Six years ago I blogged about a neat paper from Mozambique that looked at the impact of providing care on caregiver labor supply. It found a 6.2 percentage point jump. However, while it had a lot to say about child outcomes, it didn’t have much to say about adult earnings and hours. A nice new paper just published by Shelley Clark, Caroline Kabiru, Sonia Laszlo, and Stella Muthuri (an earlier, ungated version is here) does.
Clark and co. are working in Kenya, in an informal settlement in Nairobi. They work with daycare providers to randomly offer women either regular daycare or quality improved daycare (and there’s a control group too). The idea behind offering quality improved daycare (where there was extra training and materials provided to the daycare folks) is that this would improve take up by the (potentially) working moms.
The moms get monthly vouchers to cover daycare (at 7.5 hours per day) for a whole year for all of their kids. To give you a benchmark of what this is worth, at baseline Clark and co. find that the women who are paying for daycare pay about $5.40 a month per child for daycare and they have an average of 1.5 kids in the daycare age range.
Clark and co. collect data at baseline and then do a follow up survey about a year later, while the program is still running. They find that about 80 percent of mothers who got the vouchers were sending their kids to daycare. As they note, this is likely not all due to the vouchers – in the control group daycare use jumped from around 10 percent to 57.6 percent. And it turns out the quality improvements didn’t get more folks to enroll their kids, so in the analysis that follows Clark and co. combine both treatment arms.
So kids are going to daycare – what happens to moms? They work more – the vouchers lead to a 8.5 percentage point (or 17% increase) in the likelihood of working. Interestingly, this effect is entirely concentrated among married women in part because unmarried mothers were much more likely to be working at baseline.
Somewhat surprisingly, there is no increase in the (unconditional) hours worked. When Clark and co. unpack this by married/unmarried mothers, the paper takes a twist: there is a significant (at 10 percent) increase in the hours worked by married mothers of 4.25 hours per week, but for unmarried mothers there is a decline of 8.8 hours. Hmm.
Clark and co. then go on to look at income. Here the average effect is an increase of 619 Kenya shillings, or more than $6.00 per month. Recalling the daycare costs from the baseline – this would be enough for one kid, but not the 1.5 they have on average. Now, when Clark and co. break the income result down by married/unmarried women, there is no statistically significant difference by marital status. So unmarried women are working less and not earning less (in fact, the point estimate is positive).
This deepens the mystery, but Clark and co. don’t leave us in the dark. They take a look at the occupational choice of unmarried women. And what they find is that single mothers are moving away from jobs that have flexible hours and are compatible with taking care of your kids on the job (e.g. washing laundry) to jobs that have more fixed hours (e.g. teaching, working in a hotel). This is the kind of employment upgrading that often gets lost in the pure labor market participation argument for providing daycare.
Does this increased labor supply and earnings come with increased agency? Clark and co. look at a bunch of dimensions (and an index) and find that this only increased decision making over kids’ health. They also take a look at intended fertility, and here too nothing much is moving.
So this is a neat addition to the thin literature on the impacts of daycare in developing countries (and it’s a particularly thin literature in Africa). And it shows pretty sizable impacts. I look forward to more in this space, particularly for entrepreneurs and farmers – which is where most African women work.