*3.2. Multivariate Analysis*

3.2.1. Division of Housework and Childcare during the Lockdown

Table 1 presents the multivariate analysis results examining how homeworking is associated with the division of housework and childcare among dual-earner heterosexual couples during the first lockdown. Except for non-routine housework, on average (examining the constant, which are above 3, meaning equally shared), women were more likely to be the ones responsible for housework and childcare, even when controlling for other relevant factors. This is especially when we look at women's responses, as we see a significant positive coefficient (b) for female (0.5 or above for all models).

Compared to households where fathers were going into work, homeworking fathers were dividing housework and childcare tasks more equally with their female partners. This association was statistically significant for routine childcare (b = −0.191, *p* < 0.01), and cleaning and laundry (b = −0.176, *p* < 0.05). Although not at the traditional significance level of *p* < 0.05, we see some signs of this for home-schooling (b = −0.148, *p* = 0.093). When mothers worked from home, compared to households where they went into work, mothers were also more likely to be doing more if not all of the routine childcare (b = 0.169, *p* < 0.05) but were significantly less likely to be responsible for non-routine housework—namely DIY, gardening (b = −0.253, *p* < 0.05). Again, although it does not meet the significance level of 0.05, mothers' working from home was also associated with mothers doing more cleaning and laundry (b = 0.149, *p* = 0.094) and home-schooling (b = 0.161, *p* = 0.091).


**Table 1.** Model explaining the division of housework

 and childcare during the COVID-19

 lockdown for

heterosexual

 dual earning couples

In sum, our results show that in households where mothers were going into work and fathers are working from home, there was a general tendency for a more equal division of housework. However, mothers working from home resulted in women taking on a larger role in domestic work, especially routine childcare, compared to those who were going into the office. Taken together, this meant that there may be no significant difference in the division of housework/childcare among couples who were both working from home compared to couples where both went into work. To test whether or not there is an added effect when both parents work from home together, we have added an interaction term (father work from home\*mother work from home) in the model (full estimates can be provided upon request). The interaction term was significant in the case of home-schooling (b = −0.381, *p* < 0.01), entailing that it is especially when both parents work from home we see a more equitable division of home-schooling children, rather than when fathers are at home on their own. However, this was not the case with regard to the other types of housework and childcare.

#### 3.2.2. Increased Engagement in Childcare/Housework during the Lockdown

In the previous section, we found that when fathers worked from home, they were more likely to do carry out an equal share of the housework and childcare, yet when mothers worked from home some of this effect may have been cancelled out. The difference, however, lies in how much time couples spent carrying out domestic work, especially childcare. Looking at fathers' own perception of their time (Table 2), we can see a clear pattern. Compared to those who were going into work, homeworking fathers were about three and a half times more likely to say that they are spending more or much more time on routine childcare (odds ratio = 3.466, *p* < 0.05) and home-schooling (odds ratio = 3.613, *p* < 0.05) during the lockdown. Homeworking dads were also more likely to say that they are spending more time on some routine housework such as, cooking (odds ratio = 2.369, *p* = 0.089) and non-routine childcare (odds ratio = 2.737, *p* = 0.062). Although these associations were only significant at the 10% level, the effect sizes were large. However, compared to men who were going into work, fathers working from home were less likely to say they are spending more time on non-routine housework—i.e., DIY (odds ratio = 0.389, *p* = 0.061), albeit only significant at the 10% level. The low significance levels may be due to the smaller sample sizes for these models. Mothers' homeworking did not influence fathers' perception of whether or not they spent more time on housework and childcare.

Homeworking fathers' increased time spent on childcare can also be observed in our mother's data (Appendix B, Tables A3 and A4). In households where fathers were working from home, mothers were twice as likely to say their partners are spending more time on routine childcare compared to before the lockdown. Having said that, no clear patterns are observed for other types of housework and childcare tasks. The interaction term (available upon request) showed that it was especially when both parents were working from home, did mothers feel that fathers were engaging more in the home-schooling activities. This confirms what was found in our previous models exploring the division of home-schooling. This could be, on one hand, mothers being able to see the time fathers put into these activities, but on the other hand, that mothers may have been able to better organise homeschooling [79] to ensure that fathers are also taking part in home-schooling children when they both work from home. Again, mothers working from home did not impact women's own perception of father's increased engagement in housework and childcare with the exception of non-routine childcare—playing, entertaining children. When mothers were working from home, they noted that fathers were doing more non-routine childcare.


**Table 2.** Model explaining the changes in the amount of housework and childcare men carried out during the COVID-19 lockdown compared to before the lockdown for heterosexual dual earning couples with children under 18 (male respondents' perception of their own time)—dichotomous (more or much more).

Note. 1 = I do more or much more, 0= about the same or I do less/much less. Odds Ratio provided, and the 95% confidence interval in the parentheses. "sc" stands for "since COVID-19 lockdown". Odds Ratio is provided. \*\* = *p* < 0.01, \* = *p* < 0.05, † = *p* < 0.1.
