69(2): 257278.

25.

Drago

Black

Wooden

(2005) The existence and persistence of long work hours (IZA Discussion Paper No. 1720). Retrieved from http://papers.ssrn.com/sol3/papers.cfm?abstract_id=799704.

26.

Epstein

C. F.

Seron

Oglensky

Saute

(1999) The part-time paradox: Time norms, professional lives, family, and gender, New York, NY: Routledge.

27.

Finn

Donovan

(2013) PwC’s NextGen: A global generational study: Evolving talent strategy to match the new workforce reality. Retrieved from http://www.pwc.com/en_GX/gx/hr-management-services/pdf/pwc-nextgen-study-2013.pdf.

28.

Frye

N. K.

Breaugh

J. A.

(2004) Family-friendly policies, supervisor support, work-family conflict, family-work conflict, and satisfaction: A test of a conceptual model. Journal of Business and Psychology 19(2): 197220.

29.

Galinsky

Bond

J. T.

Hill

(2004) When work works: A status report on workplace flexibility. Retrieved from http://familiesandwork.org/3w/research/status.html.

30.

Galinsky

Bond

J. T.

Sakai

Kim

S. S.

Giuntoli

(2008) National study of employers, New York, NY: Families and Work Institute.

31.

Gerson

(2010) The unfinished revolution: How a new generation is reshaping family, work, and gender in America, New York, NY: Oxford University Press.

32.

Glass

(2004) Blessing or curse? Work-family policies and mother’s wage growth over time. Work and Occupations 31(3): 367394.

33.

Gordon

R. A.

(1987) Social desirability bias: A demonstration and technique for its reduction. Teaching of Psychology 14(1): 4042.

34.

Gornick

J. C.

Meyers

M. K.

(2005) Families that work: Policies for reconciling parenthood and employment, Thousand Oaks, CA: Sage.

35.

Halbesleben

J. R.

Buckley

M. R.

(2004) Burnout in organizational life. Journal of Management 30(6): 859879.

36.

Hammer

L. B.

Bauer

T. N.

Grandey

(2003) Work-family conflict and work-related withdrawal behaviors. Journal of Business and Psychology 17(3): 419436.

37.

Harter

J. K.

Schmidt

F. L.

Hayes

T. L.

(2002) Business unit-level relationship between employee satisfaction, employee engagement, and business outcomes: A meta-analysis. Journal of Applied Psychology 87(2): 268279.

38.

Isenberg

D. J.

(1980) Levels of analysis of pluralistic ignorance phenomena: The case of receptiveness to interpersonal feedback. Journal of Applied Social Psychology 10(6): 457467.

39.

Jacobs

J. A.

Gerson

(2004) The time divide: Work, family and gender inequality, Cambridge, MA: Harvard University Press.

40.

Judiesch

M. K.

Lyness

K. S.

(1999) Left behind? The impact of leaves of absence on managers’ career success. Academy of Management Journal 42(6): 641651.

41.

Katz

Allport

(1931) Students’ attitudes, Syracuse, NY: Craftsman Press.

42.

Kelly

E. L.

Moen

Tranby

(2011) Changing workplaces to reduce work-family conflict schedule control in a white-collar organization. American Sociological Review 76(2): 265290.

43.

Kornbluh

(2003) The parent trap. The Atlantic Monthly 291(1): 111114.

44.

Latane

Darley

J. M.

(1970) The unresponsive bystander: Why doesn’t he help?, New York, NY: Appleton-Century Crofts.

45.

Matos

Galinsky

(2011) Workplace flexibility in the hospitality, restaurant and tourism industry, New York, NY: Families and Work Institute.

46.

Miller

D. T.

McFarland

(1987) Pluralistic ignorance: When similarity is interpreted as dissimilarity. Journal of Personality and Social Psychology 53(2): 298.

47.

Moen

Kelly

E. L.

Hill

(2011) Does enhancing work-time control and flexibility reduce turnover? A naturally occurring experiment. Social Problems 58(1): 6998.

48.

O’Gorman

H. J.

(1975) Pluralistic ignorance and white estimates of white support for racial segregation. Public Opinion Quarterly 39(3): 313330.

49.

Prentice

D. A.

Miller

D. T.

(1993) Pluralistic ignorance and alcohol use on campus: Some consequences of misperceiving the social norm. Journal of Personality and Social Psychology 64(2): 243243.

50.

Prentice

D. A.

Miller

D. T.

(1996) Pluralistic ignorance and the perpetuation of social norms by unwitting actors. Advances in Experimental Social Psychology 28: 161209.

51.

Raley

Bianchi

S. M.

Wang

(2012) When do fathers care? Mothers’ economic contribution and fathers’ involvement in child care. American Journal of Sociology 117(5): 14221459.

52.

Rogier

S. A.

Padgett

M. Y.

(2004) The impact of utilizing a flexible work schedule on the perceived career advancement potential of women. Human Resource Development Quarterly 15(1): 89106.

53.

Rudman

L. A.

Mescher

(2013) Penalizing men who request a family leave: Is flexibility stigma a femininity stigma? Journal of Social Issues 69(2): 322340.

54.

Sallot

L. M.

Cameron

G. T.

Weaver Lariscy

R. A.

(1998) Pluralistic ignorance and professional standards: Underestimating professionalism of our peers in public relations. Public Relations Review 24(1): 119.

55.

Schanck

R. L.

(1934) A study of change in institutional attitudes in a rural community. The Journal of Social Psychology 5(1): 121128.

56.

Schroeder

C. M.

Prentice

D. A.

(1998) Exposing pluralistic ignorance to reduce alcohol use among college students. Journal of Applied Social Psychology 28(23): 21502180.

57.

Singley

S. G.

Hynes

(2005) Transitions to parenthood work-family policies, gender, and the couple context. Gender & Society 19(3): 376397.

58.

Sweet

James

(2013) What do managers really think about flexibility? (Issue Brief No. 26), Chestnut Hill, MA: Sloan Center on Aging and Work at Boston College.Retrieved from http://www.bc.edu/content/dam/files/research_sites/agingandwork/pdf/publications/IB26_TPM.pdf.

59.

Toossi

(2012) Labor force projections to 2020: A more slowly growing workforce. Monthly Labor Review 135: 43.

60.

Vandello

J. A.

Hettinger

V. E.

Bosson

J. K.

Siddiqi

(2013) When equal isn’t really equal: The masculine dilemma of seeking work flexibility. Journal of Social Issues 69(2): 303321.

61.

Wharton

A. S.

Chivers

Blair-Loy

(2008) Use of formal and informal work-family policies on the digital assembly line. Work and Occupations 35(3): 327350.

62.

Willer

Kuwabara

Macy

M. W.

(2009) The false enforcement of unpopular norms. American Journal of Sociology 115(2): 451490.

63.

Williams

J. C.

(2000) Unbending gender: Why family and work conflict and what to do about it, Oxford, England: Oxford University Press.

64.

Williams

J. C.

(2010) Reshaping the work-family debate: Why men and class matter, Cambridge, MA: Harvard University Press.

65.

Williams

J. C.

Blair-Loy

Berdahl

(2013) The flexibility stigma. Retrieved from http://worklifelaw.org/wp-content/uploads/2012/12/TheFlexibilityStigma.pdf.

66.

WorldatWork (2013) Survey on workplace flexibility 2013. Retrieved from http://www.worldatwork.org/waw/adimLink?id=73898.