Page 136 - IJPS-11-3

International Journal of
Population Studies Drivers of reproductive delay in the UK

children might be felt much earlier. However, to avoid 3. Results
simply assigning an arbitrary age to model the DCEs,
empirical tests to pinpoint when, in this sample, the timing 3.1. University-educated women
of the baby’s arrival became salient were carried out. To do Thematic analysis of the focus-group data revealed that the
so, the models were run for different age groups across the four most important attributes for this group of women
whole range of ages from which data were collected (18 – were, in order of importance:
50 years old) and at the ages when the timing variable (the (i) Partner support: this was conveyed in the sense of
baby comes in n months) are statistically significant is used hands-on-ness of their partner, someone who would
as the appropriate range for that group. share childcare duties and be a fully invested co-parent
(ii) Career costs: this was expressed as a concern that taking
The results show that for university-educated women, maternity leave would result in career stagnation or
age 33 was the first age at which the baby-timing estimate allowing colleagues to overtake them on the career ladder
became statistically significant, and therefore, all final (iii) Finances: this refers to household finances being stable
DCE models were run with women from age 33 to 50 years and sufficient, not simply the amount of personal income
(n = 160). For non-university women, baby timing was (iv) Friends: these women expressed wanting their
important from the youngest age (18 years) and decreased friendship circles to also be having babies so that they
with age. At age 33, point estimates for baby timing were may share advice, experiences, and play dates.
no longer statistically significant; therefore, only non-
university-educated women aged 18 – 33 years were These attributes were used to design the DCE
included in the final analysis (n = 184). For university men, instrument which was administered online to a nationally
the timing of the baby was no longer statistically significant representative sample of UK-based women. The DCE
beyond the age of 38 and so models for this group were results showed an overall similar story to the qualitative
performed on the subset of men aged 18 to 38 years findings but with some interesting differences. As shown in
(n = 181). For non-university men, however, it was less clear. Figure 1, the ranking of these attributes by levels revealed
Running the models for subsets of ages revealed that some that household finances were deemed more important than
ages were significant and some ages had non-significant career progression costs. Partner support was clearly and by
far the most important aspect for university women, as was
estimates for baby timing but with no clear pattern expressed in the focus groups, but now we can see just how
(i.e., either younger men or older men). Furthermore, large the effect size is. The odds of choosing a scenario with
models for the very oldest ages (41+) were not statistically the highest level of partner support were 3.75 times higher
significant so these men (n = 231) were removed from the than choosing one with having a partner who travels often.
analysis. The model that included all non-university men Having sufficient and secure household finances was also
was significant overall, so for robustness, analyses on this very strongly desired and then the value of all the other
sample and the subset were conducted, yielding similar attribute levels taper off to below twice the odds, although
results (Figure A1 in Appendix). Nevertheless, findings for they are still significantly higher than the lowest level in
this group should be interpreted with caution because baby all cases except for friendship circles. Ultimately, women
timing may not be important for some of the men in this did prefer it if most of their friends were having babies but
group. This ambivalence about the timing of fatherhood there was no difference between having some friends who
was also noticed in the qualitative component of the study; were or none who were.
focus groups with these men suggested that some had not
really thought much about the topic, and it took some time The WTP results tell us how much reproductive time
to elicit meaningful information from them (Brough & this group of women is giving up by facing these barriers.
Sheppard, 2022). Figure 2 shows the same ranking of the attribute levels
but now with time in months plotted on the X-axis,
The DCEs were piloted on a sample of 60 participants denoting reproductive postponement. This helps to
to ensure that the survey was being conducted correctly interpret the DCE in a more intuitive way, as we can
and that there was minimal drop-out. Given that no issues compare the different attributes more directly using the
emerged, no adjustments were made to the study before same “currency.” As suggested by the odds ratios, women
administering it to the larger sample. Data from the pilot value a fully hands-on partner very highly; in terms of
were included in the main study (included in the above- reproductive time, they are prepared to give up 91 months
reported sample sizes). Participants at the pilot stage were (7+ years) to have this attribute. The WTP figures for the
also given the opportunity to provide free-text feedback, other attributes are much lower, ranging from around
and none of them claimed that the DCE survey was 1 year for friends having babies to 5 years for financial
difficult to understand. stability, but the large value for partner support indicates

Volume 11 Issue 3 (2025) 130 https://doi.org/10.36922/ijps.3600