Gender, Flexibility Stigma and the Perceived Negative Consequences of Flexible Working in the UK

This study examines the prevalence and the gender differences in the perceptions and experiences of flexibility stigma—i.e., the belief that workers who use flexible working arrangements for care purposes are less productive and less committed to the workplace. This is done by using the 4th wave of the Work-Life Balance Survey conducted in 2011 in the UK. The results show that 35% of all workers agree to the statement that those who work flexibly generate more work for others, and 32% believe that those who work flexibly have lower chances for promotion. Although at first glance, men are more likely to agree to both, once other factors are controlled for, women especially mothers are more likely to agree to the latter statement. Similarly, men are more likely to say they experienced negative outcomes due to co-workers working flexibly, while again mothers are more likely to say they experienced negative career consequences due to their own flexible working. The use of working time reducing arrangements, such as part-time, is a major reason why people experience negative career outcomes, and can partially explain why mothers are more likely to suffer from such outcomes when working flexibly. However, this relationship could be reverse, namely, the stigma towards part-time workers may be due to negative perceptions society hold towards mothers’ commitment to work and their productivity. In sum, this paper shows that flexibility stigma is gendered, in that men are more likely to discriminate against flexible workers, while women, especially mothers, are more likely to suffer from such discrimination.


Introduction
Flexible working is increasingly becoming a popular method to allow working parents a better work-life balance and a major way to tackle the gender inequalities in the labour market. For example, in the European Commission's recent proposal on the work-life balance directive which includes the right to request flexible working, it is clearly stated that this directive aims to tackle the underrepresentation of women in the labour market. 1 In fact, there is evidence that flexible working, in particular those that provide workers with more control over their work such as flexitime and teleworking, can help reduce the gender wage gap. This is because these arrangements help women maintain their labour market position and stay in high paying jobs after childbirth (Chung and Van der Horst 2018b;Fuller and Hirsh 2018;Piasna and Plagnol 2017;Van der Lippe et al. 2018). To enable better labour market integrations of mothers, the UK government has introduced the right to request flexible working in 2003 for parents of young children. This has been expanded to cover all parents by 2009, and all workers by 2014 (see next section). Yet, despite the fast expansion of rights, there has been a stall in the uptake of flexible working arrangements. Looking at the four waves of the BIS Employee Survey of Work-Life Balance between 2000 and 2011, the take up of flexitime and teleworking have not changed much (Tipping et al. 2012). This is in spite of the large and growing number of workers stating that the ability to combine work with family life is important. For example, in a survey in 2010, 88% of all women and 81% for men surveyed in the UK said that the ability to combine work and family is very important when choosing their next job (Chung 2017b). One main reason behind this 'flexibility gap', i.e., the gap between the demand for more family-friendly arrangements and the actual use of it, is due to the stigma managers and coworkers have towards workers working flexibly and the negative career consequences flexible workers experience, i.e., the so called "flexibility stigma" (Williams et al. 2013). Flexibility stigma can be understood as the perception that workers who use flexible working arrangements for care purposes are less productive and less committed to the workplace. Even when flexible working arrangements are available in national and corporate policies, workers may not feel comfortable requesting or taking up flexible working arrangements when flexibility stigma is prevalent in their workplaces. The UK, which is the focus of this study, provides an interesting case in studying flexibility stigma because it combines both a relatively strong ideal worker culture (TUC 2015) with strong and fast expanding right to request flexible working. Recent reports from NGOs and the Trades Union have shown that the fear of negative career consequences is one of the most important reasons why workers do not take up flexible working arrangements in the UK (Working Families 2017;TUC 2017). Thus, for flexible working policies to be implemented properly and to achieve the goals they are meant to accomplish, we need to understand the extent to which flexibility stigma exists which will help us understand how best to tackle it.
In this paper, I argue that flexibility stigma is gendered and may be amplified during parenthood. This is because flexibility stigma is inevitably related to the discrimination and negative perception towards workers with care demands. In most societies, including the UK, there is a strong gender division of labour where mothers do and are expected to carry out the majority of care and household labour (ONS 2014;Scott and Clery 2013;see also, Kim 2018;Kurowska 2018). Similarly, men are more likely to and are more able to adhere to the ideal worker culture-i.e. be able to prioritise work above everything else, including family (Williams 1999; see also,Chung and Van der Horst 2018a; Lott and Chung 2016;Lott 2018 for evidence of this). In this sense, flexibility stigma can be gendered in that men are more likely to have negative perceptions towards those working flexibly, and women are more likely to suffer the consequences due to such perception towards those working flexibly. Again, such divisions will be amplified during parenthood, given that this is when care demands arise, and the gender division in care/household roles amplify (Schober 2013).
To show this, I use the Fourth Work-Life Balance Employee Survey for 2011, a representative survey of the population conducted by the then Department for Business Innovation and Skills in the UK. Two aspects of flexibility stigma are examined. Firstly, the biases against flexible workers-e.g., that they make more work for others, and secondly, the perception/experience of negative career consequences due to working flexibly. Despite the great deal of interest in the possible negative career consequences due to flexible working, most studies have either been based on qualitative case studies, experimental studies based in the US, or targeted surveys with limited room for generalisability. Furthermore, most studies only focus on flexible working arrangements that reduces working hours-i.e., parttime work, leaves etc., leaving us with the question what happens when other types of flexible working arrangements that maintain working hours, namely, flexitime and teleworking, are examined. This study thus aims to address this gap by examining the prevalence and antecedents of flexibility stigma, and the experience of negative consequences of flexible working, focusing on the differences between men and women, mothers and fathers.
The next section examines the definitions of flexible working, flexibility stigma, and explores theoretically why flexibility stigma may be gendered. It also introduces the UK context, in terms of family policies, labour market contexts, and gender norms to provide more background information to help us understand why flexibility stigma exists in the UK and how it is gendered. The third section examines the data and method, followed by the fourth section which provides the results. The paper ends with a conclusion and discussion highlighting the key policy recommendations to enable better use of flexible working arrangements.

Flexible Working Definitions
Flexible working can entail employee's control over when they work or where they work (Kelly et al. 2011;Glass and Estes 1997). More specifically, flexitime is having control over the timing of one's work. This can entail worker's ability to change the timing of their work (that is, to alternate the starting and ending times), and/or to change the numbers of hours worked per day or week-which can be then be used to take days or weeks off. In the broader sense, flexitime can also include annualised hours; where working hours is not defined per day or week, but calculated throughout the year, and compressed hours; where workers maintain their working hours, usually full-time, but in a lower number of days-e.g., over four, rather than five, days. Teleworking allows workers to work outside of their normal work premises, e.g., working from home. In addition to this, flexible working can also entail workers having control over how much they work-usually in the shape of reduction in working hours, e.g., part-time work, term-time only, job sharing as well as the temporary reduction of hours. In this paper, due to data restrictions I was not able to distinguish between the perceptions towards all different types of flexible working arrangements separately. However, in the analysis of the negative career consequences experienced due to flexible working, I will compare those who have used any one of the working hours reducing arrangements, against those who only use arrangements that provide workers more control over when and where they work, i.e., flexitime and/or teleworking, against those who use both work reducing and control enhancing types in combination.
Although the main focus of flexible working has been on its role in allowing workers a better work life balance (for more, see Chung and Van der Lippe 2018c), it is also used by companies to enhance workers' performance (Ortega 2009). A part of the reason behind workers' increase performance is due to a better work-life fit through flexible working (for a review, Chung and Van der Lippe 2018c; Van der Lippe and Lippényi 2018), and the subsequent increase in motivation and retention, and decrease in absenteeism and sickness (de Menezes and Kelliher 2011). However, flexible working, especially giving workers more control over their work-such as flexitime and teleworking-is also used as a part of high performance systems that increases workers' discretion over their work specifically to enhance performance outcomes more directly (Appelbaum 2000;Davis and Kalleberg 2006). As I will show later, such duality of flexible working explains why the use of certain types of flexible working arrangements are more likely to lead to flexibility stigma and negative career outcomes than others, and for whom this will more likely be the case.

UK National Context-Family Policies and Gender Norms
Before moving on, it is important to provide information on the institutional and gender normative contexts of the UK to better understand why flexibility stigma may be gendered in the UK. Institutional and gender norm contexts influence who has access to flexible working arrangements (Chung 2017a(Chung , 2018b, the nature of flexible working arrangements (Lott 2015;Kurowska 2018), and influence workplace cultures and norms (DiMaggio and Powell 1983). Similarly, the work cultures of the country, namely the prevalence of the 'ideal worker culture' (Williams 1999) and normative views on women and men's role in the family will all shape whether flexible working will be stigmatized and for whom it will carry a heavier stigma. The ideal worker culture refers to the extent to which workers are expected to work perpetually, long hours and without any other obligations outside of work, where it is assumed that work demands deserve undivided attention and priority over other aspects of one's life. Such work cultures are found to be the main culprit of why flexible working is stigmatised in the workplace (Cech and Blair-Loy 2014;Williams et al. 2013)-namely because flexible working for family purposes indicates a deviation away from such ideal worker norm.
The UK is a typical liberal welfare state (Esping-Andersen 1990), and is traditionally a male bread-winner model country (Lewis 1992). Work-family reconciliation was considered a private family responsibility till the late 1990 s, with weak public financial support for families. However, some developments have been made since the late 1990 s when the state accepted a role in the work-family policy area (Lewis et al. 2008). One of the first developments was in maternity leave: mothers can take up to 12 months off, ten of which is paid but at a relatively low rate in comparison to other European countries (OECD 2016b;Ray et al. 2010). More recently, the government has introduced a shared parental leave where fathers can take up the remainder of the statutory maternity leave when the mothers do not. This was initially up to 26 weeks after the first 20 weeks of the childbirth in 2011, but as of 2015 it has been relaxed so fathers can take up to 50 weeks of leave. The take up is low with only an estimated 2-8% of eligible fathers taking it up (BBC 2018). Only since 1998 has part-time childcare been offered to children initially for 12.5 h a week for 33 weeks/year to children 4 years of age or over. In 2011, the year of the analysis, working parents had access to 15 h of free childcare per week for 38 weeks of the year for children aged 3 and above. This has recently been extended to 30 h a week conditional on the working hours and income of both parents. The UK has one of the most expensive childcare costs within the OECD countries (OECD 2016a) and there are serious shortages of childcare and after/out of school places across the UK, with more than half of all local authorities reporting shortages (Rutter 2016).
The British right to request flexible working was introduced in 2003 "under the banner of enhancing parenting choice" (Lewis et al. 2008: 272). In the context of lack of other means for parents to address work-life balance issues-e.g., through well paid leave and public childcare-this was a policy through which the then Labour majority government aimed to enhance women's employment rates without dealing with significant costs for the government. Initially, the right was only available for parents of children under the age of six and disabled children up to the age of 18. In 2007, this was extended to carers of adults, in 2009 to parents with children below the age of 17, and finally extended to cover all workers as of the summer of 2014. The right, however, is restricted to those who have been in continuous employment with their current employer for the past 26 weeks and only one application can be made in the span of 12 months. The request has to be made by the employee, and employers can reject this request on various business grounds (see ACAS 2016).
The lack of progressive policies that support women's labour market participation and better work life balance for working parents is also reflected on the rather traditional view of division of care and unpaid labour in the UK. Despite the record numbers of women taking part in the labour market (ONS 2013) many believe that childcare responsibilities lie with the mother. According to the British Social Attitude Survey in 2012, 1/3 of all surveyed believed that a mother should stay home when they have a child under school age and only 5% of all those surveyed believed that women should work full-time when there is a child under school age (Scott and Clery 2013). In fact, there are large discrepancies between men and women in their time spent in childcare and household tasks, where it was reported that on average women spend almost three times as much time as men caring for children, and twice as much time on housework and cooking (Fatherhood Institute 2016). Furthermore, majority of women with children work part-time in the UK (Eurostat 2016). Such traditional division of labour can also be linked to the prevalence of the ideal worker culture in the UK (Lewis 1997(Lewis , 2001). The UK is well known for its long work hours culture. In 2016, full-time workers in the UK on average worked the longest hours per week of all EU 28 countries, at 42.3 h a week in paid work compared to the average of 40.3 h for the EU 28 (Eurostat 2018). Similarly, almost 3.4 million workers in the UK regularly worked more than 48 h a week in 2014 (TUC 2015). It was men that usually worked longer hour-full-time working men worked on average 43.5 h a week compared to 40.4 h a week for women, and 2.5 million men worked more than 48 h a week vs 0.9 million women. In sum, the division of labour in the UK is one where women (are expected to) devote their time to childcare and household tasks, while men spend (very) long hours in paid employment (see also, Chung and Van der Horst 2018a; Lott 2018). This consequently will shape how the use of flexible working arrangements is viewed, and who is likely to hold perceptions of and suffer from flexibility stigma. Williams et al. (2013) define "flexibility stigma" as the discrimination towards workers using various types of flexible working arrangements for family responsibilities and care purposes. They argue that it is the main cause of the slow spread of workplace flexibility in American workplaces (see also, Lewis 1997Lewis , 2001. Stigma can be defined as attributes that discredit an individual as a "less desirable kind… reduced in our minds from a whole and usual person to a tainted, discounted one" (Goffman 1990:12). Stigma can arise from an abomination of the body-e.g., disability, blemishes of the individual charactere.g., coming from mental disorder, imprisonment, addiction or unemployment, or tribal or group identity stigma-i.e., stigma towards a group of members within the same lineages/family such as nations, race and religion. Flexibility stigma stems from the fact that working flexibly can be perceived as blemishes of the individual's character in that they make workers deviate away from ideal worker image. In short, not working long hours in the office, deviating away from normal working hours, especially to meet care demands stigmatises the worker as someone who is not committed to work and thus not as productive as others. This can lead to discrimination towards theses workers and negative career outcomes for those working flexibly. Such perceptions are not only limited to managers. Colleagues and co-workers' perception towards those working flexibly, may heavily influence the take up of flexible working arrangements as well as the consequences of taking the arrangements up (see also, Van der Lippe and Lippényi 2018; Cech and Blair-Loy 2014). In sum, I define flexibility stigma as the perception of both managers and (co)workers that believe that those working flexibly for family/care purposes are not as productive or as committed to the workplace, and will effectively not contribute as much towards the company compared to those who are not working flexibly. In the rest of this paper, when discussing flexibility stigma, based on previous work of others (e.g. Rudman and Mescher 2013), I distinguish between two types. The poor worker stigma measures flexibility stigma more directly, by referring to the beliefs that workers who work flexibly are 'poor workers'-i.e., they are not as productive and don't contribute as much to the company. More specifically in this paper, it is measured through the belief that flexible workers make more work for others. The second measures the negative career consequences experienced when working flexibly due to the prevalence of the poor worker stigma. This is measured through the experience of workers, and the extent to which workers agree to the general statement, that flexible working can lead to negative career consequences-such as negative outcomes for pay or promotions. As I will elaborate in the next section, the gender dynamics in which these aspects play out will be different for the two distinct types of flexibility stigma.

Flexibility Stigma and Gender
Firstly I examine the poor worker stigma-i.e. the belief that flexible workers are less productive/committed and how gender plays a role in who is likely to perceive this. Based on the theory that self-interests shape attitudes (Chung and Meuleman 2017;Sears et al. 1980), I expect women to be less likely to believe that flexible workers are poor workers. Women are more likely to be the ones currently working flexibly for family purposes or would do so in the future (Clawson and Gerstel 2014;Singley and Hynes 2005;Kim 2018;Kurowska 2018), due to their position as the main provider of care (both child and elderly) and domestic work (Bianchi et al. 2012). In other words, for women, believing that flexible workers are less productive will entail questioning their own productivity and commitment, or that of those in a similar position. Furthermore, men and male dominated workplaces may be more aware of and more likely to (have to) adhere to the ideal worker culture (Cech and Blair-Loy 2014;Reid 2011). For men, adhering to the ideal worker/long hours culture is to a large extent a performance of masculinity, and thus a man who deviates away from this culture, especially for care purposes, is likely to be considered 'less of a man' (Williams et al. 2013). This is why many argue that although men may be more likely to work flexibly for performance enhancing purposes (Lott 2018;Lott and Chung 2016; Van der Horst 2018a), they may face stronger prejudice when using flexible working arrangements for care purposes, experiencing a further "femininity stigma" (Rudman and Mescher 2013). This is because the use of family friendly arrangements make workers deviate away from the masculine worker's image of being the providers rather than carers (Vandello et al. 2013). Thus, men may be more likely to perceive that workers using flexible working arrangements for family purposes are less productive/committed, and more specifically for this paper, make more work for others. They may also be aware of its negative career consequences more than women for the same reason.

H1
Men are more likely to believe that flexible workers make more work for others.

H2a
Men are more likely to believe that flexible working leads to negative career consequences.
However, I argue that in the UK social context, where traditional norms on gender division of labour prevail, it is more likely that women will be the targets of flexibility stigma and their careers are more likely to suffer than that of men. Again, the key cause of flexibility stigma is the perception that flexible working for family purposes will not allow workers to adhere to the ideal worker norm (Cech and Blair-Loy 2014). In the UK, women are already expected to be the main person responsible for care and domestic work, thus subsequently expected not to be able or willing to adhere to the ideal worker culture. This is one of the major reasons why women suffer from a penalty in their pay/career when they become mothers-i.e., the motherhood penalty (Budig and England 2001). Similarly, women, especially mothers, are more likely to but also are expected to use flexible working for family friendly purposes rather than for performance enhancing purposes (Sullivan and Lewis 2001;Hilbrecht et al. 2013;Brescoll et al. 2013;Kurowska 2018;Kim 2018). Again for women, employers may assume that they will use the flexibility in their work to conform to gender roles (Clawson and Gerstel 2014), meaning restricting work to facilitate family demands, resulting in a negative career outcomes in terms of pay/promotion. In fact, Lott and Chung (2016) show that when men use flexible working arrangements, they increase their overtime and gain income premiums. On the other hand women, especially mothers, seem to (have to) trade off longer overtime hours to work flexibly-not receiving any additional pay. In addition, for UK mothers, flexibility stigma may be compounded by the stigma for having taken up significant amount of maternity leave, which majority of mothers would have done (Pronzato 2009).
On the other hand, I do not expect fathers to suffer from the same degree of penalty when using flexible working arrangements as mothers. Fathers are expected to and actually increase their working hours after childbirth because of the male breadwinner ideal people hold. In fact, many studies provide evidence for fatherhood bonus in pay (Hodges and Budig 2010), and have shown that while mothers face discrimination due to their parental status in job search, fathers benefited from it (Correll et al. 2007). There is also evidence for a fatherhood bonus when requesting flexible working arrangements. For example, Munsch (2016) shows that fathers were evaluated more positively than men without children and women with children when requesting teleworking even when it was requested for childcare purposes-i.e., the progressive merit badge (Gerstel and Clawson 2018). Such positive perception can also be due to the prevailing gender norm, and sex role stereotypes, that expect men to still prioritise work even when they work flexibly. Thus I come to the following hypotheses.
H2b Women, especially mothers, are more likely to believe that flexible working leads to negative career consequences and have experienced it directly.

H2c
There is no significant difference between fathers and childless men in their perceptions or experience of flexible working leading to negative career consequences.
Finally, this paper will examine whether there are differences in the experienced negative career outcomes of flexible working between arrangements that reduce working hours versus those that enable workers more control over their work-namely, teleworking and flexitime. Work reducing arrangements inevitably make workers deviate away from the ideal worker norm in that it reduces the numbers of hours worked. As such, there are numerous studies that have shown that part-time working and other types of work reduction can result in negative outcomes for one's careers (Coltrane et al. 2013), especially for women (Connolly and Gregory 2008;Tomlinson 2006). On the other hand, as mentioned in the previous section, flexible working that allow workers more control over when and where they work can result in performance enhancing outcomes (Chung and Van der Horst 2018a; Lott 2018). Thus studies have shown that it can lead to income premiums (Van der Lippe et al. 2018; Glass and Noonan 2016) especially for men (Lott and Chung 2016), particularly when used for productivity enhancing purposes (Leslie et al. 2012). Thus, I come to the following hypothesis.
H3 Using flexitime and teleworking is less likely to lead to negative career consequences compared to using arrangements that reduce working hours, especially for men.

Empirical Evidence
Majority of the work on flexibility stigma has been done looking at case studies focusing on a certain group of workers (e.g. professionals, academics), and qualitative or experimental studies on small number of cases. For example, experimental studies (Rudman and Mescher 2013;Vandello et al. 2013) show how male leave takers and those working part-time were in fact associated more with weak, feminine traits, while lower on masculine traits. However, there were no clear evidence that men were targeted more harshly compared to women when they worked flexibly. In addition, the former study shows that women are harsher in their poor worker stigma against male leave takers, while the latter study found no gender differences in their perceptions against workers who requested reduction of hours. Cech and Blair-Loy (2014) in their analysis of academics in STEM subjects in the US found that women and parents of young children are more likely to believe that flexible working results in negative career consequences, yet the gender difference disappeared when comparing mothers and fathers. Few studies have actually looked at larger and more representative groups of workers and at outcomes that are more objective. Coltrane et al. (2013) examined the wage penalties of mothers and fathers that stop working or reduce their hours for family reasons, and come to the conclusion that although father's wage penalty was slightly higher, there are no statistical differences between the two groups. Some studies examine how flexitime and teleworking led to income premiums and penalties. Some conclude that men are more likely to gain income premiums when using flexible working arrangements (Lott and Chung 2016), yet others say that rather than gender, it depends more on whether or not the flexible working arrangements are used for personal vs productivity purposes (Leslie et al. 2012).
Finally, some recent studies examine the prevalence of flexibility stigma specifically in the UK. Sampling British working parents, a report shows that when asked why they are not using flexible working arrangements, 12% of fathers said it was because flexible workers are seen as less committed and 10% said it was because they believed flexible working will negatively impact their career. In both cases, only half as many mothers agreed to these statements at 6% and 4% respectively (Working Families 2017). In another survey focused on managers in the UK, 23% agreed that flexible workers create more work for others who do not work flexibly. 30% of managers agreed that employee's promotion prospects are harmed by flexible working but with a strong gender division of 37% of female managers compared to only a quarter of male managers (Vontz et al. 2018). Finally, in a recent survey in the UK in 2017, it was shown that 47% of all mothers surveyed felt that their careers suffered from working flexibly (Workingmums 2017). These studies provide us with useful insights on the extent to which flexibility stigma may exist, but are limited in the representativeness of their sample and the way the questions are posed. This study is thus the first, to the author's knowledge, that examines the extent to which flexibility stigma exists and the extent to which workers have directly experienced negative outcomes due to flexible working (in the UK) using a data set covering a larger representative sample.

Data
For the purpose of this study, I use the fourth wave of the Work-Life Balance Employee Survey, conducted by the Department for Business Innovation and Skills in the UK via telephone in the first quarter of 2011. The sample sizes are 1874 for the core sample, and another 893 additional boost sample that includes parents with children as well as those with non-children caring responsibilities to allow for a more detailed analysis of this population. I use the total sample including the boost sample to allow for a larger number of cases to be analysed. However, as a robustness check, I conduct a weighted analysis, which is only available for the core sample, to check whether the results change when conducting the analysis on a representative sample of the population based on the Labour Force Survey data. For more information about the data see Tipping et al. (2012).

Dependent Variables
In the first step of the analysis, the general perception towards two distinct types of flexibility stigma is examined. The first is the poor worker stigma, i.e. workers' beliefs that those who work flexibly are not as productive. This is measured by asking people the extent to which they agree with the following statement, "People who work flexibly create more work for others." The second relates to workers' belief on the negative career consequences of flexible working-measured through the variable "People who work flexibly are less likely to get promoted". For both questions, respondents were asked to choose between strongly agree, agree, neither agree nor disagree, disagree, and strongly disagree. For both variables, there was a bi-modal distribution. Thus, I have created two dichotomous dummy variables, where those who agree and strongly agree to these statements are considered as those holding poor worker flexibility stigma perceptions or perceptions that flexible working results in negative career outcomes. Note that it is not possible to know what respondents think of when asked about 'flexible working'. However, in an earlier section of the same survey, a series of questions are asked on the respondent's knowledge of the right to request flexible working as well as the availability and take up of a wide range of flexible working arrangements (more specifically; part-time working, term-time working, job sharing, flexitime, temporary reduced hours, working from home on a regular basis, compressed working week, annualised hours). Thus, it is likely that the respondent will think of these arrangements when prompted to think about 'flexible working'.
In the second stage of the analysis, I examine the direct experiences of workers. The survey asks respondents what they perceive as the negative effects on themselves due to colleagues working flexibly. This was only asked to those that have mentioned that any one of the flexible working arrangements were available in the company and were being used within the company (78% of those who said there are flexible working arrangements available in the company responded that it is currently being used). This was an open question where a wide range of options were coded. Here, I distinguish between those who have said there were any negative consequences versus those who noted that there were no negative consequences (see the online appendix for detailed coding scheme and frequencies of the responses). I also examine the negative career consequences experienced by those who have said they are currently or have taken up any one of the flexible working arrangements in the past 12 months. This is measured through the question "What have been the negative consequences of working in this/these ways?" where the respondents were given a list of answers but were also encouraged to provide other possible answers (for more see online appendix). Of the different answers, I grouped lower pay/salary, damaged career prospective, miss out on certain projects, negative relationship with colleagues all as negative career outcomes.

Key Independent Variables
One of the key questions asked in this paper is whether flexibility stigma perceptions vary between gender and parental status. In other words, I distinguish between men and women as well as parents, and workers without children. To distinguish parental status I define parents as those with children under 12, vs those without children, or with children 12 or above. I use this definition for the simplicity of the models, based on the different levels of childcare demands for the different age group of children, as well as due to data restrictions. In the second part of the analysis, which examines the actual experience of negative career outcomes of workers who have taken up flexible working arrangements, I distinguish between the types of flexible working arrangements used to see how it leads to different consequences. In the data, workers were asked whether they have the following arrangements available in their workplace, and if so whether they have used it in the past 12 months. The arrangements include; firstly, arrangements which I consider broadly relating to working hours reduction, namely part-time, term time only, job share, work reduced hours for a limited period; secondly, arrangements that entail flexibility in one's schedule, namely, flexitime, compressed working week, annualised hours; and lastly teleworking-namely working from home on a regular basis. I distinguish between (1) workers who only use (or used) working hours reduction arrangements (part-time), to (2) those who only use (or used) either or both of the schedule control arrangements-namely flexitime and teleworking, to (3) those who use(or used) both types together.

Control Variables
Based on previous studies (e.g., Cech and Blair-Loy 2014) and other related studies (e.g., Chung 2017a; Wiß 2017), I include the following variables as control variables; age, a dummy to measure whether the respondent is living with a partner, whether the partner is in paid employment, care responsibility for someone other than children, and whether the respondent has a disability. I also include various work characteristics, such as occupational level, supervisory role (dummy), permanent contract status (dummy), working hours categorised as less than 30 h (part-time), 30-48 h, and 48 h or more (long hours), union membership, size and sector of the company, both in terms of the line of business and whether or not it is a public sector. I also include gender composition of the work place using the question "Thinking about the place where you work, are the people there mostly women, mostly men, or is it about half women and half men?", where two dummies were derived indicating mostly-female workplace, and mostly-male workplace with the equally represented workplace as the reference group. Although there may be other factors that can contribute to explaining our dependent variables, I have restricted the number of controls due to the sample sizes of the data. All description of the variables can be found in the Online Appendix.

Models
Four sets of multivariate logistic regressions are carried out to examine how gender and parental status explain the likelihood of an individual holding both types of flexibility stigma, and the likelihood of having experienced negative consequences due to flexible working-for their career or due to colleagues working flexibly. For each dependent variable, I run the analysis with the total sample, and then separately for men and women, as well as use gender and parental status as an interaction term to see how the two interact in explaining perceptions of flexibility stigma and its negative consequences. For the fourth variable-namely the actual experience of negative career outcome due to working flexibly, the type of flexible working arrangements used is first excluded and then included in the model to test its effect. I use STATA 15.1 logistic for all models.

Descriptive
As shown in Fig. 1, a large proportion of workers in the UK hold flexibility stigma perceptions, with 35% believing that flexible workers creates more work for others, and 32% believing working flexibly decreases chances for promotion. Similarly, as shown in Fig. 2, on average, 39% of workers who have had someone in their work environment use flexible working, have experienced some sort of negative outcome due to colleagues around them working flexibly. However, examining the proportion of those who believe that they have experienced some sort of negative career outcome due to flexible working, the proportion is much lower at 18% of all those who has used any one of the flexible working arrangements listed in the survey saying that it resulted in a negative career outcome. There may be an understimation here, because this is only considering those who have used flexible working arrangements in the past 12 months, effectively excluding those who may have used it in the past but do not any more, possibly due to the negative career outcomes.
As expected the perception of flexibility stigma vary depending on gender and parental status. Figure 1 indicates that both gender and parental status made a difference. Men were more likely to agree to both flexibility stigma statements, especially regarding the poor worker stigma (significant examining t-test results). Non-fathers were more likely than fathers (41% vs 36%) to perceive flexible workers as poor workers. On the other hand, mothers were more likely to perceive that working flexibly will lead to negative career outcomes compared to non-mothers (34% vs 25%). Examining t-test results, parents were in general more likely to agree that flexible working leads to negative outcomes compared to non-parents. Similarly, men were significantly more likely to have directly experienced negative consequences due to colleagues working flexibly compared to women (36%), but in this case, fathers were more likely than men without children (47% vs 42%). Also, again women, especially mothers (26%) were more likely to have expressed that their careers have taken a hit due to flexible working compared to women without children (18%) or men both with and without children (11% and 13% respectively). The gender difference in having experienced negative career outcomes due to flexible working is also significant having examine t-test results. These figures already show that gender and its interaction with parental status makes a difference in worker's stigma towards flexible workers, and its consequence on those working flexibly. However, these figures do not take into account a wide range of other factors that may influence these perceptions-so a multivariate analysis is needed. Table 1 examines who is most likely to believe that "People who work flexibly create more work for others". As found in Fig. 1, even when other factors are taken into account, men are significantly more likely to agree to such flexibility stigma statements-at around 1.6 times more likely compare to women accepting hypothesis 1. Parental status does not make a significant difference. Examining the interaction term in model 1-4, this is confirmed. The difference lies mostly between men and women rather than parents and non-parents. Table 2 examines who is likely to agree to the general statement that those who work flexibly are less likely to get promoted. As shown here, there are no significant differences between men and women, partially rejecting both hypothesis H2a and H2b. Parental status is an important factor for workers feeling that flexible work can result in negative career consequences, but only for women. Mothers are almost twice as likely to believe that flexible working can come with negative career consequences compared to women who are not mothers of children under 12-partially supporting hypothesis H2b. On the other hand, parental status does not make a difference for men, accepting hypothesis H2c. Examining the interaction term between parental status and gender, I find that it is actually mothers that can be distinguished from all other groups in their perceptions of the negative consequences of flexible working. It is worth mentioning that in both flexibility stigma perceptions, the models only explain a small proportion of the variation across the population (2~6%) entailing that there are other factors that need to be considered to understand why workers hold such beliefs. As the next step, I examine the real experiences of workers, having faced negative outcomes due to colleagues working flexibly and due to working flexibly themselves. Table 3 examines the likelihood of workers to have experienced some sort of negative consequence due to colleagues working flexibly. Similar to what was found in Table 1, it is men that are more likely to perceive that they have experienced some sort of negative outcome due to colleagues working flexibly, again supporting hypothesis H1. Parental status does not matter, similar to what was found for workers' perception of flexible workers making more work for others. Table 4 examines the likelihood of workers having experienced some sort of negative career consequence due to flexible working themselves, of those who have used any flexible working arrangement in the past 12 months. Women and parents are both significantly more likely to experience negative career outcomes when working flexibly. Examining the models Table 1 Explaining flexibility stigma for men and women (odds ratios): flexible workers make more work for others ***p < 0.001, **p < 0.010, *p < 0.050, † p<0.100 N=2358 (total), 967(Men), 1391 (Women), unweighted analysis

Model
Flexible workers make more work for others 4-2 and 4-3, it seems like parental status only matters for women, and the largest difference lies between mothers versus other groups of the population again confirming hypothesis H2b. Although the difference between mothers and fathers in model 4-4 is not significant, the coefficients indicate that in fact the averages between these groups are large. These models do not take into account which arrangements have been used by the respondent, and there are large differences across groups in the types of Table 3 Likelihood of workers having experienced negative consequences due to colleagues' flexible working (odds ratios) ***p < 0.001, **p < 0.010, *p < 0.050, † p < 0.100 N=1636 (total), 607 (men), 1029 (women), non-weighted analysis Model Experienced negative consequence due to colleagues working flexibly 3-1 3-2 3-3 3-4 arrangements they take up (see Appendix Figure A1). For example, women, especially mothers, are more likely to take up working time reducing arrangements alone and with other arrangements.  Table 5 takes the types of arrangements used into account. Examining model 4-5, using part-time working and other work reducing arrangement is more detrimental for one's career compared to the arrangements where workers get more control over their work, namely flexitime and teleworking. For example, those who use part-time and work reducing arrangements, were about 14 times more likely than those who only use flexitime/teleworking to have experienced negative career consequences due to flexible working. In fact, less than 5% of workers who only use (have used) flexitime and/or teleworking have experienced any negative career consequence due to working in such a way (see Appendix Figure A2)-entailing that these arrangements may not necessarily lead to negative career outcomes. Those who use both part-time and flexitime/teleworking together were only half as likely to have experienced negative career outcomes compared to those who only use part-time and other work reducing arrangements. This confirms our hypothesis H3 where it was expected that the working hours reducing arrangements were more likely to lead to negative career outcomes. There seems to be a gender difference in the extent to which the combination of work reducing arrangements with the control enhancing arrangements led to negative career consequences. As shown in models 4-6, 4-7 and the interaction term in 4-8, the positive effect of using part-time work with flexitime/teleworking in reducing the likelihood experiencing negative career outcomes, seems to be stronger for men. For women, part-time working seem to lead to negative career outcomes even when combined with other control enhancing arrangements. Finally, when the types of arrangements used are taken into account, the larger difference now lies between parents versus non-parents rather than men versus women with parents more likely to experience negative career outcomes. However, when considering that very few fathers in the data (and population) actually work part-time, and are much more likely to use flexitime/teleworking (see Appendix figure A1), it will be problematic to conclude that gender differences do not exist based on this result. Finally, I take a brief look at the control variables (full analysis text is available in the Appendix). It seems that organisational level factors play a role in the prevalence of flexibility stigma (see also, Van der Lippe and Lippényi 2018). For example, workers in large and public companies are less likely to think that flexible working makes more work for others, and those working in public sectors are also less likely to think that working flexibly can damage your career. On the other hand, ideal culture norms may impact perceptions towards flexible working. Those who work long hours are more likely to agree that working flexibly leads to negative career outcomes and say that they have experienced negative consequences due to colleagues working flexibly. Similar to the penalties parents face, those with care responsibilities and disabilities were also significantly more likely to have experienced negative career outcomes when working flexibly. This confirms the earlier conclusions that workers who are most likely to (or expected to) use flexible working for care purposes (including self-care), are most likely to be the ones to face career penalties when working flexibly confirming some of the assumptions made in the flexibility stigma literature (Williams et al. 2013).

Robustness Check
As a robustness check, I run the analysis only using the core sample with weights to see whether the results will be different if the sample population is representative of the total labour force. Although there were some changes in the significance levels, the main conclusions made in the previous section generally remains the same especially for the first part of the analysis. Even with the core weighted sample, men were significantly more likely to agree that flexible workers make more work for others, and women especially mothers were the ones who were more likely to agree to the statement that flexible working results in negative career outcomes once other factors were controlled for. There were some changes in the latter part the analysis examining the direct experience of workers, in terms of significance levels. Although men on average were still more likely to say they've experienced negative consequences due to colleagues working flexibly, there were now significant differences between parents and non-parents in their experiences as well (online Appendix table A5 A3-4).
Again parents (both men and women) were more likely to have directly experienced negative career outcomes due to flexible working, but in the analysis with the core weighted sample (online Appendix Table A6/A7), even without having controlled for the types of arrangements used. Similarly using flexitime/teleworking alongside working hours reducing arrangements seem to have the same effect for both men and women removing the gender differences found in Table 5. It is worth noting that this result may be due to the sizes of the sample. For example, there were only 22 fathers only using part-time/working time reducing arrangements in the core sample, and only 20 cases of fathers using parttime with flexitime/teleworking. Using the core sample weighted data also showed some changes in the control variables, especially in relation to sectoral positions (more in online Appendix Tables A3 to A6) but again in general, majority of these changes are significance level changes and not directional changes.

Conclusion and Discussion
This study examines the extent to which flexibility stigma exists in the UK. Flexibility stigma is defined as the discrimination and negative perception towards workers who work flexibly, and consequently the negative career outcomes experienced by them. The results of the study show that flexibility stigma is prevalent with more than 1/3 of all workers agreeing to the statement that workers who work flexibly make more work for others, and 32% saying that those who work flexibly are less likely to get promoted. Similarly, more than a 1/3 of workers (39%) say that they themselves have suffered due to colleagues working flexibly, and 1 out of 5 workers (18%) who use/have used flexible working arrangements in the past 12 months experienced some sort of negative career consequence due to it. This study also aimed to see whether there were gender divisions in who is more likely to hold negative perceptions towards flexible workers, and who is more likely to suffer from such perceptions. The results show that men were more likely to be the ones to believe that flexible workers make more work for others, and more likely to say that they themselves have suffered due to colleagues working flexibly. This is true even when controlling for a whole range of factors. On the other hand, it was women, especially mothers (of children below 12) that are likely to have experienced some sort of negative career consequence due to flexible working, and are more likely to agree to the general statement that flexible workers are less likely to be promoted once other factors were controlled for. Some of this effect was due to the different types of arrangements used by workers. Fathers are most likely to use flexitime and telework, i.e. arrangements that predominantly give workers more control over their work, possibly for performance enhancing purposes (Lott and Chung 2016;Lott 2018;Chung and Van der Horst 2018a). On the other hand, women, especially mothers were more likely to take up part-time work and other working time reducing arrangements (such as job sharing, temporary reduction of working hours, term-time only), which is the most important factor explaining why workers experience negative career outcomes when working flexibly. The use of flexitime or teleworking is much less likely to lead to negative career outcomes, even when it is used alongside working time reducing arrangements. Thus, when controlling for the type of arrangements used, there were no significant differences between mothers and fathers, mirroring previous studies (Cech and Blair-Loy 2014;Coltrane et al. 2013). This leads us to believe that perhaps rather than gender, the purpose for which the flexible working arrangement is used for, i.e. care purposes, matters in the negative career outcomes it can result in. This is confirmed by the fact that those with care responsibility and disabilities were also more likely to have experienced some sort of negative career outcome due to flexible working. With the exception of fathers, these groups are more likely to experience stigma and discrimination in the labour market irrespective of whether or not they use flexible working arrangements (Jones 2008;Blau and Kahn 2000)-meaning that flexible working may result in a double stigma for certain workers. Policy makers should thus make sure to put protective mechanisms in place 2 to ensure that flexible workers do not suffer unfairly due to misconceptions of their productivity and commitment, especially for the already more disadvantaged groups in the labour market. This is especially important, when we take into account the fact that there is no evidence of flexible workers shirking away from work, and rather the evidence shows that flexible workers usually work harder, and are more productive for the hours worked (Chung and Van der Horst 2018a; Durbin and Tomlinson 2010;Young 2018;Kelliher and Anderson 2010).
One major implication for future studies is the need to distinguish between the different types of flexible working arrangements when looking at flexibility stigma. In the data used here, as well as all other previous surveys done before (e.g., Working Families 2017;Cech and Blair-Loy 2014;Vontz et al. 2018), respondents are asked about flexible workers and the perceived performance and outcomes of/for these workers without clear distinctions between different types of arrangements used. Flexitime and teleworking have been shown to increase overtime hours worked (Glass and Noonan 2016;Lott and Chung 2016) and have been associated to increased productivity and performance outcomes (de Menezes and Kelliher 2011). Although previous studies have shown stigmatised views towards those using flexitime and teleworking (Munsch 2016), as this study shows, these arrangements may not necessarily result in career penalties at least not nearly as bad as compared to using arrangements that reduce working hours. In fact, other studies have shown that the use of flexible schedules can actually lead to income premiums, especially for men (Lott and Chung 2016;Langner 2018). Future studies should thus be more precise when asking workers about their perceptions of flexible workers and their performance outcomes/ stigma around them to distinguish much more clearly between the different types of flexible working arrangements. The flexible working arrangements that predominantly provide workers more control over their work may be viewed more positively than once believed, but more data is needed to uncover this. Similarly, future researchers should endeavour to develop the theory of what stigma means for different types of flexibility. As presented in this paper, flexibility stigma can be complex and multi-dimensional with different possible applications for different types of flexibility. The context in which flexible working is being used may also be of importance (see also, Van der Lippe and Lippényi 2018; Kurowska 2018; Chung and Van der Lippe 2018c), which may explain the difference between this and studies done in the US (e.g. Munsch 2016). Again further work is needed.
The findings that flexitime and teleworking may not neccessarily lead to a large negative career outcome is a welcome one. Actually, these arrangements have been shown to reduce women's likelihood of working part-time after childbirth (Chung and Van der Horst 2018b), reduce the motherhood penalty (Fuller and Hirsh 2018;Van der Lippe et al. 2018), and potentially increase wage premiums for women especially in the longer run (Langner 2018). Then why despite the stigma and negative career outcomes do women continue to work part-time after becoming mothers? One major reason can be found in the gender norms in the UK, which expect mothers to work only part-time for the sake of children's well-being, especially before they reach school age (Scott and Clery 2013). Another reason is because other arrangements such as flexitime and teleworking are not readily accessible for women (Chung 2018a). In other words, women may have no choice but to resort to parttime work. So despite all its problems, to remain in the labour market, women may involuntarily choose to work part-time (Young 2018). However, another interpretation is possibleperhaps because part-time work is predominantly used by mothers to adapt work around family demands, it comes with a larger stigma. This is despite the fact that part-time working usually results in increased productivity for the hours worked (Künn-Nelen et al. 2013;Durbin and Tomlinson 2010). On the other hand, flexitime and teleworking is used by other groups of workers, mostly fathers in this survey, and not necessarily used for family-friendly purposes (Clawson and Gerstel 2014). These arrangements may not be associated with such stigma because of the expectations employers and co-workers have towards these groups of workers in the extent to which they will uphold the work devotion schema.
Encouraging employers to readily provide alternate options of flexible working for mothers, and provide easy access to and encouraging fathers and other groups of workers to use flexible working for broader work-life balance purposes could be helpful in tackling such flexibility stigma, and to ensure that it does not result in further traditionalisation of gender roles. Providing a stronger legal entitlement to flexible working can also help in ensuring that workers who work flexibly are not penalised, and may ensure that a more gender equal take up of flexible working for family purposes. It is likely that as flexible working for family purposes become more of a norm, such stigma towards these workers may be removed (see also, Van der Lippe and Lippényi 2018). Ensuring the wide spread of the message that flexible working leads to increased productivity and other benefits for the company and the society especially in the longer run, will also be useful in making sure that flexibility stigma is eradicated. A large and growing evidence shows that there is a strong business case for flexible working (de Menezes and Kelliher 2011;Beauregard and Henry 2009). It is time that views about flexible workers' commitment and productivity are brought into line with current realities.