Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=

# A Free Lunch? How Changing Childcare Defaults Increases Parental Labor Supply

Authors: Justus Bamert  
Discussant: None  
Video: https://www.youtube.com/watch?v=0mE9GNhwdY8&t=15171s


## Talk (04:12:51 – 04:57:25)

[04:12:51] right, everybody. Welcome back. If we could get everybody to come on in and take a seat.  
[04:13:00] We will have two 45 minute presentations to start off the afternoon. These will be of the wait 10 minutes to ask your question format. As a reminder, please do use a microphone. There are three  
[04:13:14] around the room. Um, and with that, we have Bammer who will have the first presentation.  
[04:18:38] So those I I well  
[04:20:51] I I know that I'm two minutes ahead, but okay. I I I view this as very different from the other types of defaults. Very different in the sense that  
[04:21:05] uh individuals are not opting in because there's a real friction of the usual type, but that individual mothers feel guilty. And  
[04:21:18] that's a very different type of friction. Okay. And so this is a case in which if nobody did it and then one person decided to do  
[04:21:31] it, she would be viewed and we know the words that are used in Northern Europe to to shame that woman. But if everybody does it, then she's not going to be  
[04:21:46] shamed. And that's extremely different from the retirement defaults.  
[04:22:03] They were called raven mothers, right? your child at 3:30 and then take them to  
[04:25:34] another location >> and for lunch as well if you opted in before the reform.  
[04:25:52] I was just wondering like even if you opt in, did you have to show up and move them because that still disrupts your workday? Okay.  
[04:29:54] Is enrollment does enrollment closely track actual utilization or are there lots of kids who are enrolled but don't regularly attend the lunches?  
[04:31:29] how to reconcile 50% usage in the population overall with potentially a a a shame or um social norm mechanism. Do you see sort of like a biodal distribution of schools ahead of time where there are schools with really high  
[04:31:44] utilization and schools with really low utilization or is it more evenly distributed following up? Is there variation across  
[04:32:03] grades within the same school in New Zealization or you know cohorts that have more or few fewer working moms.  
[04:32:19] >> Exactly. But not in the for the in the kids, right? Yeah. Because that's real unfortunately.  
[04:32:26] [laughter] >> I have a question about Wednesdays. What happens on Wednesday?  
[04:32:34] [laughter] Because we can fix up Monday, Tuesday, Thursday, Friday, but without Wednesday, we're sort of missing the middle of the week.  
[04:35:17] But are you taking into consideration that some parents of older children have a younger child?  
[04:35:25] >> No. the previous slide, >> you said no effects on parents with older children, >> but is that conditional on the fact that  
[04:35:35] they don't have a younger child? I think maybe building a bit on  
[04:37:25] Claudia's question about what happens with Wednesdays, are there certain types of occupations or or fields that you could see that allow for say like Friday through Monday work schedules versus not. And can you look at whether there's  
[04:37:39] heterogeneity and responsiveness depending on whether if you're in a field that allows for a part-time work schedule that is conducive to that.  
[04:37:58] Uh, so I was also going to ask, so since your effects are um concentrated among parents of kindergarteners and you said 50% of families were using it at baseline, but is that much lower among families of kindergarteners?  
[04:38:10] Oh, it is about the same. Okay. Um, because I also had in mind a coordination story where like your kids more willing to go if they know their friends are going, but maybe if 50% are going, it's okay.  
[04:38:37] because you've already have established a habit with you and your child and you're older Keterus Parabus and you know your child is used to you picking them up and going to you know to eat lunch at home  
[04:38:51] >> or not picking them up but they walk to you at home but they walk back home and they eat lunch at home.  
[04:38:57] >> So could be habit. >> I I have a question. I don't know if you mentioned do you have time use survey data >> just Oh okay. I was going to ask if you could measure impact on home home  
[04:39:11] production on like domestic work. I just wanted to say that early release on Wednesdays exists in the US too. You know, exactly the same as [laughter] that  
[04:39:26] >> in some in some places. has the same thing.  
[04:41:57] Do you know which days? Because that would help you distinguish between a change in norms about the acceptability of having the kid there versus coordination that the kids are only willing to be there on days that their  
[04:42:11] friends are also there. I I know that I haven't explored it to that extent but um we will go later into this co coordination direction with the  
[04:42:26] Thanks. Sorry I'm a bit distracted. Sorry about that.  
[04:42:30] >> Didn't working. Okay. Okay. So I just showed you right the default is a key driver. Um now I will go into the mechanism. So um yeah there was one more question  
[04:42:43] >> if you qualified for the subsidy before. And then now you want to opt out. Do you have to regrow through hoops to qualify for the subsidized again price or is it does it automatically translate back to  
[04:42:58] the price >> if you opt out at some point? Yeah, exactly. You would have to re reapply.  
[04:43:04] Yes. >> Okay. So there could be some Yeah.  
[04:43:07] there. >> Yeah.  
[04:43:08] >> Yeah. Thanks. >> Okay. I just showed you that the default is very important in this setting. Um in standard household decision models we would think the default should have very little effect right because parents are  
[04:43:22] fully optimizing. Um so to understand what's going on I will proceed as follows. So first I will investigate which families are actually driving these results. Second um I will analyze or investigate whether this is a setting where the decisions of other parents  
[04:43:37] matter. And then finally I will use uh surveys that the department of education conducted with parents to understand how do parents themselves explain this decision.  
[04:43:48] So first here I show you the results from complier analysis where the effects of the reform are driven by these compliers and I define a complier to be um a family that is induced to take up public lunchtime childcare. And to  
[04:44:01] characterize the compliers I comply I compare um a complier with a the share of compliers with a certain characteristic to the population share and you can see right so the compliers are very similar in ter terms of  
[04:44:15] household earning or place of birth of the parents to the population but they're more likely to be from uh two parent households from those families that had a male breadwinner model before the reform and from th um from those  
[04:44:28] countries that have a high agreement with this statement that children suffer with a working mother.  
[04:44:35] So [clears throat] right so we kind of postpone the puzzle by one step. So now the question is okay these are families that are very likely not indifferent between home care and public childare.  
[04:44:46] So the question would be why do these more conservative families stay in public childare? Here I just want to focus on one key feature of the reform.  
[04:44:54] So I think some of the questions already went into this d direction right. So with the reform, all children are initially signed up in public into public lunchtime childcare unless parents opt out, right? And it could be that this actually changes how parents  
[04:45:07] perceive public lunchtime childcare. And I think it also connects a bit to this this literature regarding misperception of norms or P effects can be a driver of norm or cultural cultural changes.  
[04:45:18] >> Question is that the norms question? >> Yeah. Who answered the norm's question >> in this heterogeneity before?  
[04:45:25] >> And when you said that they from uh high agreement >> no that's yes >> who answered that question. Yeah, thanks. So I mean here I don't want to oversell this result, right? So what I do is I take the country of origin of  
[04:45:40] the parents and then just take the average. So it's not an individual response. It would be really really nice to have it but >> so I approximate right like it's just a kind of approximation of of the the likely norms that the parents.  
[04:45:53] >> So but then in fact it makes even more sense if I come from a country in which the norms are different from mine.  
[04:46:03] >> Yeah. then I might not want to deviate. But now I'm being told everyone's deviated and then I say great and so that may be why.  
[04:46:14] >> Yeah. Yeah. Nice. Yeah. >> Okay.  
[04:46:18] >> Thanks. All right. So regarding the peer effect story um I will do two tests. So first they will um analyze whether there are peer effects uh for childcare usage before the reform and then in the second step whether the reform has stronger  
[04:46:33] effects in classes where there was low pre reformed childcare usage. This is basically exactly what I find. So um first I find that children are more likely to use child care when their peers do. Um so here I show you the  
[04:46:47] results from kind of this Hawkby style peer regression um with the outcome whether child is enrolled in any lunchtime childare in grade three. And on the x-axis um I measure um how many  
[04:47:01] of your grade one peers have used child care before. And this shows you that if you qu kind of quasi randomly have more peers in grade one that have used childcare before in grade three you will  
[04:47:15] be um way more likely to also use uh public uh public lunchtime childcare.  
[04:47:23] Then the second piece of evidence now I split the sample by this baseline peer childcare usage and you can see on the left kind of in the left estimate that the effects of the reform are way stronger in classes where there's low  
[04:47:36] baseline peer childcare usage. And together this suggests right this is a setting where um the decisions of other parents matter and that the reform partially substitutes uh for this peer  
[04:47:49] induced take up. All right. So now coming to the last piece of evidence um on the mechanism.  
[04:48:00] So here I show you results from these parental surveys and um these surveys parents are asked here this is just for the result for parents that stay in this automatic lunchtime childcare and  
[04:48:14] parents asked okay why does your ch child attend automatic lunchtime childcare and they can select up to three um three answers and what is yeah many parents they highlight the the  
[04:48:26] direct benefits of this uh automatic lunchtime childare So they say, "Oh, it makes work family compatibility way easier and it's also beneficial for the for the child." And right, I think  
[04:48:40] that's that's surprising given that public lunchtime childcare was before.  
[04:48:45] And um if these are parents that have not used childare before, it suggests that they changed or update their beliefs about the benefits of lunchtime childcare.  
[04:48:57] They also highlight uh it's just part of the new concept. So it's kind of it's just what you do now. And the the the the friends of my child, they're also there going into this direction of the pure effects.  
[04:49:10] Then um what what seems to matter a bit less is it's not about logistics. So it's not a bit not the reason that many select. Um it's also not about the price and interesting interestingly it's also not about the child's preferences. So  
[04:49:25] those somehow don't seem to matter. So I have I have enough time.  
[04:49:33] Um here I just show you the results for um parents that actually opted out of public lunchtime childcare. So um you can see here you can see that the reasons for opting out are still about  
[04:49:48] connected somewhat to these preferences for home care. So they say ah someone is always at home or we eat lunch together as a family are very prevalent reasons but I think here right it's important to to know that there are way fewer parents  
[04:50:02] that actually opt out so and also given the complier results there are still many that that have these characteristics that indicate that they have more conservative norms. Yes.  
[04:50:13] >> These are mostly family style meals in the classrooms right?  
[04:50:16] >> Yeah. And so how do they accommodate students who have dietary restrictions especially perhaps due to like religious restrictions?  
[04:50:28] >> Yeah, that's a good question. So uh >> so the reason I'm asking right is because students whose dietary choices are constrained by religious restrictions may also be students from families who have particularly  
[04:50:42] conservative norms. >> Yeah.  
[04:50:44] >> Okay. I mean, I think like this hypothesis wouldn't probably show up in these answers, right, if it was really affecting a lot of those parents that that opt out. But, um, >> well, it's not actually clear to me what  
[04:50:59] I mean, I guess it might be under other reasons.  
[04:51:02] >> Um, but if we think about this as being a response to norms, >> Yeah. Right. Then I think understanding whether you know what this is signaling  
[04:51:15] for whom is this normal and is it more are are you more included by staying in the classroom with the other children but eating a separate diet or are you  
[04:51:27] more included by not staying in the classroom such that your family's restrictions are not as observable >> to the the school or the aspects.  
[04:51:39] >> Yeah. Yeah. No, it's a good point. I never I've never looked into that, but thanks.  
[04:51:47] >> All right. Just to uh to wrap up the mechanism part, I don't find much support for alternative me mechanisms.  
[04:51:54] So, I don't find that this is a one one story could be that this is about information. So, parents don't know about public lunchtime childcare or that they don't know how to sign up. Um that doesn't fit very well the evidence.  
[04:52:06] So many most of the parents they report that they're very well informed and also don't find that there are differences between parents that have used childare in the past. Um another story could be adjustment costs. So it's very difficult  
[04:52:18] to um to opt out or sign in. Um again like the process is very simple.  
[04:52:25] You can just send the mail to the school and take your child out and also right it doesn't fit this evidence that parents actually actively add services and also respond on the labor supply.  
[04:52:35] Yeah. But as you said before then you'd have to re-qualify for the subsidy right so there's a segment of parents for which it's not that easy >> separate question I have the mic do you have information on cost to the  
[04:52:50] government of all this or should we you said supply then expand so should we think of this as low cost zero cost >> it's coming up on the next slide so I defer the question so just the last point quality improvements also doesn't  
[04:53:02] seem to expl explain these results So um exactly in this surveys for example parents are also asked about the food quality and it's actually one of the most contentious topics and also spending per child on the side of the government falls um right because  
[04:53:17] they're just way more children involved now.  
[04:53:22] Okay. Um so now I have five more minutes and the last part will just be about um kind of the welfare implications um of this reform. Um what I do in this  
[04:53:33] section is um I uh I I estimate the marginal value of public funds. So the it's defined as the willingness to pay divided by the net um net costs of this  
[04:53:46] reform and um maybe regarding the the costs right given your question before um there there are costs for the for the government to provide this additional lunchtime child care for example in  
[04:54:00] additional um so every meal is subsidized and um they also have to hire more child care personnel some of the schools also make adoptions in the in the school facility ities. So overall um  
[04:54:11] it costs around $1,000 US per child per year to um to provide this automatic lunchtime childcare to the for the government regarding the the willingness to pay. Um  
[04:54:25] so here that's a key conceptual choice in this calculation of the marginal value of public funds. Um so um because it depends on the underlying economic model or decision model of the  
[04:54:39] parents right if parents are fully optimizing then this earnings gains that parents experience would be uh offset by the utility of work that they experience. So in this world I will for this world I will define the willingness  
[04:54:53] to pay as uh the cost of provision of the government um for uh providing this service. If we are in a world where parents were previously constrained, right? So when they experience these uh  
[04:55:06] behavioral frictions, then the earnings gains could reflect the value of reducing these frictions. And and for this benchmark, I will define the willingness to pay as the earnings gains that parents experience.  
[04:55:19] So what how high are these marginal value of public funds? So in the fully optimizing benchmark, I find that that it's slightly above one. So each Frank invested has a slightly more in  
[04:55:33] benefits. If we focus on parents with children in kindergarten, right? Um these these effects are slightly the MVPF is slightly larger. And then finally if we move to this constraint  
[04:55:45] benchmark, the MVPF goes up to 3.4 uh 3.4.  
[04:55:51] So the together the takeaway from this welfare analysis is that the the reform has positive welfare returns. in particular when we assume that parents were not fully optimizing before and that the reform reduces these frictions and allows parents now to to experience  
[04:56:06] this earning scales. So just to conclude, I showed you um that the changing public childcare defaults have very strong effects on parents decisions, right? They influence  
[04:56:19] public childare usage as well as these labor supply decisions. And the opt out default really prompts parents to reorganize or rethink their childare and labor supply arrangements. And the results are consistent uh with reducing  
[04:56:34] these uh with reducing uh behavioral frictions. and the implication what are the implications. So um I I think I hope I showed you that behavioral frictions can be an explanation for these small uh  
[04:56:47] labor supply uh responses that we sometimes see in uh in opt-in public childare systems and the results highlight the importance of policies such as this opt out default that can reduce such frictions.  
[04:57:01] All right. Thanks a lot. [applause] [laughter]