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.
=

# Activating Parental Managerial Capital: How Role Clarity Improves Education Outcomes

Authors:   
Discussant: None  
Video: https://www.youtube.com/watch?v=PHYYOyrSItw&t=15161s


## Talk (04:12:41 – 05:02:40)

[04:12:41] Okay, I think we're ready to uh start again.  
[04:12:49] Okay. So, uh the next speaker is uh Mahunan Yedomifi from Dartmouth. You have 50 minutes. Welcome.  
[04:12:56] >> Thank you very much. Uh thank you to the organizer for you know adding my paper uh to the program. So uh I will give you you know a high level you know uh uh motivation of uh this uh project. So uh  
[04:13:10] several years ago when I was you know in lower secondary uh in uh Ivory course in the neighborhood you know most kids you know we were playing together but mostly all of us you know our parents had little to no education and in the neighborhood there were few friends  
[04:13:25] whose parent were kind of educated but we're all going to the same school you know we all had I would say you know all the same books and and so on and whenever we are playing there these few friends whose parent had some education would just stop at a specific time and  
[04:13:40] be like we need to go home and we will get mad because they are ruining the whole fun that we were having right so fast forward when you I look back I realized that you know these few friends you know uh did really well when it comes to education but at least as far  
[04:13:54] as I remember we're all going to the same school we all had the same books and so on so that got me uh really thinking about you know the large and persistent gap that we see in terms of you know performance when it you know uh across social so social economic status  
[04:14:08] and we mostly view, you know, those differences through the lens of, you know, differences in in resources, financial resources, you know, parents, you know, uh, direct, you know, instructional input, you know, uh, that they can provide to their children and  
[04:14:21] so on. But uh you know uh looking at what happened you know with my friends there is another question that you know I started thinking about which is is it possible that some household are just better in terms of translating you know the educational investment into uh  
[04:14:36] learning which does you know has nothing to do with maybe you know uh uh financial resources or the ability of the parents to uh read to their kids and things like that you know. So what this project uh tried to do is to argue that  
[04:14:49] when we think about parents you know yes parent play this financial role and you know sometime you know uh this uh direct role of you know uh you know supporting the children education by helping the children read and so on but on top of  
[04:15:03] that parents you know manage a critical input that's uh play an important role in human capital uh uh formation and that critical input is children time and effort you know and think of this as you know things that those who are here who  
[04:15:17] are parent already already do right you have a clear routine at home you know when to stop you know telling your children to stop watching TVs and and things like that and that's exactly you know the input that you know I want to  
[04:15:30] focus on you know uh in this uh paper so the question here is you know can you know activating you know this overlooked role basically just you know managing children time and effort can activating this you know with parents who have  
[04:15:44] little to no educ education is that something that can improve their children uh educational outcome and the way I you know proceed to answer this question is to you know uh you know work with eighth grade uh student and a  
[04:15:56] parent in uh rural uh southern Benin and first thing that I will show you today is that I document a a large gap between you know what parents perceive as being their role in their children education and the actual  
[04:16:10] action that they they they take toward that and then I designed a low cost phone based is you know intervention that uh uh try to makes parents managerial role salient and provide concrete and feasible recommendation to  
[04:16:24] these parents and support parent through uh the implementation of uh you know uh these managerial uh practices and then I estimate the impact of that intervention on parental practices you know student  
[04:16:36] effort learning and and dropout a year after the intervention. So here is a preview of the result. So first you know I will show you that uh parents you know endorse you know this managerial role  
[04:16:49] that I'm talking about overwhelmingly parents endorse that role but uh when it comes to actual you know action toward those role you know most of most parent do not uh uh take any action and second uh I will show you that the intervention  
[04:17:03] that I designed help narrow this role action gap and third uh parents become better informed about their children academic standing while the intervention did not provide any you for academic standing information to to uh these  
[04:17:16] parents and also absentis student absentism also false and then I will show you that the intervention improve student test score GPA and grade completion and grade completion is very critical in this environment because on average 30% of student need to repeat  
[04:17:30] the grade where you know they are and for some subgroup it's up to 70% you know uh grade repetition and finally after the intervention I go back and collect data on dropout and I find that dropout decreases for girls and for  
[04:17:42] students transparent are illiterate. So this project makes a few contribution to the literature. uh you know we've seen a lot of project today on you know managers and and so on and we have you know literature that shows that you know differences in management practices  
[04:17:57] affect you know productivity within firms but if you think about it household are also like organizations you know I would say you know the oldest organization maybe you know on earth where you know every household have you know some resources that they need to  
[04:18:11] turn into some output. So I bring that idea to the you know of uh uh you know this idea of uh uh uh uh management to the household setting and view these parents you know as managers you know  
[04:18:24] and in doing so you know this paper introducing you know introduces parental managerial capital as an overlooked uh input in human capital you know uh formation and not not only that this paper also suggests that uh those  
[04:18:39] differences that we see in terms of uh you So a learning gap or achievement gap between you know across social economic status could also be driven by the fact that household are different in their ability to manage you know their  
[04:18:52] children time and effort. And the second uh contribution that I make is in this big literature on information friction in education. There's been a lot of work that focuses on providing parents with you know academics you know information  
[04:19:06] about the ch you know children academic standing to parents or how the child is behaving at school and and and things like that. And if you take a step back these information that you know we focus you know on providing to parents are  
[04:19:20] information that actually readily available. You just need some time to talk to your child to get those information. So that's a kind of you know the stance that this paper is taking which is maybe the fact that parents do not have that information is a is because parent are not even playing  
[04:19:34] the managerial role in their children education and by you know empowering parent to uh play that role they end up getting information as you know uh a product of just you know being engaged in their children education. And the  
[04:19:48] second you know uh contribution uh in this literature is u you know basically you know bringing you know uh drawing our attention on the fact that you know we focused you know uh a lot on you know providing this information about  
[04:20:02] children academic you know performance to parents but maybe we need to also you know uh look into parent ability to really you know play a role in their children education you know because there have been some work that show that  
[04:20:15] for instance when you provide uh performance information to parents they might reduce investment in their children and uh by the way I'm I'm happy that maybe I wasn't part of some of those intervention because maybe my parents would have reduced investment in me and I wouldn't be here today right  
[04:20:30] you know so uh those are the few contribution that you know uh the paper make so this is the outline of you know uh what I'll be showing you we discuss the context the research design timeline  
[04:20:43] and then uh present the data that I collect and some uh baseline and uh motivating uh fact and then I will discuss the intervention and uh conclude with uh the result.  
[04:20:55] So why you know um should we you know can we think about you know parental u uh uh management being uh critical. So this is in a setting where you know uh you know uh student uh in eighth grade  
[04:21:09] takes uh six subjects and each of these subject are very important uh in uh defining you know you know uh in computing your GPA and also making grade completion uh decision. But these tests  
[04:21:23] happen throughout the you know uh academic you know year. And these are settings for those who work on education. We know that when these children go to school, most of the learning, you know, happen outside the school. You go, you get the material and  
[04:21:36] you go, you need to, you know, review all the material that you got from, uh, school. So if you go you know once you are outside the school and you are you know at home if you do not uh really spend time with the material there's no way you will learn you know anything and  
[04:21:50] be able to do you know uh well and that's exactly when parents are important to create an environment conducive uh uh uh to to to learning and interestingly there is no parent school communication I would say very limited  
[04:22:05] parent school communication uh uh if your child is not doing well or they didn't show up to school nobody will inform you nobody would call you the child can leave the house and say I'm going to school but do not go to school you won't even know uh about it and these parents you know have little to no  
[04:22:19] education so there is no scope for them to even say they will provide direct you know instructional support or even hire a private tutor to help the parent so if any margin you know exists for this parent to really you know play role uh  
[04:22:33] it can be along the dimension of you know creating an environment where children uh could really uh learn.  
[04:22:40] So that's a little bit uh the context. So here is uh the study uh design. So I randomly selected 40 schools uh from uh this uh region of Benin. And then I randomly allocate this school into what  
[04:22:53] uh I call intervention schools and pure control schools. And everything that I will be you know most of the result that I'll be showing you today will come from the intervention school. So within intervention school what we did you know is to invite parents to a meeting at the  
[04:23:07] school and then we enrolled around 55% uh of them. The other 45% you know uh a large majority uh was not uh contacted but there were also uh a couple that we contacted and couldn't make it uh for uh  
[04:23:21] various uh reason. So once parents came to the meeting and took part of took part in the baseline uh survey uh data we then within each classroom we then  
[04:23:34] create literacy strata because I uh uh collected information on parental literacy during the baseline and then within each classroom we create a literacy strata and then randomly uh assign uh parents into treatment and  
[04:23:48] control uh within each of those classroom. So my treatment and my control will all be student that are in the same classroom. They have the same teachers, same peers, everything. The only difference will be that the treated will receive the parents will receive  
[04:24:02] this information intervention and the control uh won't uh uh receive it. Yes, question.  
[04:24:09] >> Um so of course parents are busy and chasing after kids and how they're spending their time. It's a freaking pain. So is is this a I mean how should we think about this intervention? Is there a cost that's easily measurable or  
[04:24:21] is it pretty costless? Um is there that just getting them to use their time more efficiently or you know >> so we'll get to that. So if I go back to the example I gave at the beginning of you know my friends growing up when every time that we leave the game and  
[04:24:36] ruin the fun that we were having the parents were not around but the parents had a structure that the children knew that at 7:00 p.m. at whatever parent need to go home to be sitting and studying and in this setting which is  
[04:24:48] rural area the parent they go to work or you know whatever and they come back they're at home so the as you will see in the intervention uh these are not things that the the the father need to be there necessarily for some of these thing to be going on it's you know you I  
[04:25:03] will show you in a couple of slide that it's really uh uh light touch in terms of you know what it require parent to do but there are also some component of the intervention in terms of the action that parent can take some of them are more costly. So in the intervention I will  
[04:25:17] see I have three type of action that we encourage parent to take and those three type vary by how expensive quote unquote it is you know to implement you know.  
[04:25:25] Thank you. So uh in the period control school what I do is there is no meeting no calls nothing but basically given uh that my treated and controls you know the variation comes from within classroom uh  
[04:25:40] we might be you know concerned about uh spillover uh uh uh on those uh uh control uh uh student. So this is uh the timeline of uh the study. Uh the  
[04:25:54] academic year start from September to uh May and is divided into two. So the first semester is from September to uh February. And what happened is that these student take uh two tests per  
[04:26:06] semester in all of the six subject that I you know that I just uh mentioned. So right after the first test we went and collected our baseline you know uh survey data and a couple of days before the second test of the first semester we  
[04:26:21] started the intervention and then the intervention uh uh continue until the end of uh the school year. So these different tests that happen after the intervention uh started are you know where we expect you know performance you  
[04:26:34] know any change in performance we that's where we expect you know uh the channel for which we expect you know to see any change uh in in performance.  
[04:26:44] So uh I will in you know briefly show you the different data that we collected and then introduce uh some key object that are important in uh what I will be showing you here. So in terms of data we collected uh in interventional school we  
[04:26:58] collected parent survey data at the baseline and also at the end line and we also conducted student survey at the end line basically because some of the outcome we'll be looking at are self-reported outcome from parents. So at the end line we wanted to also see if  
[04:27:12] when you ask the children directly do you see the same pattern in the responses that uh you get and then we collected also school administrative uh data this subject that I showed you uh these different tests that I showed you we collect data on that and we collect  
[04:27:27] data on student absentes uh student uh GPA and grade completion and then a year after the intervention uh we went back and collect data on uh enrollment in the pure control school to be able to look  
[04:27:39] at some speed over what we do is to just collect data on annual GPA and grade completion. I didn't have enough money to you know collect detailed data in those school but I hope those two outcome you know will be enough to  
[04:27:52] convince you uh that you know uh you know uh the result we have is not coming uh from negative spillover.  
[04:27:59] So this is what my sample look like. These are household with six members you know on average and as you can see here the average years of schooling for the mother is around two and for the father is around uh five. So the children are  
[04:28:13] already more educated than them. The children are you know uh in eighth grade and half of the uh survey respondent were illiterate. You know we as uh gave them the basic literacy a simple text to  
[04:28:25] see if they can uh read it. And 97% of my sample uh have a mobile phone, a basic phone. And that's you know how we'll be providing the intervention. And the other three person who do not have a mobile phone, there is someone in the  
[04:28:39] household beside the father or the mother you know with a mobile phone.  
[04:28:44] So here you know uh based on uh the focus different focus group that uh we uh conducted we defined you know uh four  
[04:28:56] very specific domain where you know uh that you know school official and teacher and so on think you know are very critical you know domain where parents should be taking actions and what we do is that for each of these  
[04:29:09] domain at a baseline we try to see if parents believe that that's their responsibility to be doing that. So and that's what we measure by role endorsement. So we ask parents whose responsibility it is for instance to ensure that student attend school  
[04:29:24] regularly. So they can choose different you know multiple you know answers as long as you know one of the option that they select is parent themsel then we define that as parent endorsing that role. So we do that for these four domains. So the first one is ensuring  
[04:29:38] the child attend school. The second is checking the child you know academic progress and the third is ensuring that the child when the child is at home the child review you know uh their lesson and then uh the last one is ensuring that the ch the child notebook is uh up  
[04:29:53] to date. Why? Because if the notebook they some of these children go to school and they don't take note if you don't take note there's nothing you are going to review and if you don't review you can't do well in the test and you're going to repeat the grade. So that's the logic behind uh these four key domain  
[04:30:07] that uh uh we have here and then we collect data also on action implementation for each of these domain you know after you have endorsed you know this you know uh domain and you said okay it's your responsibility to do it the next thing is to uh look into it  
[04:30:22] and see if you did take an action toward that so with those two we can measure now what I call the role action gap and the role action gap here basically is for every domain that you endorse what are the you know uh the action the domain where you did not take the  
[04:30:36] corresponding action and this is going to be a very uh important uh you know object in uh the following so let's take the case of those four uh domain you know here what I'm basically showing you I'm showing you you know the you know  
[04:30:51] how you know the role endorsement question and the implementation uh action question are kind of uh related so here I'm now showing the result of you know the role uh uh uh action gap.  
[04:31:04] So when it comes to attendance as we can see for all of these you know uh domain we can see that uh more than 90% of parents endors this as being something that they should be doing. But when it comes to the action as you can see we  
[04:31:16] you know uh except you know checking the progress where we see a 63% you know in terms of you know 63% of parent uh taking action on that you know all the other one you will see very low you know uh action implementation. So this  
[04:31:31] clearly show that there is a large you know role action uh gap you know uh uh being present here. Yes. Question.  
[04:31:41] >> The action is the specific action that you prescribe rather than some other I think attendance to school. You could say ask other parents to go visit instead of you, right? But you wouldn't count that.  
[04:31:52] >> No, we count that. >> Okay. Yeah.  
[04:31:54] >> What's the domain of that? If I ask other kids that my child went to school, that >> So, yeah. So, in the survey, what we did is that we clearly restrict, you know, the kind of action that uh would count toward a specific domain. So for  
[04:32:07] instance in for attendance if you say that I asked my cousin to go and check yes we we are not asking if you exactly you went and uh took the action. No but if you say I asked other kids we wouldn't count that because  
[04:32:20] children themselves they you know Yeah. Exactly. Exactly. Yes. Sep.  
[04:32:27] >> How much of this is sort of institutional versus parental action?  
[04:32:33] So, I'm thinking of like if my kid doesn't go to school, I get a text message from the school um or and I get an attendance report and the report cards need to be sort of viewed and signed and sent back in the Thursday  
[04:32:46] folder. Um so, is that like h how do you want us to think about >> that's not happening here? Yeah.  
[04:32:53] >> Yeah. [laughter] >> Definitely.  
[04:32:56] So, that's not happening here. Like, you know, uh you can miss school for whatever days your parents won't be informed. You can be doing very well and suddenly start doing very poorly. Your parents would never be informed. They would never be contacted and things like that you know. So it's quite fascinating  
[04:33:11] you know when I was going to school my parents never told me go and sit down and study you know. So it's you know this is the kind of environment where we are working. Yeah.  
[04:33:20] >> To mention the the the weekly phone calls you said.  
[04:33:24] >> Yeah. Right. They do like you ask whose responsibility to check attendance and did you did you make sure the kid goes to school? Is that what you do? And is that the intervention?  
[04:33:35] >> No, that's not the intervention. We >> What is the intervention?  
[04:33:38] >> So here we are I'm showing you uh the role action gap because this is you know a key object that we define that we determine how we design the intervention you know. So that's uh what I'm doing here. I haven't you know presented you know  
[04:33:52] >> my question wasn't that I I I assume that this is not happening here. Uh but the the question was more like is this you know is this something that could be solved with a different kind of intervention where schools take on this  
[04:34:06] responsibility rather than you uh calling up the parents or the parents.  
[04:34:11] You see what I mean? Yeah, that's definitely possible like any institution that can be set up to ensure that you know uh parents know understand you know the managerial role and implement this role you know any such institution will  
[04:34:25] also uh do the work but basically what the paper is trying to show is to draw our attention on the fact that we think about you know financial support you know we have intervention where we encourage parent to help their children read those are things that are not feasible to parent to most parents at  
[04:34:39] least in these settings and there is a dimension that's visible to them which is managing the children time and effort. Yeah. But how we get there might you know vary. So the next thing I do is that I take this role action gap and I try to you know uh you know look at it  
[04:34:53] from you know based on different characteristic in the sample. And as you can see here there's not much difference you know uh based on the gender of the child and not much difference based on uh baseline performance of the student but we can see a big difference uh based  
[04:35:07] on parental uh literacy. So the next thing uh that I will uh try to do now is going back to the question the first question that I received which is that some of these action that I mention mentioning here differ in terms of the  
[04:35:20] implementation cost you know so let's look at the four you know domain that I just uh uh uh presented and we will see that you know the we create three categories the first is what we call conversation the second is verification  
[04:35:33] the third is higher friction you know action so for things such as just asking your child about the test score. This is uh something that goes into the first uh category where we see that those are the low hanging fruit that parents can implement. And then the second category  
[04:35:48] are things that still happen within the household but could be a little bit costly because it requires some verification from the parent. And the most you know expensive quote unquote set of actions are things such as you know setting a study routine and visiting the school for instance. These  
[04:36:03] are action where you know parents need to coordinate you know and sustain the effort that they are doing toward that or even they need to leave the house and go to the school you know. So let's look at how the role action gap vary based on  
[04:36:16] this different level of cost and you as you can see you know the more you know expensive quote unquote you know the actions are the uh the bigger the role action gap. So this support you know  
[04:36:30] what we've learned from the field which is you know these parents are not you know yes they endorse this rule but they are not taking action because first it's very costly for some of them to even get started and second some of them do not really understand how important or  
[04:36:44] critical these things are in the children you know performance although they endorse the role and that's you know what we will use to try to you know that's what we have in the back of our mind when we are designing uh the intervention. So what's the intervention  
[04:36:59] uh about? So it's this structured uh phone based uh coaching sessions over four months and we have standardized script uh that facilitators uh needed uh to uh to follow and you know the two  
[04:37:13] friction that I just mentioned you know are kind of in the back of our mind you know uh you know when we were designing this so we have a you know a three-step uh you know approach in the first step we clarify the why meaning you know bringing this parent to really  
[04:37:27] understand why are this thing important and If you don't do it basically you know nobody's going to force your child to read or to do anything they won't do when you know the teacher won't be at home with you to you know ask your child to sit down and read and things like so parent needed to be really convinced and  
[04:37:42] understand that this is something that's on them you know and then once we do that we help them set you know achievable goal and achievable goal here is things uh related to you know bringing parent to understand that it's not you per se that need to be there and  
[04:37:55] ensuring that they are doing it but have the structure if there's a cousin there's an called just at 6 p.m. or whatever you choose. You want you know to ensure that your child is seated for instance and reviewing. And in the second step what we do is that we try to  
[04:38:08] build accountability in way that uh fit uh the household. And finally we discuss setback and trouble uh uh uh should any challenge that these parents are facing.  
[04:38:17] And as you can see here with the characteristic of this parent here we can't do the typical phone based text messages and things like that. you know these parents have little to no education and on top of that you know they face so many challenges when they try to do these little things for  
[04:38:32] instance you tell them to go to the school you know to encourage them to go to the school they will go to the school but then they will be like I got there I don't know what who to talk to and things like so they need a sounding board of you know of someone that they can talk to when they are going through the challenge so what are the action  
[04:38:47] that we encourage them to take so again the three categories so when it come to conversation we encourage parent to just ask the chat about how was your day at school today, you know, basic and also encourage them to ask about test scores  
[04:39:00] and encourage the parent to ask about you know upcoming you know exam day.  
[04:39:04] These are things that regardless of when you come home, you can do as a parent.  
[04:39:08] And these are also things that I'm sure that most of us here who you know have kids, we we do that and we maybe take it for granted. But if there is any good news at the end of this presentation is that you know the parents who are here who are doing that you know you all  
[04:39:21] deserve the best parent award you know. [laughter] So the second category is verification and basically uh what we encourage parent to do here is to you know check homework and what we mean here by check homework the parent can't read can't  
[04:39:34] know anything but just asking your child you have homework do it and come and show me but the parent the kid show you you don't know what they did but at least you just you know encourage them uh to do that and uh for notebooks we encouraging parent to check you know the  
[04:39:47] children notebooks and this might sound very funny but when the children don't take note what they do is that they leave empty pages. So these are things that parent despite the regardless of their literacy they can catch if they can you know scroll through the notebook  
[04:40:02] and see if there are many empty pages you know left in there. And then we also encourage parent to check on the child when the child is studying to just ensure that okay you are seated you are studying you are indeed you know doing that and not playing. And for the higher  
[04:40:15] friction you know actions we encourage parent to set a study routine. Most of this household the child will study whenever they want or even if they want you know they might not study you know uh so we encourage parent to do that and to follow through and we also encourage  
[04:40:29] parent to adjust chores around exam uh date and that's why it was important to initiate conversation to know the exam date you know uh to start off with and then we encourage parent to visit schools and when they visit the school there are three things they can do one  
[04:40:42] is to you know uh check attendance to make sure that okay the child left the house is my child indeed at school and the second is to engage in conversation with school official or any teacher you know uh to ask about you know uh the the the chat.  
[04:40:57] So this is a picture that shows you the kind of things that I'm talking about you know for many people that might you know you know you might not relate but you know this is uh the typical you know uh household in you know uh a developing country in the rural area. So basically  
[04:41:12] here you see a picture where the mother is cooking but she set a study uh routine the child is seated and she has an eye on that child while you know cooking that's you know uh is already enough at least the way I think about it  
[04:41:25] you know here in my intervention and if I go back to what I was saying about you know my friends when we were growing up uh if I if I go back you know uh okay question yeah yeah so if I go back  
[04:41:40] to that when my friends you would leave us and go and you know study or whatever you know we were always praying that our parents would never you know know about what their parents were doing so that you know they won't constrain us and things like that so just getting parent  
[04:41:54] to take these actions you know are things that can change uh something yes >> I might have missed this but so is this with one parent or uh both of them because I guess I'm wondering if part of what this is doing is solving a  
[04:42:08] freewriting problem where if you you know each person says they 90% they're it's their role. Uh but no one did it.  
[04:42:16] So >> this is with one parent.  
[04:42:18] >> Yeah. And so could part of what's going on is just like you know solving this coordination problem of who's actually going to do the action.  
[04:42:26] >> Yeah. So we don't constrain we don't set any constraint in terms of who will take the action. So you know in in a household let's say it's a father who is the person who is receiving intervention. the father can talk to you know the mom that these are the things  
[04:42:41] that you know we are I'm learning now let's you know uh create this routine and if I'm not around just ensure that you know uh the the kids you know follow it so that's a little bit you know if I understand your question >> yeah and I guess what I'm saying is that  
[04:42:54] you know your sort of theory of what this intervention did was it gave people the concrete steps to turn what was their role into actions and I'm >> you know like I guess an alternative is both parents thought this was their  
[04:43:07] role. There was a freewriting problem and by intervening with one parent and like having these conversations, they became the natural person who was responsible for doing it. So, it was like less about the content and more it  
[04:43:20] was uh um yeah, coordination device. >> Yeah, that's a very good question. So I uh we don't I can't really tell you because we didn't collect data from both parents but when we're collecting the data and asking if there is someone who  
[04:43:34] check on the child progress it's not that specific person in front of us like as a household you know is it a you know do you check your child you know progress and things like that so I think we might have captured some of that we are not saying the specific person in  
[04:43:47] front of us did you take this action if you know that your you know your spouse took the action and things like that you will still answer that yes we do that for our child. Yeah, but that's a very uh uh uh good question. So, it's  
[04:43:59] possible that that's going on. So, here is uh you know uh the take up rate. So, parents receive on average around 10 phone calls and that's like a 88% you know uh success rate and this is  
[04:44:13] balanced across you know uh the you know uh different characteristic uh in the sample and this uh parent you know the average length of the phone call is around 5.5 minutes. So in total it's less than an hour in terms of uh conversation and discussion that uh we  
[04:44:28] had with our parents. So let's uh look at the result. But uh what I run is you know pretty straightforward uh uh where you know we regret the outcome on our treatment indicator and then control for  
[04:44:42] baseline outcome whenever possible and then I include this strata fix effect that we had in there and whenever the outcome that we are looking at is an outcome where we have repeated measure for you know the student in those case  
[04:44:55] we cluster the stand at individual uh level. Okay. Question >> quickly. Do you also collect that information for the siblings to see if there are spillovers on siblings?  
[04:45:05] >> No. So, you know, uh I didn't have that was something that was in my I didn't have enough funding to, you know, collect detailed performance data on the siblings and and things like that. Yeah.  
[04:45:17] >> Yes. >> I guess this is related to question about costs. Uh but it's it sounds to me like one of these context where you're providing some inputs and that could be complimentary to my my effort so for this period happy to go  
[04:45:32] along getting these calls and helps me organize but I'd be really curious to see what happens once you stop the intervention but they learned that now they're you know they saw that it did have positive after >> yeah that's a very good question so I  
[04:45:46] have some data on parents the action that parent take a year after the intervention and I see some uh persistence and not only that usually when you know at least for typical performance information that we provide to parents when parents learn that their  
[04:46:00] children are not doing well they they tend to reduce investment and those children will drop out which is the opposite that I find here a year after the intervention we will see that dropout decreases for you know parents uh children whose parents are illiterate and also for girls you know so that kind  
[04:46:15] of maybe suggests that you know there is some form of learning that happened during this process and although some of these you know student are low performing parents now understand how they can support and they maintain their  
[04:46:26] children you know in school. Yeah. So I was wondering whether you have any  
[04:46:41] data on that something where like you know you be like learning something that you know your parents check  
[04:46:59] >> I'm not sure I really understand the question. Yeah. Could you please reframe the question >> and maybe tomorrow  
[04:47:21] but a child would always start.  
[04:47:27] So I'm wondering what happens once you get the what the child or something.  
[04:47:38] >> Oh, I see. I do not have data on that. I do not have data on you know when you know parents you know start monitoring or you know taking this action how do children you know uh actions adjust to that. I do not have uh data on that but  
[04:47:53] as you said if you know children are children you know if you left you leave them alone you know the effort that we put in is minimal and it's not because they did poorly today that tomorrow they will be like I need to study hard study hard and that's why great repetition as  
[04:48:08] we will see here it's high you know yeah definitely yeah that's a very good uh question so let's look at some of the results so the very first result that we'll be looking at is how the  
[04:48:21] intervention shrink the role action gap and role endorsement was almost universal at the baseline. So at the end line I don't collect data on that. I just collect you know action implementation you know information and then I use that to uh look at how role  
[04:48:34] action gap you know uh behave. So here as we can see if we focus on uh uh the four you know domain that I showed you earlier as we can see here for conversation the role action gap uh decreases by 9 percentage uh uh point  
[04:48:48] and for verification 11 percentage point and for higher friction uh 6 percentage point and as you can see in the control group you know based on how difficult you know the the the action is the gap is still high you know in the control uh  
[04:49:02] group and overall we can see that there is an eight you know uh percentage decrease in uh the role you know action gap. So as I said at the end line we do not collect role endorsement data. We already have that from the baseline. So any change that we see here is coming  
[04:49:15] from parent and you know taking actions and that's what I try to show here. So for the four you know action that we you know and you know that I showed you earlier we can see that for conversational you know uh uh there's  
[04:49:28] increase of you know 11 percentage point you know uh for conversation and 12 percentage point for verification and 6 percentage point for higher you know uh friction and uh when I put you know all  
[04:49:41] these you know action together you know and look at the simple average of them you can see that there's a 9 percentage point increase you know in the likelihood of parent uh uh you know uh taking uh these actions and then in partner B what I have here is showing you the full set of action that we  
[04:49:56] actually encourage parent to take you know as part of the intervention not just the four where we have role endorsement uh data and as you can see the trend uh is uh quite uh uh similar you know in terms of the result yeah question  
[04:50:10] >> related to my previous question here you're asking about that specific child if they have three four children is this as for each ch child or just for the one just for the one >> yeah just for the one that we are following here. Yeah, thank you very much. So, uh the next thing that I uh  
[04:50:25] look at here is now uh the recommendation we give parents around chores you know. So in column one here what we are measuring is has it happened that you know the child was asked not to do a job that they usually do and take that time and go and study. That was  
[04:50:38] basically the question and we asked the question to both parents and student and as you can see that probability increase you know the probability that you know parents are student are being asked not to do a job that they usually do and to go and study and then if they're not  
[04:50:52] doing it there is someone else in the household doing it. So obviously we can see here that parents are doing more and also siblings are also chipping chipping in you know. Yes question around  
[04:51:08] >> exam time. Yeah >> it's kind of interesting but this seems like a more measure. So I'm just kind of curious how you think about kind of like this that kind of specific advice translating into this about just kind of the importance of education more generally.  
[04:51:22] Yeah, that's that's that's a very good question. Uh so defin you know there is it's possible here that parents you know uh learn you know uh maybe that you know  
[04:51:35] chores might be uh conflicting with education not only around exam time and I have another project you know uh that we go in the field that kind of really focus on this idea of you know household chores dynamic within uh the household.  
[04:51:48] So that's a very good question. So, everything that I've shown you so far are self-reported. Yes.  
[04:51:55] >> Just as a followup to that, like I um >> yeah, I guess like the way the intervention was framed was about uh you know, you want your child to be studying and you're just not doing being  
[04:52:09] proactive or doing the right things. Uh but it's it seems like people might have just heard education is important, right? And like because this >> but their children are already in school, right?  
[04:52:20] So I feel like maybe if the children in school and school is not free, they kind of know that it is important.  
[04:52:30] I guess this just has much more of a trade-off than the misallocate, you know, like there's the you're not on the production possibilities frontier and we're going to get you there versus there are some trade-offs and we're the behavior change was just shifting like  
[04:52:45] along that chore, you know, do I want to you have my child do chores or study and yeah, you know, >> that's a very good question. That's that's an a very good question at least uh based on my personal experience and you know uh what I know from uh the  
[04:52:59] setting and as I said my own parents are illiterate it's not like they don't care about education or they don't know the value of education you know they will invest like crazy amount of money in education but on they do those financial  
[04:53:12] investment it's like this like they don't really because they haven't been through the process they don't really understand things that you and I take for granted you know that's like a a little bit the way you know and understand the whole problem you know.  
[04:53:26] So based on that, yeah. >> Yeah. Suppose like there's an intervention that taught parents that time invest in a child studying is important for education. Like they weren't they were thinking about their financial investments but not the child's time. And like you would get  
[04:53:39] this kind of result from that. And that feels different to me than uh like you knew that studying was important, but you weren't doing the like you weren't kind of executing on that plan of how to get them to study um as well. So anyway, it just feels  
[04:53:54] different to me to to like convince them that having kids study is more valuable than you thought. And I don't know if that I guess I don't know if that's a friendly amendment like you're happy with that interpretation that that's what parents got away from the  
[04:54:07] intervention or that feels like no no no that that's not management skills that's um you know changing their beliefs about >> yeah that's that's a very good good question. I I I'll you know keep thinking about it and then we can chat  
[04:54:20] you know after uh the talk. Yeah, siblings education siblings are also doing more.  
[04:54:29] So it be good to know if those siblings are also in school and is this substitution across children in which case it would be consistent with this.  
[04:54:39] >> Yeah. So definitely uh you know I don't think parents you know what we provide here as information is something that parents would see oh this is important for child one and this is not important for child two. That's not how I see it.  
[04:54:51] But unfortunately you know I didn't have enough funding to collect you know those detailed data of other you know children within the household their performance and all that I didn't have you know uh funding for that. So uh everything that  
[04:55:05] I showed so far one second everything that I showed so far were based on self-reported outcome but if indeed parents are taking this action we you know are expected to see some changes you know in some objective you know measures. So in column one here it's  
[04:55:18] student absentism and as we can see you know student absentism you know uh there is a a negative impact there where student absentism decreases and in column two and three what I try to measure there is information where  
[04:55:32] parent knowledge about their children academic standing and in column two what we do is that we ask parents what's your child GP but given their literacy level most of them wouldn't have that information that's why you see that very small sample size in the third column we  
[04:55:46] rather Ask parents whether you think your child GPA is in the pass or fail you know category and there as you can see you know we can see that parents knowledge about their children academic standing has improved you know which goes back to what I was saying which is  
[04:55:59] that maybe this basic information that parents can get from their children and they do not have you know is you know uh comes from the fact that they are not you know playing this role to start with you know so now let's look at uh the  
[04:56:12] impact on student you know performance. So in the first column what I'm showing you here is grade completion and as you can see uh there's an increase of four percentage you know point in grade completion. So around 32% of student  
[04:56:25] here will repeat grade 8. And in the second column I show you year end GPA we see a 0.11 standard deviation increase.  
[04:56:33] And then you know given the timing of the intervention as we can expect in the first semester we are not able to pick up something significant. In the second semester we see that there's a 0.12 you know standard deviation increase but when I compare first to second you know I'm not able to reject you know  
[04:56:48] equality. So the next thing that I do is that you know I look at the subject because that's what you know where the changes need to happen so that we see those changes that I just show you in per in GP and so on. So as we can see they take six subject and where the  
[04:57:03] subject where we see an impact are things such as physics, biology and you know uh borderline you know uh history and uh geography and I try to do pair wise comparison and all that and these result are statistically uh uh different  
[04:57:18] from you know the subject where we do not see uh any uh effect and this you know test score that I I showed you I just showed you are you know the three tests that happened after the intervention started. So I try to also boil that down in terms of first  
[04:57:31] semester and second semester. And as you can see the overall pattern that I just showed you is present in the second semester but not so much in the first semester. But interestingly although the intervention started a few days before that only case of the first semester we  
[04:57:45] do see pick up something there uh in physics. Yes David.  
[04:57:49] >> Yeah I was did you try to break it down by gender of the kid and does it also gender of the parents? Yeah. So you know uh you know I I do that and I do not see in terms of performance I do not see any  
[04:58:03] difference by by gender h but I did not interact that with the gender of the parent but one of the result I look at when it comes to parent is to see the gender of who received the intervention and you would see that when mothers receive the intervention the effect is  
[04:58:16] really big but you know that's not you know really causal because you know uh we didn't assign. Yeah. Yeah.  
[04:58:34] So I do not have data on that. That's a very good question and I think that's something that I need to collect for uh the next project on gender on household chores. Yeah.  
[04:58:44] I I didn't record the phone conversation. Yeah. I didn't have you know authorization to to do that. Yeah.  
[04:58:50] But we receive anecdotal evidence where where there was some kind of conflict between you know like the father will give the instructions before going to work come back the mother would wouldn't follow you know those kind of you know  
[04:59:04] uh those instruction that the father you know uh left before leaving you know so next what I do is try to ask myself are performance gains larger when parents have more room to improve you know and you know the take away from you know uh that result is that you know we don't  
[04:59:19] see really uh uh something going on based on you know uh you know whether you know uh when parents have more room to improve. So overall what the result that we see are driven by you know uh almost everyone you know uh in the  
[04:59:32] sample. So the next thing that I want to show you before I run out of time. Yes.  
[04:59:42] >> How much parents are doing this for their kids as a mechanism of like intergenerational systems right of poverty or whatever. I was trying to think about like whether you say anything that like if you could estimate  
[04:59:55] the cross-sectional relationship between you know parential education and or poverty levels or something and these kinds of crossction and then you know think about like an ID changes that like  
[05:00:08] something about sort of the degree to which this mechanism of like parental involvement is playing >> I think you just gave me an idea for another paper you know [laughter]  
[05:00:22] >> that's that's a very interesting one yeah I will look into that I will look into that yeah thank you very much for you know that idea yeah maybe we'll be quoted on that so next what I'm showing showing you here before I run out of time is to just  
[05:00:35] you know look at some iterogenity. So basically I look at across three dimension. The first is by student baseline performance and as we can see the grade completion you know result is mainly driven by student who are at the bottom quartile of you know uh  
[05:00:50] performance and exactly those are the student who are more likely to repeat because in that subsample 70% of them repeat the grade you know. So great repetition here is uh very high stake uh you know great competition is very high stake and then when we look at those two  
[05:01:03] other subgroup we don't see much and I do the same thing by parent literacy and as we can see here you know we don't see much going on and same thing for you know uh student uh gender so the next thing that I you know want to discuss in  
[05:01:16] my few few last minutes is drop out a year after the intervention. So overall we do not see you know anything going on but when we look at it originity by parent uh literacy and student gender we can see that uh dropout decreases for student whose parent are illiterate and  
[05:01:30] also uh for girls. So this goes back to the question of you know uh you know whether there is some persistent or not because this is a year after the intervention you know so this may be potentially suggest that something is going on and some other result that I'm  
[05:01:44] not showing today is you know uh result based on you know the action that parent take you know continue to take a year after the intervention and there I do see that there is uh persistent so to conclude uh parents uh largely endorse  
[05:01:59] responsibility you know for their children you know education but often fail to implement the corresponding action. I design you know this low cost intervention and the intervention is able to reduce uh uh that role action  
[05:02:12] gap and the intervention is also uh al also help at improving uh uh student uh performance. So the takeaway if anything is that parental managerial capital is a consequential you know malleable input  
[05:02:26] into human uh capital formation and I think the best way to uh finish a presentation that's talking about management practices is to make sure that I finish on time. So thank you very much. Yeah.