Humans are inherently social creatures – we like to fit in and take many cues about how to behave from those around us. But what happens when we misinterpret group norms? What is the impact when we adjust our own behaviours to fit in with others based on incomplete knowledge?
According to recent research, 76% of mothers and 73% of fathers would like to work flexibly to spend more time with children – a remarkably similar proportion. Yet mothers are much more likely to work part-time than fathers. 28.5% of mothers with a child aged 14 or younger have reduced their working hours for childcare reasons, compared to just 4.8% of fathers. In the UK, fathers and non-childbearing parents are eligible to take up to 50 weeks of Shared Parental Leave (SPL) and 85% of men in one large survey believed they should be as involved in caring for children as women. However, since the introduction of SPL in 2015, estimates for uptake vary from just 0.5% to 8% among eligible fathers.
So if dads want to work flexibly, what’s stopping them? One barrier might be related to the idea of pluralistic ignorance. This is the tendency of people to hold a particular opinion privately while mistakenly believing the majority of people disagree with that opinion. For example, if men mistakenly think that their colleagues and managers would disapprove if they worked flexibly, then they avoid doing so, anticipating negative social and career repercussions.
Working in collaboration with Santander UK and a second global bank, we set out to investigate whether pluralistic ignorance could be holding men back from taking more parental leave and working flexibly. First, we ran a baseline survey at each bank, where we asked male participants whether they supported male colleagues who took longer parental leave and worked flexibly, in order to balance work and non-work responsibilities. Then, we asked them to estimate what percentage of their male colleagues they thought would be supportive of the same behaviours. This allowed us to measure participants’ perceptions of norms.
Combining the responses from the two questions enabled us to see whether there was a gap between private opinions and perceptions of norms. We found very similar trends at both banks:
- Respondents at Santander UK would encourage men they work with to take 8 weeks of leave, but thought their male colleagues would only encourage around 6 weeks. At the second bank, this was 12 weeks and 8 weeks.
- Men at Santander thought that roughly 65% of their peers would encourage male colleagues to work flexibly, while in reality 99% would do so.
- At the second bank, twice as many men supported male colleagues to take at least 4 months of leave as men assumed.
Once we’d established that pluralistic ignorance was occurring we provided participants with feedback on their peers’ actual beliefs – allowing men to update their perceptions of social norms closer in line with actual norms.
We ran a two-arm Randomised Control Trial at each bank where we gave the treatment group feedback about their colleagues’ true views, based on the baseline survey. The control group did not receive any feedback. At Santander UK, we told participants:
We have already asked some male colleagues in Santander their views on family leave and flexible working. The survey revealed that: The majority of male respondents would encourage their male colleagues to take 5 weeks or more of family leave
At the second bank, we tweaked the feedback and said:
“We invited 1,100 men at [this bank] in the UK to tell us their thoughts on men taking parental leave. Of the respondents: 7 in 10 managers told us that they would be supportive of men they manage taking at least 6 weeks’ parental leave. Of those managers, 74% were supportive of men they manage taking at least 12 weeks parental leave.”
The surveys were sent to over 2,200 men working at Santander UK (with a response rate of 50%) and over 4,000 at the second bank (with a 25% response rate). We would have liked to measure actual parental leave taking and flexible working behaviours within the treatment and control groups post-intervention. The reality was that we could only measure future intent – as our partners’ personnel systems, like many employers, could only capture contractual flexible working changes (such as reductions in working hours), and not some of the more informal flexible working arrangements we were equally interested in encouraging. Foro parental care, the infrequency of uptake meant we would not be able to detect changes during our trial timeline, and parental leave was not always recorded consistently.
Before we ran these interventions, the average level of leave taken was 4 weeks at Santander UK and 2 weeks at the second bank. Both trials produced a statistically significant shift within the distribution of the responses. At Santander UK we saw a 62% increase in the proportion of men intending to take 5-8 weeks of leave in the group that received feedback. At the second bank, we saw a 50% increase in the same category.
Figure 1: Self-reported intended weeks of parental leave at Santander UK
Figure 2: Self-reported intended weeks of parental leave at the second bank
Overall, the average number of intended weeks of parental leave did not differ significantly between the control and treatment groups. At Santander UK, we saw an unintended effect whereby the number of men intending to take more than 16 weeks decreased by 59% in the feedback group. We think this may be because we potentially anchored people to the single length of 5 weeks referenced in the feedback. We managed to avoid a similar effect at the second bank because we instead referred to a couple of lengths of longer leave (6 and 12 weeks) in the feedback.
We only tested whether pluralistic ignorance was having an impact on flexible working amongst men at Santander UK. There we told participants ‘Almost 100% of male respondents would encourage their male colleagues to work flexibly in order to balance their work and non-work responsibilities’.
We found that participants who were given this feedback were 4% more likely to say they would work flexibly in the future than the control group – a statistically significant finding. Whilst the increase was fairly small, it was likely limited by a ‘ceiling’ effect i.e. there was already a high average likelihood that the control group would work flexibly in the future, so there was not much room for movement in the treatment group.
Figure 3: Self-reported likelihood of working flexibly in future in the control and treatment groups at Santander
Taken together, our full analysis suggests that sharing accurate feedback about high levels of support for men to take parental care and work flexibly, has several positive outcomes:
- it encourages men to plan to work flexibly and take parental leave in the future
- it increases men’s support for other men to do the same
- and it reduces the gap between actual and perceived norms
Employers who want to implement a similar intervention, should start by measuring existing attitudes among staff, to understand whether there is strong support for men to work flexibly or take longer parental leave.
Whilst we provided feedback via a simple survey, it should be feasible for employers to share feedback via different comms channels – for instance using email, posters, or intranet systems. This is a low-cost intervention for employers who would like to increase male uptake of flexible working and parental leave. This in turn may help to reduce the gender pay gap and create more equal sharing of work and care responsibilities between women and men.
This work was funded by the Government Equalities Office as part of the Gender & Behavioural Insights Programme.