ABCDE 2024—Day 2—Session 3: Economic Inclusion of Youth

videoType

dynamic-media

videoDmUrl

keyFrameImage

timestamp

00:06 Welcome to your last session of the day.

00:09 In fact,

00:10 the last session of the conference,

00:12 and it's going to be a good one.

00:14 Today was meant to focus on the economic inclusion

00:18 of women and youth.

00:20 You will have noticed that the 1st 3 panels we had

00:24 all talked about women,

00:26 clearly close to my heart,

00:28 but this is the panel that's gonna be talking about the economic inclusion

00:32 of youth.

00:33 So please stay tuned.

00:35 My name is Tamar Manuel Yan Atten.

00:38 I'm a senior non-resident fellow at the Brookings Institution

00:42 after a long career at the World Bank,

00:46 um,

00:47 and we have a wonderful group of

00:50 presenters here who,

00:51 I,

00:51 as I said,

00:52 are gonna be talking about

00:54 the topic of economic inclusion of youth.

00:57 Now

00:58 I don't really think this topic needs much of a motivation.

01:03 But I'll give you one in any case,

01:06 uh,

01:06 I believe you all know that youth employment

01:10 is a challenging issue everywhere around the world

01:14 and it is particularly acute in some regions of the world,

01:18 Middle East and sub-Saharan Africa in particular.

01:21 According to World Bank estimates,

01:24 the working age population

01:26 in sub-Saharan Africa

01:28 will increase by 740 million

01:31 by 2050,

01:33 more than doubling its current size,

01:36 which is at 630 million

01:39 today.

01:40 If unemployment and underemployment in general

01:45 and for youth are a problem now,

01:49 they will be an even greater drag on growth,

01:53 source of frustration,

01:55 and a source of instability in the future.

01:58 Rachel touched upon that point this morning.

02:01 What programs may help foster

02:04 the economic inclusion of youth

02:07 is the topic of this panel.

02:09 So let me introduce the speakers who are all gonna be talking,

02:13 presenting papers that all are in the African continent,

02:18 um,

02:19 appropriately.

02:20 Our first speaker is Oyebola Okunogbe.

02:24 She's an economist in the Development Research Group of the World Bank.

02:27 She was born and raised in Nigeria.

02:30 Next will be Andy Zeitlin.

02:32 He is an associate professor at the McCourt

02:35 School of Public Policy at Georgetown University.

02:38 He was born and raised in Stow,

02:39 Massachusetts.

02:40 I don't think I know where Stow,

02:42 Massachusetts is.

02:44 He's also a non-resident fellow at the Center for Global Development.

02:48 Next will be Isaac Mbiti,

02:51 who's an associate professor of public policy and economics at

02:54 the Frank Batten School of Leadership at Public Policy.

02:57 Prior to this,

02:58 he was an assistant professor of economics at Southern Methodist University

03:02 and a Martin Luther King visiting assistant professor at the,

03:05 at MIT.

03:07 He's a JPA bred MBR and ISA affiliate.

03:10 He was born and raised in

03:11 Kenya.

03:13 And our final speaker will be Owen Ozi who's an

03:16 associate professor in the department of economics at Williams College.

03:20 Um,

03:21 you will remember one of our previous speakers is also Williams College.

03:25 It made me

03:26 wonder about the

03:27 in-household bargaining that took place for that outcome to happen,

03:31 but we'll leave that aside.

03:33 So Owen was previously a senior economist

03:36 in the World Bank's development research group.

03:38 He's a JPal Breed and ISA affiliate.

03:41 So

03:42 why don't we go ahead and start?

03:43 Oyebola?

03:44 Great,

03:45 thank you so much for the introduction and thank you

03:47 to the organizers for inviting me to this great event

03:50 and thank you to you all for staying.

03:52 1 2nd to get the

03:54 amount of time you have straight.

03:56 I was told 12 minutes and I think so were all the presenters,

04:00 but the last session got 14,

04:02 so I'm.

04:02 Going to ask for equity and treatment and ask for 14.

04:06 Thank you.

04:07 Great,

04:08 yeah,

04:08 I'm saying thank you to you all for staying till the glorious end.

04:12 I'm looking forward to all the discussion we'll have.

04:15 So today I'm going to be talking to you about a livelihood program in Nigeria,

04:20 uh,

04:21 one of the.

04:23 Comments we got in one of the earlier sessions on cash transfers was

04:28 what is it why you know

04:29 instead of teaching people how to

04:31 instead of giving them fish,

04:33 how about you teach them how to fish

04:34 and many governments apparently think in a similar way and many

04:38 there's a lot of policy interest

04:40 in helping people to graduate out of poverty.

04:43 So

04:44 many programs have a basic targeted cash transfer program

04:48 to provide consumption support

04:49 and then there's now um light layering on top of that a livelihood intervention

04:54 to help households develop a sustainable income generating activity.

04:59 Now

05:00 one can think about how this should be targeted within

05:03 the household so who in the household should be the recipients

05:06 and then how could one promote positive household dynamics

05:10 and this is where the focus on youth employment is going to

05:14 become apparent

05:14 as you'll see later on because

05:16 what we find in our studies is that

05:18 it ends up

05:20 once the households get the choice of who to target

05:23 young people in the household tend to be the ones that are selected.

05:27 Let me walk you through this.

05:29 So this is the Nigeria National Safe Social Safety Net Program.

05:34 We call it NASP,

05:35 and it's a very large,

05:38 uh,

05:38 program in Nigeria.

05:39 It's reaching over 2 million people now.

05:41 It started in 2017 and has grown rapidly.

05:44 There's much I can say about this program,

05:46 but I'm going to focus on a key element,

05:48 on some key elements.

05:49 And before I go further,

05:50 I'm just going to go back and

05:52 introduce my co-authors because one of them is in the room.

05:56 OK.

06:01 I was just excited to get into it.

06:02 So Kindi and Jay is here,

06:05 um,

06:05 Robin,

06:05 Tama,

06:06 Ayoilli,

06:07 Naira,

06:08 um,

06:08 are all at the World Bank.

06:10 OK,

06:11 let's continue.

06:13 OK,

06:13 so

06:14 in the NAS program,

06:16 the first thing to point out is that there is a basic package

06:19 where,

06:19 um,

06:20 households get 5000 naira,

06:22 uh,

06:22 for each month.

06:23 This is about equivalent of $12

06:26 and

06:27 depends on which exchange rate you use.

06:28 I know we're sticking with the

06:30 one from a few years back.

06:31 OK,

06:32 so

06:32 they also get um

06:34 savings group mobilization which is where um

06:37 the recipients in.

06:40 Communities,

06:40 they get a lot of training on

06:42 saving some of the the importance of saving

06:45 even some of this cash transfer so

06:46 that they can channel it into productive activities

06:49 and the recipient of this is the caregiver who is typically the senior woman in the

06:53 house.

06:54 So if it's a

06:56 male headed household,

06:57 which is the majority,

06:58 usually the wife of the household head

07:00 or um someone similar category,

07:02 and there's usually

07:03 an alternate who can receive the cash transfers if the caregiver is not available.

07:08 Um,

07:09 we worked with the government to develop a randomized,

07:13 um,

07:13 experiment to be able to

07:15 look at other components that are layered on top of this basic package.

07:18 So there was a top-up package

07:20 which,

07:21 in addition to these two elements I described also got a core responsibility.

07:25 Support

07:25 which involves training around things like early childhood

07:30 development,

07:30 nutrition,

07:32 sanitation,

07:33 and in for this the households get

07:35 an additional 5000 so they get double the amount

07:39 again here the recipient of all this training is a caregiver.

07:42 And there's a 3rd arm which is the livelihood package.

07:45 That

07:46 has

07:47 life skills training components,

07:49 business skills training,

07:50 mentoring and coaching,

07:51 and the goal of all of this is to help the recipients to think through

07:56 what they would do

07:57 when they eventually get the last component,

07:59 which is the live productive grant

08:01 so it involves things like

08:02 goal setting,

08:03 um,

08:04 budgeting,

08:05 management,

08:06 time management,

08:07 how do you develop a business plan.

08:09 How do you market your

08:11 products?

08:12 Basically,

08:12 all kinds of skills that they will need

08:14 along these three dimensions and the mentoring and coaching

08:16 also involve one on one component where you

08:18 actually walk people through developing a business plan

08:20 and what exactly they will do once they get this grant.

08:23 And the goal of this is that you didn't want people to get the grants and then

08:26 not have a plan in place

08:28 for all of this.

08:29 So,

08:30 all the work I'm going to be

08:31 describing today happens within this livelihood package.

08:34 We then um

08:36 among the group that are randomized to receive the livelihood package,

08:38 we

08:39 randomize some elements of the way it's delivered.

08:45 And the key questions that we're trying to answer,

08:46 number one is who within the household

08:48 should participate in the Livelihoods program.

08:50 The um.

08:52 In some programs,

08:53 the

08:54 default is the caregiver.

08:55 So this is the person that the government has

08:57 been investing in so far.

08:59 They've

08:59 been receiving the savings group training.

09:01 If there was a lively,

09:02 there was a

09:04 core responsibility training.

09:06 This person has been the one

09:07 that the government has really been engaging with.

09:09 And so it seems like the natural thing is to continue with giving it to them,

09:13 and many programs do this.

09:15 But we might think that

09:16 the qualities that we might be interested in for a caregiver.

09:21 Who's the recipient of the cash transfer may be

09:23 quite different from who we want to be the recipient

09:25 of the livelihoods training program

09:27 so for the caregiver maybe we want to be

09:29 the rationale like many programs is um you want to give it to the woman

09:35 in the house thinking about the

09:37 welfare of children

09:38 and

09:39 spending the finances on

09:41 household needs household consumption.

09:43 Or for the productive members where we care about the person being able to convert

09:47 this um investment into profitable activities

09:51 maybe we want to

09:52 have people who are more um business oriented

09:55 who have ideas who are creative or entrepreneurial

09:58 there might be other activities

10:00 qualities that we seek in them.

10:03 And now we can think that the households might also be

10:05 the people who have the best information about their members.

10:08 They know their characteristics.

10:10 They know their strengths and weaknesses

10:12 and so might think that they probably have the

10:13 best information on who should be the productive member.

10:17 But at the same time there's the risk of capture

10:20 if the household head just says,

10:21 OK,

10:21 then it's going to be me because I want the money to go to myself or

10:24 um or is making the decision on other criteria that might not be the most um

10:30 productive ones.

10:31 And so

10:32 we might think that it might be

10:33 important to also foster positive household dynamics

10:36 in the setting to ensure that

10:38 the most productive members chosen.

10:40 Let me

10:41 now tell you specifically what we do in this in the intervention.

10:45 So the first thing is we

10:47 want to examine the impact of allowing households the choice

10:52 of who

10:53 should be the productive member

10:54 who is going to receive the

10:56 train all the trainings and also receive the

10:59 productive grants once it's given.

11:01 And

11:02 the instructions that they're given

11:04 is that they

11:05 choose someone

11:06 who is

11:07 between that should be 1818 to 45 years old and able bodied.

11:12 And so this is the same instruction that's given in the

11:15 um.

11:17 So when it's the default arm

11:19 it's the caregiver

11:20 if they fit this criteria if they don't then

11:23 they um choose the alternate if they fit it

11:25 and then if they don't then they can move to anybody else in the household that fits it

11:28 and if it's in the treatment um they're

11:30 free to choose anyone that meets the criteria.

11:34 Second thing we

11:35 did

11:36 was to

11:37 do a household sensitization

11:39 to encourage collaborative decision making so we

11:42 had a facilitator go to the household

11:44 and to just help them to have a conversation

11:47 on the productive potential of all the household members,

11:50 so.

11:50 All

11:51 everyone,

11:52 you know what,

11:53 let's think about person A,

11:54 what's the education?

11:56 What's their experience,

11:57 what would they be good at?

11:58 What are they currently doing?

11:59 who was basically trying to get a profile of each person as a

12:02 way of helping the household to really be proactive and thinking about this.

12:06 We also had a video to illustrate

12:08 positive models of collaborative household decision making

12:11 and the goal of all of this was especially interacting

12:13 with the choice was to see if it will lead to

12:17 uh the choice of a more productive member.

12:20 So this is what the design looked like there were

12:23 2000.

12:26 Households

12:26 and then we had on top is the.

12:30 Um,

12:30 so the number one block,

12:31 uh,

12:32 the default,

12:33 um,

12:34 caregiver is selected

12:35 and there's no sensitization and then we have the sensitization arm,

12:39 the household choice arm with no sensitization,

12:41 and then

12:42 the combined.

12:43 I'm

12:45 Before going I just tell you a little bit more about the sample

12:48 for those of you who know Nigeria,

12:50 we have 6 states,

12:52 so these are in each of the 6 geopolitical zones,

12:55 and

12:56 we're in 12 rural LGAs.

12:58 They wanted to see

12:59 the government was particularly interested in seeing how

13:01 this type of live loop program would work

13:03 in the poorest of the poor.

13:05 Um,

13:06 we have the average of 7.4 household members.

13:10 90% of our caregivers are women.

13:12 Average of 41 years old.

13:14 Less than half of them have ever attended school.

13:19 Um,

13:20 90% of the households are farming,

13:22 and

13:23 this is important,

13:23 the 63% of them already have some kind of

13:26 non-farm business.

13:28 I'll explain that more.

13:30 Some pictures

13:31 to see on some of our recipients.

13:35 OK,

13:35 so just pulling um

13:37 the timeline.

13:37 So starting in 2018

13:40 we had our baseline survey

13:42 which was in a sample of people who

13:44 are all already receiving the basic cash transfer.

13:47 So the cash transfer line goes

13:49 the entire time because the whole everyone in

13:51 the sample is a cash transfer recipient.

13:53 Um,

13:56 Implementation is delayed by some time and then COVID hit

13:59 so we don't actually implement the interventions until early 2021

14:04 where we have the household

14:06 choice.

14:07 um,

14:08 households can make their choice and then we have the sensitization visits

14:11 and then

14:12 that's very quickly followed with the trainings that households are receiving.

14:17 And then the cash transfer

14:19 is bursed.

14:22 While the cash grants are being disbursed is

14:24 when we need to conduct our midline survey

14:27 for World Bank people for logistical reasons,

14:29 the project and you know,

14:30 so we had to

14:32 have this time in

14:33 which

14:34 leads to really interesting results on our show in a minute.

14:37 So we have.

14:39 About a third of our recipients have received,

14:41 but it's really right as they're receiving,

14:43 so

14:43 all the results I'll be showing you today are really coming from

14:46 people having all of this training

14:48 and the anticipation of the grant

14:50 and then any savings that they have

14:52 or whatever activities that they've started doing,

14:54 it's all coming from there

14:55 as opposed to

14:57 the,

14:57 so it's all pre-grant training

14:59 and we're going back this fall

15:01 to collect more data,

15:03 um,

15:03 now that people have actually had the grant for a couple of years.

15:07 OK,

15:07 so now finally let me show you some results.

15:09 So the first thing is on the selection of the um.

15:14 Of the beneficiary of the productive member,

15:16 so what we find is that when households have the choice,

15:19 a third of the time they shift away from the caregiver when they have the option,

15:24 and we find that the shift is equally to men and women.

15:28 So in the first

15:29 paragraph there about 80% of people.

15:32 Um,

15:32 in the default,

15:33 the caregiver is in,

15:35 in the control group,

15:36 the caregiver is selected.

15:38 This reduces by 25% points,

15:41 and

15:42 this

15:43 is

15:44 about equal split between a shift to women and to men.

15:47 So when it's no longer the caregiver,

15:49 half the time it's a woman and half the time it's a man.

15:52 Um,

15:52 it's less likely to be the household head's wife.

15:54 That's just a reflection of the fact that

15:56 many household heads wives are the caregiver,

15:58 and it is more likely now to be the child

16:01 of the household head,

16:03 and it's also more likely to be a young person,

16:05 someone under the age of 40.

16:08 And moving now just to summarize what happens with the.

16:13 Very early results that we're seeing again this

16:15 is all pre-grants but we're seeing an increase in

16:18 household employment

16:20 um

16:20 and

16:21 we think this is coming,

16:22 um,

16:22 primarily from

16:24 the fact that the caregiver is

16:26 a lot of them

16:28 just from the cash transfers the

16:29 savings groups they're already mobilizing into already

16:32 working in a,

16:33 um,

16:34 productive,

16:35 um,

16:35 household enterprise so they're doing

16:37 little petty trading they're doing something

16:39 and so there are other members who are not.

16:42 Economically active who are not engaged and so we find higher employment now for

16:47 these other members we see an increase in household earnings as well

16:50 and we see some

16:51 um increases not statistically significant but once we um

16:55 uh combine all of these

16:58 different sources of income we see increases in the overall household earnings.

17:03 On the household sensitization we don't see any effects on the

17:07 choice of the productive member.

17:08 We see some improvements in women's mental

17:11 health and their participation in paid work

17:13 and in children's nutrition,

17:15 um,

17:16 but we don't see any effects on some of these

17:19 outcomes I just showed about economic activity and consumption.

17:22 And interestingly we also don't see anything in this intersection of the two,

17:26 so our ideas that helping them to discuss

17:29 would help apparently they can make those decisions on their own.

17:34 OK,

17:34 so just um wrapping up,

17:37 these are all very early results like I've pointed out

17:40 we're um

17:42 going to get more data but

17:44 what we see so far is that it seems that

17:46 private information that households have among themselves is very important

17:50 and it's and we need to think about how we can leverage them in design of this program.

17:54 Yesterday and even part today there was a lot of talk about

17:57 how

17:58 governments are in a very tight fiscal space

18:00 and you know

18:01 these investments in

18:03 social protection programs

18:05 uh

18:06 uh fiscally demanding and here we see

18:08 really promising evidence of how a simple tweak

18:11 in the within household program targeting

18:14 can lead to significant improvements in outcomes.

18:17 And here we see specifically that it's

18:19 moving towards the younger members of the household

18:21 which then deals with this

18:23 big big problem that we've discussed many times here

18:26 about the youth unemployment so

18:29 um we're

18:30 cautiously excited

18:32 and look forward to um telling you more about what we find in a few months.

18:36 Thank you.

18:42 Ask the baton.

18:45 Uh,

18:46 OK,

18:46 great,

18:46 um,

18:47 so,

18:47 uh,

18:48 it's gonna take 10 seconds.

18:50 That's fine.

18:50 Uh,

18:51 so I'm standing in for Thomas Gin,

18:52 uh,

18:53 who's,

18:53 uh,

18:53 based at CGD,

18:54 and my fantastic co-author,

18:56 uh,

18:56 and this is joint work of ours,

18:58 uh,

18:58 together with Travis Bassler at Rochester,

19:01 uh,

19:01 Belinda Muya at the,

19:02 uh,

19:02 IRC,

19:03 uh,

19:04 and Ibrahim Kassiri at EPRC in Uganda,

19:06 um,

19:07 and so we're gonna be talking about work that is

19:09 done,

19:12 waiting.

19:13 We're gonna talk about some work that we've done,

19:15 uh,

19:15 focusing on,

19:16 um,

19:17 uh,

19:18 programs to sort of,

19:19 uh,

19:19 provide both cash assistance and mentorship,

19:22 uh,

19:23 to youth,

19:24 um,

19:24 in the,

19:25 in urban.

19:26 Did you click.

19:28 There's nothing.

19:28 It's not me.

19:29 There's nothing,

19:30 uh,

19:30 in,

19:30 in Kampala in Uganda.

19:32 There we go,

19:34 come back,

19:34 um,

19:35 OK,

19:35 so we'll be talking about work,

19:36 uh,

19:36 that is,

19:37 uh,

19:38 uh,

19:38 taking place in Kampala in Uganda,

19:40 um,

19:40 and I think with a special eye,

19:42 I think we're,

19:42 we're still allowed to talk about women in this session.

19:44 So with an eye towards,

19:46 uh,

19:47 sort of gender cleavages and then in particular,

19:49 uh,

19:50 cleavages acro uh.

19:51 Sort of another dimension of potential economic exclusion

19:54 which is between host country

19:57 nationals and refugees in this urban context

20:00 as we think about what are the particular

20:02 barriers that might be there for these particular populations

20:05 that might impede

20:07 economic integration for them.

20:09 So

20:10 to think about where we're coming from this and and it's nice

20:12 to kind of follow Oyebola's uh presentation because I think there's some features

20:16 that I think you'll you'll hear across several presentations about

20:19 the types of livelihood interventions that are that are there.

20:22 So here I think you know the starting point is that this is

20:24 a population that is engaged in informal micro entrepreneurship to a large extent,

20:29 uh,

20:29 but in a way that has relatively low profits,

20:32 uh,

20:32 for most,

20:33 um,

20:33 and we hypothesize that some of the constraints

20:35 that they're up against are both financial constraints.

20:38 So there are liquidity constraints stopping

20:39 people from expanding their businesses,

20:42 um,

20:42 but also maybe managerial capital that would

20:44 impede them or or the presence of which would enable

20:47 them to sort of better put those resources to use.

20:50 um we've seen lots of cash interventions with some positive effects,

20:53 at least in the short run,

20:55 sometimes these are lower for women

20:57 and so we want to think about kind of what are the

21:02 potential design features that you could build around that

21:03 to make sure that those returns were there for all

21:06 and.

21:07 When we think about managerial capital

21:09 training interventions that might sort of provide formal business training

21:12 um

21:13 have varied and I think sort of modest effects

21:16 in general relative to the relative high costs

21:20 and so thinking about on uh mentorship models

21:23 that might provide an alternative to that that sort of

21:25 provide the decentralized delivery of information at lower cost to the

21:29 uh to the NGOs that are trying to support these things or the governments

21:32 that are trying to support these things

21:33 may be a viable and interesting alternative.

21:36 Uh,

21:36 which raises,

21:37 I think,

21:38 potentially

21:39 interesting questions about how these things

21:40 get tailored to the particular recipient,

21:43 how it matters who is mentoring whom for the types of

21:46 constraints that they might unlock,

21:48 um,

21:48 so that's part of what we're going to explore in this study.

21:51 Um,

21:52 and so we're gonna think about kind

21:53 of how mentorship might improve business practices,

21:55 uh,

21:55 the networks that people have access to,

21:57 um,

21:57 and also knowledge,

21:58 knowledge exchange in ways that might be important,

22:00 uh,

22:01 for these populations in particular.

22:04 OK,

22:05 so we're working with the IRC to study uh an,

22:08 an aspect of their rebuild programming.

22:11 Uh,

22:11 we do so,

22:12 um,

22:13 with a sample of Ugandans,

22:15 uh,

22:15 who we'll,

22:16 we'll describe as the host country nationals and then refugees

22:19 and also purposefully sampling across gender cleavages,

22:22 um,

22:23 to,

22:23 to look at both impacts on men and impacts on women.

22:26 Um,

22:26 and in the large part we're gonna be focusing on a,

22:30 uh,

22:31 cohort of 2000 relatively inexperienced perspective

22:35 entrepreneurs who fall into this kind of

22:37 broadly defined youth category,

22:39 um,

22:40 and then,

22:40 uh,

22:40 they are in some instances gonna be paired with mentors who come from

22:44 a slightly more experienced cohort and tend to be a little bit older.

22:48 At the core of the interventions that everybody's going

22:50 to get are a set of cash grants.

22:52 These are $450

22:54 equivalent,

22:56 and the control group is getting these

22:58 sort of at the end of the study after,

23:01 after

23:01 all the measurement is wrapped up.

23:03 And then this is

23:05 in some instances paired with this mentorship

23:07 intervention that puts people not in one

23:09 on one mentorship but into groups of three mentees together with a single mentor.

23:13 They met weekly for a period of six months,

23:15 or that was the program to which they were assigned.

23:18 Um,

23:18 this is about 2/3,

23:20 almost 2/3 of our sample,

23:22 um,

23:22 and consists of a couple of pieces.

23:24 Uh,

23:24 at the core of it is

23:25 some actual kind of structured content to give a conversation starter

23:30 and maybe to sort of,

23:31 uh,

23:32 provide a framework where,

23:33 where,

23:33 um,

23:34 uh,

23:35 where,

23:35 uh,

23:35 for the kind of exploration of some formal skills

23:38 to the extent that those kind of managerial

23:40 skills are important and standardizable in some way.

23:44 But we're also going to think about how

23:46 the particular pairings might matter for the types

23:48 of information that people can give each other,

23:50 so we experimentally vary whether the groups are homogeneous

23:54 or heterogeneous with respect to gender and nationality,

23:58 and we'll sort of think about that.

23:59 And then following the literature that

24:00 suggests that some of this idiosyncratic advice

24:03 that can be particularly helpful but that fades away as pairings fall apart,

24:07 we also have a kind of layer in this study that thinks about whether.

24:11 Whether part of the constraint to the success of these things might not be

24:16 might be in the in the incentives within

24:18 the mentorship models and whether some kind of

24:20 shared fate might be important to getting

24:23 to perform better,

24:24 so there's some group level incentives in some of these groups

24:27 and what I'll share with you here

24:28 are preliminary results from a baseline

24:30 and then quarterly surveys over a year of exposure.

24:34 We've got one

24:35 last study round

24:37 coming up behind that.

24:39 In broad terms,

24:39 what we're going to find is positive and

24:42 persistent both over time and over groups,

24:45 I think to a surprising extent effects of cash,

24:48 and I'll show you how that holds up.

24:50 And then on average,

24:52 essentially a null effect of the mentorship layer put on top of that,

24:56 but a null effect that masks some,

24:58 I think,

24:58 quite interesting.

24:59 Heterogeneity that's central to the design here

25:02 and in particular we'll see positive effects

25:05 for some groups,

25:06 for men in particular,

25:07 and especially when they are matched with men,

25:09 and some negative effects for women,

25:11 particularly when they're matched with women,

25:13 and that challenge,

25:13 I think,

25:14 will be in terms of the design is a little

25:15 surprising to all of us and something we want to

25:17 unpack in this and further work.

25:21 OK,

25:21 so,

25:22 uh,

25:22 to be brief,

25:23 uh,

25:23 with background,

25:24 so Uganda's a,

25:25 a,

25:25 a progressive,

25:26 uh,

25:27 uh,

25:27 refugee hosting country with lots of people there and with,

25:31 uh,

25:31 pretty broad rights to work and to move.

25:33 Um,

25:34 this is a population in this urban setting that

25:37 they are not,

25:38 um,

25:38 because they are registered refugees,

25:40 they are not refugees who have just arrived.

25:42 the previous month or anything like that,

25:46 many of them have

25:47 businesses.

25:48 70% have businesses at baseline.

25:50 That's a little higher among the Ugandans and a

25:51 little higher among the women in our sample.

25:54 Their profits

25:55 are low from these businesses,

25:57 are modest from these businesses at $28 a month as we measure it,

26:02 but similar.

26:02 Between the Ugandan and refugee populations,

26:05 bearing in mind that this is a host

26:08 population that's selected to be sort of comparable in

26:10 in sort of the economic barriers that they face,

26:13 uh,

26:13 a little higher for men

26:14 and significantly higher for the mentors who are brought in

26:16 precisely because of having established businesses along those lines but

26:20 not

26:21 sort of a whole world apart you might say.

26:24 Um

26:25 And one of the key questions here will be what might be

26:28 the importance of the business context that people can be introduced to,

26:31 so,

26:32 so it's notable then that Ugandan men have the most business contacts at baseline,

26:37 um,

26:37 but not a lot of cross-nationality or cross gender contact,

26:41 which suggests that that might be,

26:43 there may be frictions there if there if there's useful information

26:46 exchange to be had that might have some positive effects.

26:50 So let me show you then some of our our core results.

26:54 So as I said,

26:55 I'm I'm showing you for brevity results that are just

26:57 presenting the kind of pooled effects across all of the

27:02 subject groups for the time being

27:04 and across all of the rounds,

27:05 but these are these are temporarily persistent

27:09 and persistent across the

27:12 host and refugee populations to

27:15 to a

27:16 surprising extent.

27:18 So we see quite substantial effects on profits.

27:21 These are moving from about $40 equivalent in the post-intervention period

27:26 to close to $80.

27:28 We see a near doubling of the capital stock,

27:30 so a lot of the cash that people are getting

27:32 is showing up in increases in the capital stock.

27:35 We see against even that relatively high base

27:39 some

27:40 increase in the prevalence of business ownership.

27:43 And then these things

27:45 filter through into better

27:49 mental health,

27:50 so people report being sad most of the time to to a lesser extent.

27:54 We see that these don't seem to be these

27:57 sort of related to Oyebola's question in a way

28:00 these targeted interventions don't seem to be crowding out

28:03 income elsewhere in the household,

28:04 so it's not the case that the

28:05 household earning effects are substantially less than the

28:09 beneficiary business effects,

28:11 and we see this sort of

28:14 passing through into.

28:15 Measures of improved consumption stability and the

28:17 like as the skipping of meals decreases substantially

28:21 in this population.

28:22 So I think the first thing as we think about the

28:24 sort of fundamental constraints in this population is that liquidity constraint

28:28 is substantial.

28:29 And while there's a broad literature here,

28:31 I think we contribute something to the sort of sense

28:33 of stability and persistence over time and across populations.

28:38 Turning to the mentorship

28:39 dimension of this,

28:41 um,

28:42 we see that these effects are really quite modest.

28:44 We look at this across subgroups and

28:47 thinking about who people whom people are paired

28:50 with when we want to think about what the consequences of mentorship might be.

28:53 Um,

28:53 so showing you first in the panel on the left,

28:56 uh,

28:56 the benefits of mentorship for Ugandans and refugees

29:00 again across both genders in our sample here.

29:04 So relative to cash,

29:06 you see that Ugandans are in point estimate terms slightly worse off

29:10 in the mentorship program

29:12 than without it,

29:14 and the refugees are sort of,

29:16 you know,

29:17 more or less the same

29:18 to a large extent

29:19 with some hint of something positive when they're

29:22 mentored by a refugee rather than by Ugandan.

29:25 The gender dimension is where this becomes particularly interesting.

29:29 So

29:30 as you can see,

29:31 so at the 10% level,

29:33 men being mentored by men have statistically significant

29:38 profit effects in this sample over the study period.

29:41 And these are sharp,

29:42 especially in the early

29:43 early rounds,

29:45 and then

29:46 mentorship seems to,

29:47 on the other hand,

29:48 have negative effects for women,

29:51 particularly when women are mentored by other women,

29:53 contrary to something of our expectation here

29:57 and the expectation of our implementing partners.

29:59 Um,

30:00 so thinking about how the decisions that are made by people are shaped by

30:05 the experience of mentorship and,

30:06 and how the access to information that they get

30:09 may sort of be

30:12 traded off against the ways that they,

30:14 they,

30:14 they make slightly different decisions or or or or decisions that

30:17 exposes seem not to have worked out I think is an interesting

30:20 and key question for us.

30:22 We don't see.

30:23 These effects driven by changes in the sectors to which people enter into,

30:27 we don't see them driven by the likelihood of opening a business,

30:31 so,

30:31 so some of the obvious channels that you might have expected

30:35 are not there,

30:37 um

30:37 but

30:38 uh but I think this is potentially

30:40 potentially important

30:41 where we do have some sense of what might be going on is

30:44 when we look at how this works across the distribution of business profitability.

30:49 Um,

30:50 and so when we,

30:50 so what I'm showing you here

30:52 are quantile treatment effects

30:54 looking across the range of,

30:56 of,

30:57 uh,

30:57 business profitability in the,

30:58 in the after intervention period,

31:01 um,

31:01 and first showing you in the upper panels for men and for

31:04 women what the impacts of the cash grants are by quantile.

31:07 So unsurprisingly,

31:08 you know,

31:08 we see these big effects at the top of the distribution

31:11 um that are that are potentially important here.

31:14 Strikingly then,

31:14 when we look at the effects of mentorship,

31:17 and so the bottom graphs are going to show

31:18 you the effect of mentorship relative to cash alone

31:21 across the quantile of the of the profit distribution.

31:24 We see that the negative female effect seems to be coming particularly at the top,

31:29 so it seems like

31:31 some high return things that might have happened in the absence of mentorship

31:35 don't get invested in

31:37 because of the conversations that are happening in these groups,

31:40 particularly when women are being mentored by other women in this context.

31:44 But exactly how that's being driven and what trade-offs,

31:48 what other benefits might be there that are being traded off against,

31:50 I think

31:51 are questions we want to explore a little bit more.

31:53 OK,

31:54 so then just to,

31:55 to wrap up,

31:56 um.

31:57 We see in this

31:59 work

32:00 substantial effects of the cash grants on business and household outcomes.

32:03 These effects are persistent over time and across populations,

32:08 and,

32:08 and I think one contribution here is to sort of line up in time and

32:13 at the same time.

32:14 Time and space,

32:15 uh,

32:16 a trial studying both the refugee and host population impacts

32:21 and to show that we have some evidence of the generalizability of

32:24 the

32:25 sort of

32:26 interventions to alleviate

32:27 liquidity constraints across that that

32:29 barrier or across that division.

32:32 We have these marginally positive effects of mentorship among men,

32:35 particularly when they're being mentored by other

32:37 men,

32:37 and these marginally negative effects of mentorship among women,

32:40 particularly when they're being mentored by other women

32:43 and seemingly concentrated among the upper quartile of the profit distribution,

32:48 whereas

32:49 we don't see this being driven by substitution of other economic activities,

32:52 we don't see this being driven by differences in the sectors that they choose,

32:56 and we don't see this being driven by,

32:58 for instance.

33:00 The taxation by their mentors or something like that,

33:03 that there might be some bargaining going on within the mentorship

33:05 group that that results in funds not going to them.

33:08 So

33:09 some evidence on those constraints and some mysteries about where why

33:13 why this mentorship model has fallen down in some instances for some people.

33:17 Perfect,

33:17 thank you.

33:22 So thank you so much.

33:23 Uh,

33:23 it's really great to be here.

33:24 Um,

33:25 and thank you for being here.

33:27 I know it's where I'm keeping you from,

33:28 or we're keeping you from dinner,

33:30 so,

33:31 uh,

33:31 we'll try and be,

33:32 uh,

33:32 engaging and interesting,

33:33 and,

33:34 uh,

33:34 but yeah,

33:35 we'll try.

33:36 So,

33:36 uh,

33:37 I'm gonna give a um

33:39 Uh,

33:39 a presentation on,

33:41 uh,

33:41 some work I've been doing in Ghana with co-authors,

33:43 uh,

33:44 Gabriel,

33:45 Morgan,

33:45 Jamie,

33:46 and Isabel,

33:46 and we've had some outstanding,

33:48 uh,

33:48 RA support from,

33:49 uh,

33:50 DT and Michelle as well.

33:51 So I wanna just give them a

33:52 shout out to people who are behind the scenes who

33:55 do great work in helping us move things along,

33:58 um,

33:59 so we've been working in Ghana for some time and.

34:02 Um,

34:03 you know,

34:03 I think,

34:04 and I,

34:04 I,

34:04 I'll probably skip a lot on,

34:06 on,

34:06 on this motivation here and this sort of framing,

34:08 but

34:08 essentially,

34:09 you know,

34:09 we're all thinking about youth unemployment and underemployment,

34:12 and,

34:12 you know,

34:13 uh,

34:14 it's not good,

34:15 right,

34:15 though I think we can,

34:16 uh,

34:16 all agree on that,

34:17 and we're all trying to find ways on,

34:19 uh,

34:20 to,

34:20 to address this problem,

34:21 right?

34:21 And so

34:22 what we've been doing in Ghana for some time is looking at apprenticeships and,

34:25 uh,

34:25 in many,

34:26 you know,

34:27 uh,

34:27 low and middle income countries we

34:29 have apprenticeship systems they're pretty common.

34:32 Um,

34:32 I think often when people think of apprenticeships they think of

34:34 sort of the sort of German model in some sense of,

34:37 you know,

34:37 well established big firms,

34:39 etc.

34:39 but

34:40 in these contexts that we're looking at we're often thinking about or,

34:43 you know,

34:44 speaking about sort of,

34:45 um,

34:46 working in very small scale informal sectors,

34:48 right,

34:48 where the firm owner,

34:49 you know,

34:49 has maybe like 2-3 employees and a couple apprentices,

34:52 and they're the one that's providing the,

34:54 the training,

34:54 and that's kind of the model of training that we're looking at and studying in,

34:57 in,

34:57 in we've been studying in our work here.

35:00 And so why do people sort of get interested about apprenticeships in general?

35:02 I think

35:03 they're seen as this promising way of sort of providing skills,

35:06 um,

35:07 but also and with the idea that you know I can

35:10 boost the employment prospect uh at the end of the day,

35:13 um,

35:13 and we,

35:14 it,

35:14 it's seen as maybe practical right because again you're learning sort of on

35:18 the job you're learning from someone who actually is active in the trade,

35:21 right?

35:22 and in some cases this is paired

35:24 with on you know classroom instruction sometimes not

35:27 um.

35:28 And you know there's sort of some thought that

35:30 you know what ends up happening really is that

35:33 at least in the sort of the systems that we're

35:34 studying is that the apprentice basically learns from the trainer

35:37 and then basically replicates exactly what the

35:40 trainer does and sort of copies their business

35:42 at the end of the day and goes sets up their own firm right so.

35:45 Again,

35:45 unlike sort of what you see in developing countries,

35:47 there's no real retention of like the firm hiring the apprentice at the

35:50 end of the day where this is sort of a screening mechanism for

35:53 for your workers,

35:53 but this is really just a

35:55 uh a training um

35:57 system right that then sort of trains other folks that enter

36:00 um and start new businesses,

36:02 right?

36:02 so.

36:03 Uh,

36:04 you know,

36:04 this arguably could be more relevant for sort of the context

36:08 we're talking about where this is a large informal sector,

36:10 right,

36:11 um,

36:11 you know,

36:11 in Ghana,

36:12 you know,

36:12 88% of men,

36:14 95% of women are in this sort of informal sector that's,

36:17 you know,

36:17 arguably low productivity,

36:19 right,

36:19 and so.

36:19 You can imagine that this is,

36:20 you know,

36:20 you're learning from someone who's actively in the space who's been

36:23 experiencing in this space and so this might actually be a good

36:26 um

36:26 um form of training,

36:28 a relevant form of training for that relative to let's

36:30 say a vocational training uh institution which is very classroom,

36:34 uh,

36:34 based right and so

36:36 um.

36:37 You know,

36:37 and you know,

36:38 it's,

36:38 it's somewhat surprising,

36:39 or at least you know one surprise people often get surprised,

36:42 but just how common these,

36:44 these sort of types of training are and,

36:46 you know,

36:46 in Ghana for example,

36:47 like,

36:48 uh,

36:48 this type of training in formal sector apprenticeships has trained almost

36:52 4 times as many individuals as all other sort of formal

36:55 training of,

36:56 um,

36:56 alternatives,

36:57 right?

36:57 So if you look at sort of people who went to

36:59 sort of vocational schools and all these other forms of schooling,

37:02 you know,

37:02 apprenticeships basically dominates it yet.

37:05 The

37:06 actual academic literature looking at apprenticeships like the economics

37:09 literature is actually pretty thin relative to the importance

37:12 um.

37:14 Now,

37:15 um,

37:15 well,

37:16 again it could be relevant in terms of,

37:17 you know,

37:18 actually training you for,

37:19 uh,

37:20 the,

37:20 the,

37:21 the sectors,

37:22 um,

37:22 I think a lot of,

37:23 uh,

37:23 observers have raised

37:25 quality concerns,

37:25 right?

37:25 And so

37:26 things that they've raised are some factors

37:28 have sort of pointed to specifically are,

37:30 you know,

37:31 many of these firms are using outdated technology.

37:33 There's a lack of monitoring that's going on

37:36 there's a lack of formal contracting

37:39 the

37:40 standards in the curriculum are often sort of not there,

37:43 right?

37:43 They're just sort of everyone's kind of doing their own thing in some sense,

37:47 um,

37:47 there's no sort of formal certification or it's limited in many contexts,

37:51 um.

37:52 And there's also the strategic concern right where basically because again I'm not

37:56 unlike you know sort of like a German system where

37:58 the firm sort of hires a person at the end of the day

38:01 here you're basically after the training they go off and

38:04 start their own business they replicate your own business,

38:06 right?

38:07 And so there's this

38:09 concern that you might be like well I'm

38:10 actually I don't wanna train Andy Owen Oyobola because

38:12 they're gonna start competing with me so I better maybe slow the pace a little bit,

38:17 maybe withhold some things so there are these

38:18 strategic concerns that come into play as well.

38:21 So

38:23 How

38:23 could we address this concern,

38:25 and that's what we're going to look at,

38:26 um,

38:27 or that's what I'm gonna talk to you about today is basically one potential way to

38:31 address,

38:32 uh,

38:32 some of these concerns,

38:33 these quality concerns,

38:33 and one way to do that is by introducing financial incentives

38:37 very much in the spirit of what Andy was talking about or just sort of

38:40 getting everyone's skin into the game,

38:41 right,

38:41 and sort of a shared,

38:42 um.

38:43 Uh,

38:43 skin in the game,

38:44 so to speak,

38:45 right,

38:45 so the idea here is that we will,

38:47 again,

38:47 economists love incentives,

38:48 right?

38:49 That's any,

38:50 that's always a solution to everything that economists sort of come up with,

38:53 um,

38:54 but you know,

38:54 in this case we thought,

38:55 you know,

38:56 um,

38:56 the trainer remuneration or,

38:57 you know what they actually get right in terms of cash,

39:00 um,

39:00 and what they apprentice,

39:01 you know,

39:02 if we can strengthen that link,

39:03 right,

39:04 that could potentially increase,

39:05 um.

39:06 Um,

39:07 you know,

39:07 the sort of skills,

39:09 uh,

39:09 uh,

39:09 attainment right amongst,

39:11 among apprentices,

39:12 right,

39:12 and so we thought,

39:13 well,

39:13 this could encourage training effort and improve skill acquisition

39:17 and maybe,

39:17 uh,

39:18 sorry,

39:18 some of these things didn't render very quick uh correctly,

39:21 um,

39:21 better skills

39:22 could lead to better.

39:22 The labor market outcomes right

39:24 now again this was I think motivated a lot from I think,

39:27 uh,

39:27 the academic literature in,

39:29 in,

39:29 in schoolings right where you see in schoolings where you see basically

39:33 financial incentives for school teachers in

39:34 K-12 education can be quite effective.

39:37 There's a lot of literature on this,

39:38 um,

39:39 that's,

39:40 that's sort of done shown that.

39:42 Now

39:43 Well that's true in the sort of schooling literature,

39:46 it's unclear if sort of financial incentives

39:48 will be effective in the apprenticeship system,

39:50 right?

39:50 So

39:51 again,

39:52 when I come to a trainer,

39:53 right,

39:53 there's a large opportunity cost of training,

39:55 right?

39:55 So like if I'm training someone,

39:57 right,

39:57 if I spend a little bit more time training or Ebola Andy

40:00 and Owen,

40:01 right,

40:01 I'm

40:02 potentially not spending time making stuff,

40:04 dealing with customers,

40:05 right?

40:05 So,

40:06 so there's a large opportunity cost of training,

40:07 right?

40:08 And there's also these strategic considerations like,

40:10 you know,

40:11 if I do put more effort they're gonna be better at competing against me,

40:13 right?

40:14 So,

40:14 so it's not clear this will automatically

40:16 just translate from the education sector or,

40:18 you know,

40:18 the formal K-12 education sector into this apprenticeship system,

40:22 um.

40:23 Now,

40:23 on the other hand,

40:24 um,

40:24 what you do see if you just to sort of survey sort

40:26 of a lot of training programs that do exist is you see

40:29 many of them sort of feature these sort of,

40:31 um,

40:32 outcome-based contracts for trainers where basically the trainer payment is

40:36 linked to training outcomes.

40:37 So you've seen this in,

40:38 you know,

40:38 some work that the World Bank did in Liberia by I think lots of folks,

40:42 um.

40:43 And other places,

40:44 right,

40:45 so that's

40:46 become a common feature of,

40:48 of,

40:48 of many training programs

40:50 but again that's we actually don't have limited evidence of

40:53 the effectiveness of that sort of contractual structure so.

40:57 So,

40:58 so in this experiment what we're gonna do is,

41:00 uh,

41:00 now it's embedded within a larger experiment,

41:03 uh,

41:03 that we're doing,

41:04 uh,

41:04 that's,

41:05 you know,

41:06 still sort of working on it,

41:06 on that's trying to look at the returns to

41:10 apprenticeships in,

41:11 in Ghana

41:12 and which was that RCT was in collaboration with the,

41:15 the Council for Technical Educational Education and Training in Ghana,

41:18 uh,

41:19 it's a government agency.

41:20 And

41:21 so what we're doing here,

41:22 what I'm going to talk about today here is basically a

41:25 sub experiment within that larger experiment where we're basically looking at,

41:29 you know.

41:30 What is the effect of giving incentives to

41:32 training providers within this,

41:34 uh,

41:35 larger RCT?

41:36 OK.

41:36 The apprenticeship training that we're,

41:39 uh,

41:39 looking at is gonna focus on construction trades,

41:42 uh,

41:42 cosmetology and hairdressing and garments and tailoring,

41:44 right?

41:45 So those are the sort of the trades that were offered.

41:47 We didn't pick those as,

41:48 as researchers.

41:49 This is what the,

41:49 the,

41:50 the government sort of had picked.

41:52 Um,

41:52 and,

41:53 uh,

41:53 due to logistical considerations and just constraints that we had,

41:57 um,

41:57 basically there were no syllabi,

41:59 right,

41:59 that we could use,

42:00 um,

42:00 we could only sort of implement this,

42:02 uh,

42:02 for cosmetology and hairdressing and tailoring and garments.

42:08 So this is just a broader um research design uh sketch just

42:12 to sort of give you an idea of what's going on.

42:14 So in 2012,

42:15 so yeah we,

42:16 this has been going on for a while,

42:18 um,

42:19 we sort of did the baseline for the larger RCT,

42:21 you know,

42:22 recruited us applicants,

42:24 um.

42:25 You know,

42:25 we sort of put people into different buckets of control treatment.

42:28 Uh,

42:29 priority you'll see there is basically we allowed,

42:31 uh,

42:31 district officials to pick,

42:32 um,

42:33 oh,

42:34 only 5 minutes.

42:34 Alright,

42:35 uh,

42:35 pick,

42:35 uh,

42:36 people,

42:36 so I'll keep going.

42:37 I'll talk a little bit faster

42:38 and then we moved to match apprentices with trainers,

42:42 uh,

42:42 that was sort of

42:43 2013,

42:44 and then we,

42:45 uh,

42:45 layered on the incentives program just among those that were in training,

42:48 OK,

42:49 and so we have about,

42:50 uh,

42:50 you know,

42:51 225 firms in each,

42:53 in each,

42:53 um,

42:54 in each treatment arm.

42:55 Um,

42:56 what are we doing basically in the treatment group,

42:57 um,

42:58 trainers are gonna give a,

42:59 uh,

42:59 get a financial bonus based on a skills test

43:02 that we're going to administer to their apprentice,

43:04 right?

43:04 So basically

43:05 if

43:06 they,

43:06 we,

43:06 we come in,

43:07 we're gonna test the apprentice,

43:08 um,

43:09 see how,

43:10 you know,

43:10 how.

43:11 Well they've learned hairdressing or tailoring

43:13 and then basically the better the apprentice does within the district,

43:17 right,

43:17 so if they're the top ranked,

43:19 uh,

43:19 tailor of this test within the district,

43:22 the apprentice gets a bigger bonus.

43:23 So it's a rank order tournament structure,

43:25 um,

43:26 within the district

43:27 and in the control group they're going,

43:29 the trainers are going to get a fixed payment of 100 Ghana CDs

43:32 this,

43:32 uh,

43:32 if the if the apprentice takes the skills test,

43:35 right?

43:35 So,

43:36 um,

43:36 basically there's no incentive there to,

43:39 you know,

43:39 increase your effort.

43:41 Uh,

43:41 and the average payment is equalized across,

43:43 across both groups.

43:45 Um,

43:46 here's a bonus schedule.

43:46 I'll skip it,

43:47 but,

43:48 um,

43:48 and I'll skip all of this.

43:50 So we have lots of data,

43:51 uh,

43:52 basically we have our baseline,

43:53 we have our skills data,

43:54 we have,

43:55 uh,

43:55 tracking data,

43:56 so we're going through,

43:57 um,

43:57 Um,

43:58 the intervention we have,

43:59 and we have two follow up data,

44:01 uh,

44:01 we have one,

44:02 sort of what we'll call a midline,

44:03 so to speak,

44:04 in 2017-2018,

44:06 which is like

44:07 two years after,

44:08 um,

44:09 the intervention of the incentives,

44:10 and then we have a 2022 follow up,

44:12 uh,

44:13 as well.

44:13 So I'm actually gonna present

44:15 both and the 2022 follow up with 2022,

44:18 2023,

44:19 that's fresh off the

44:20 this data,

44:21 so

44:22 it's very exciting.

44:23 Um,

44:24 so who are these folks,

44:25 um,

44:26 you know,

44:26 apprentices,

44:26 they're young people,

44:27 they're about,

44:28 they're about 23 at,

44:29 at the time of enrollment into the program,

44:31 about 7 years of schooling,

44:33 um,

44:33 uh,

44:34 because again,

44:35 we're focusing

44:36 primarily on cosmetology and tailoring,

44:38 they're basically all women,

44:40 OK,

44:40 um.

44:43 So skills assessment really hard to do

44:45 um we basically had to hire people who were

44:48 external experts to go around and conduct these assessments

44:52 uh so folks had to take a theoretical portion like there was a test that was developed

44:56 uh on the theory of hairdressing,

44:59 the theory of garment making,

45:01 and then we actually had to do a practical test.

45:03 Basically,

45:03 and you can see there on the pictures,

45:05 like,

45:05 you know,

45:05 how well can you braid hair,

45:07 how well can you,

45:07 you know,

45:08 style the hair,

45:09 and then,

45:10 um,

45:10 for the garments,

45:11 you know,

45:11 we had they had to like,

45:12 you know,

45:13 sew zigzag scallops,

45:15 straight edge,

45:16 and then they also had to like you know,

45:17 actually make a dress,

45:19 uh,

45:19 and we graded all that,

45:20 uh,

45:21 about 2/3 of the folks that,

45:22 um,

45:23 uh,

45:23 were assigned treatment actually took participated in the assessment,

45:26 and that's balanced across treatment and control.

45:29 And what do we find

45:30 um

45:31 in terms of skills

45:33 in the treatment group practical uh practical assessment,

45:36 right,

45:37 so how well can they sew,

45:38 how well can they do hair,

45:40 uh,

45:40 we had about a 0.13 standard deviation increase uh in um

45:44 in,

45:44 in,

45:45 in learning outcomes,

45:46 right?

45:46 So they actually did a lot better

45:47 in,

45:48 in skills.

45:48 Assessment,

45:49 um,

45:50 in a theory

45:52 it wasn't really much that was stati statistically significant,

45:54 and they,

45:55 and then they could get a certificate if they passed a,

45:58 um,

45:58 a threshold,

45:59 and again there was no real difference between

46:01 the treatment and control in terms of passing

46:03 this,

46:03 this test,

46:04 um.

46:05 We come back sort of in 2017 we give him a quiz,

46:08 um,

46:08 that basically says,

46:09 you know,

46:10 if you were to,

46:11 you know,

46:11 and that was sort of very practically,

46:13 um,

46:13 very practical questions about like,

46:15 you know,

46:15 how would you do this hair,

46:16 how would you sort of,

46:17 uh,

46:18 um,

46:19 you know,

46:19 do this stitch,

46:20 etc.

46:21 and again still this is again.

46:22 Two years after the the incentives

46:25 they were

46:26 better on they did better on that quiz

46:28 and they also did better on sales on the sales skills index,

46:30 right,

46:31 basically we asked them how would you actually you know sell this stuff

46:34 right?

46:34 so

46:35 basically skills went up as a result of the treatment and I think we're pretty

46:38 convinced about that.

46:40 In terms of labor market outcomes,

46:41 um,

46:43 we don't see much in terms of a change in

46:45 labor supply like we know which sector they worked in,

46:47 whether they worked for self-employment or wage work,

46:49 um,

46:50 but we see increases in earnings,

46:52 right,

46:52 so we're about,

46:53 uh,

46:54 you know,

46:54 25%

46:55 essentially increase in earnings in total earnings over the past 12 months

46:59 we're seeing sort of,

47:00 uh,

47:00 and that's driven in,

47:01 in,

47:01 you know,

47:01 in 2017 by self-employment earnings,

47:03 right?

47:03 So they're basically

47:05 much more productive in self-employment,

47:07 um,

47:07 as a result of that,

47:08 right?

47:08 And then.

47:10 Now,

47:11 why is that?

47:12 Well,

47:12 obviously skills went up.

47:14 They also worked more hours during the apprenticeship,

47:17 right?

47:17 So they got more

47:18 training,

47:19 so to speak,

47:20 um.

47:22 Their trainers used

47:23 more syllabi,

47:24 right,

47:24 so there was a standardization that sort of came into play

47:27 and there was an increase in the pace of training,

47:29 right,

47:30 so basically there were actually we see that

47:32 apprentices were more likely to complete their training

47:35 and um quicker.

47:39 What was surprising to us was these

47:41 results were persistent through 2022.

47:44 Right,

47:44 so.

47:46 Couple of years ago we come back we measure their labor supply

47:49 and earnings

47:50 we see that uh

47:51 in total earnings is still up by,

47:52 you know,

47:53 25,

47:53 28%.

47:54 Wage earnings is what's driving this now,

47:56 which is sort of surprising,

47:57 um,

47:58 so it's actually being driven by wage earnings

48:00 and there's no real increase in self-employment earnings,

48:03 um.

48:04 And uh

48:05 again it's it's all just coming through wage and and now we're seeing in

48:08 terms of labor supply they were more likely to work for a wage.

48:13 So

48:14 we were asked to do some policy sort of thoughts too which I'll,

48:17 I'll I'll,

48:18 I'll do my minus 1 minute I know,

48:20 um,

48:21 but essentially what I'll say is well

48:22 designed incentives could improve apprentice learning and

48:26 uh in a cost effective manner,

48:27 right?

48:27 So it's hard to do.

48:28 You have to really pay attention to the design details.

48:31 Uh,

48:31 and in our setting the program paid for itself in,

48:34 you know,

48:34 about 15 months,

48:35 right,

48:35 in terms of apprentice earnings,

48:37 right,

48:37 so it's actually pretty cost effective to do this.

48:39 Unclear if this will scale up,

48:41 right,

48:41 because again,

48:42 as you scale this up,

48:43 people can learn how to game and and sort of,

48:45 um,

48:45 game the system.

48:47 But on the other hand,

48:48 as I think,

48:49 uh,

48:49 Owen and co-authors work and Andy's work actually,

48:52 um,

48:52 has shown,

48:54 the incentive structure could attract a different or better trainer to the system.

48:58 OK,

48:59 so we just need to do more research

49:00 on the scale of potential,

49:02 uh,

49:02 and I'll stop there.

49:03 Thank you.

49:10 OK,

49:11 you've stuck around to the end.

49:12 You've heard about,

49:14 uh,

49:14 some long term follow up.

49:15 You've heard about

49:17 a cash grant program.

49:19 You've heard about a program that built

49:22 livelihood skills,

49:24 uh,

49:24 teaching people to fish and not merely giving them a fish,

49:27 and

49:28 in my presentation,

49:29 I hope to talk about all of those things.

49:32 Uh,

49:33 this is a firm of one's own.

49:34 It's joint work with Andrew Bruteval Newman

49:36 and Madalena Honorati at the World Bank.

49:39 Gerald E Papa,

49:39 who's just finished his PhD at University of Delaware,

49:42 and Pamela Jaquila,

49:43 who along with me,

49:44 we're at,

49:45 uh,

49:45 Williams College,

49:47 so.

49:48 Uh,

49:48 I want to show a picture,

49:49 a couple of pictures that illustrate things that Tamar said right at the beginning.

49:53 This is a picture taken from a report that,

49:55 uh,

49:56 Dion Filmer in the room and Louise Fox wrote.

49:58 What you can see

50:00 is that,

50:00 uh,

50:01 on the left side,

50:02 sub-Saharan Africa's population pyramid has a tremendous amount of youth,

50:06 OK,

50:07 compared to almost anyplace else in the world.

50:09 And so when we talk about

50:11 employment,

50:12 we need to think about what are young people going to do.

50:14 That's really the focus of policy and you pointed that

50:17 out earlier and this is a picture that illustrates it.

50:20 Uh,

50:20 when we

50:21 talk about sub-Saharan Africa,

50:22 this is a,

50:22 a graphic taken from Bandiera et al.

50:25 Uh,

50:26 look at the bottom thing with the pink bar and the green bars,

50:30 OK?

50:31 The green bars are self-employment,

50:33 the pink bars

50:34 are salaried employment,

50:36 wage employment.

50:37 The top one is Africa,

50:39 Sub-Saharan Africa,

50:40 28 countries I think,

50:41 and the bottom bar is other countries in the world,

50:43 and the stylized fact here

50:45 is that

50:46 self-employment is really big in sub-Saharan Africa.

50:50 Labor demand

50:51 in

50:51 wage employment

50:53 is quite low,

50:54 and there are factors that you can think about changing that.

50:56 We've heard talks today and yesterday about

50:58 different ways that,

51:00 uh,

51:00 industrialization and a revolution in services may change labor demand,

51:04 uh,

51:04 but this is the situation as it is now.

51:08 This is another picture of the same idea.

51:09 The pie chart

51:10 on the right side

51:12 shows that

51:13 there's this little quarter in the upper left that is wage sector

51:16 and everything else is not

51:18 in sub-Saharan Africa.

51:20 Also from Filmer and Fox.

51:21 So the project I'm gonna talk about today

51:23 happens in,

51:24 uh,

51:25 neighborhoods on the sort of east side of Nairobi,

51:28 Babadogo,

51:29 Dandora,

51:29 and Lunga Lunga.

51:30 Uh,

51:31 people who live and work in these neighborhoods live

51:33 and work with varying degrees of formality and informality,

51:38 uh,

51:39 and this is going to be an intervention a lot like the ones that you've heard about.

51:43 The intervention uh is run by the International Rescue Committee which

51:47 was working on the project in uh Uganda as well.

51:51 It's a micro franchising program,

51:53 so it's going to bring together a bunch of

51:54 the elements of things that you've heard about.

51:57 The

51:57 target population and the population in this study is young women

52:01 aged 18 to 19.

52:02 They're living in these neighborhoods and they express an interest

52:05 in starting a business.

52:07 The idea,

52:08 the distinction between micro franchising

52:10 and other multifaceted interventions is that

52:13 in addition to providing a kind of life skills training,

52:16 which you heard about in Oyebola's presentation,

52:18 uh,

52:19 and a specific business skills training,

52:21 in addition to providing

52:23 capital

52:24 so that you can start your business to overcome any

52:27 capital constraint,

52:27 credit constraint you may face,

52:29 in addition to all of those things,

52:31 micro franchising is going to attach.

52:33 You and your small business

52:36 to a nationally recognized brand.

52:38 You're gonna have a business model to do,

52:39 OK?

52:40 Now it's self-employment,

52:41 so you can decide whether you want to do it or not,

52:44 but

52:45 there's gonna be a business model all worked out and recognized branding.

52:48 So that's like

52:49 we're trying to overcome the IRC is trying to overcome with this program as many

52:54 of the possible constraints that may inhibit young people

52:57 from launching their businesses as possible along the

52:59 lines that we've heard in other presentations today.

53:02 So you're matched with one of a few different franchise

53:05 models on the basis of the preferences that you express,

53:07 and you get all this training,

53:08 and you even get mentoring,

53:10 that was a feature of several of these things as well.

53:11 You get mentoring

53:13 after your micro franchise is launched.

53:16 Uh,

53:17 the two sectors that this ends up being in,

53:19 the partners that the International Rescue Committee had to work

53:22 with young women in the east side of Nairobi,

53:24 uh,

53:25 were hair salons.

53:26 So here you can see an academy

53:28 where people are learning how to do different things with hair,

53:31 uh,

53:31 and here's one of the salons that they've started after coming out of the academy.

53:36 And uh mobile food carts where they sell-prepared,

53:39 ready to eat foods,

53:41 uh,

53:41 and there's like a,

53:42 a company that they're associated with and a national brand in each case.

53:47 Another thing I mentioned you heard about was a cash grant.

53:51 So we're going to run

53:53 a sort of a horse race of sorts between two interventions that are both meant

53:58 to help people in this population.

54:00 So one of them is the micro franchising program.

54:03 The other is a completely unrestricted cash grant.

54:06 It's about $230

54:09 20,000 Kenya shillings,

54:11 and

54:11 there was no particular encouragement

54:13 to do anything with this

54:15 money.

54:15 So you could think a lot of things might happen

54:17 as a consequence of getting a few 100 extra dollars.

54:21 Now,

54:22 what does this do for kind of thinking like economists?

54:25 Uh,

54:26 this will relax the credit constraint,

54:28 so if you're just short a few $100 and you don't have a way to get credit,

54:31 this will overcome that problem for you.

54:33 But if you face all kinds of other constraints,

54:36 this won't necessarily do anything about them.

54:38 So

54:39 we think of these as competing models,

54:42 um,

54:43 and so if,

54:43 for example,

54:44 a variety of skills that you could get in

54:46 the training program that accompanies the micro franchising intervention,

54:49 if those are the skills you're really lacking,

54:50 then it will do much better.

54:53 So which of these is going to help people

54:55 get out of poverty better?

54:57 Um,

54:58 one thing to notice and one

55:00 often described advantage of a cash grant program

55:03 is that it's much easier to implement.

55:05 It's very simple.

55:06 You don't have to find a partner organization

55:08 of the kind who's willing to do training and launch people with a brand.

55:11 It's much simpler than that.

55:13 Uh,

55:14 and

55:14 if I'm running a training session

55:16 next Monday

55:17 starting at 9 o'clock in the morning.

55:19 It's going to be a logistical constraint to get people to be there.

55:23 Some people have something already next Monday at 9 o'clock in the morning,

55:26 but if I just need to get you a couple of $100

55:28 we can meet at almost any time to do that.

55:31 So takeup is going to be very high in the cash grant program

55:34 relative to the micro franchising program,

55:37 and that's common.

55:38 The fact that it's not 100% take up,

55:40 more like 2/3,

55:41 60%,

55:42 that's common in the literature on active labor market programs.

55:45 People who are looking to change the way they interact with the labor force.

55:50 Have other stuff going on

55:51 and so

55:52 they aren't all going to show up at a training at a specified time,

55:55 even if they express an interest in doing so.

55:57 So for that reason,

55:58 what I'm going to show you are going to be what we call intention to treat results.

56:02 So I'm going to say here's who's assigned to this group,

56:04 here's who's assigned to that group.

56:05 It's a randomized trial

56:07 with 3 arms.

56:09 So

56:11 People in this randomized trial express interest in the program.

56:14 The program has limited spots,

56:16 so in the middle,

56:16 in the blue,

56:17 and blue will be a color associated with

56:19 micro franchising throughout the rest of the presentation.

56:22 Several 100 people are assigned to the micro franchise group.

56:25 In the green,

56:26 you have people who are assigned to the grant group,

56:28 and in the gray you have

56:30 the control group.

56:33 As I said,

56:34 uh,

56:34 the takeup is going to be lower in the more complicated program,

56:38 so we see that we get everybody who completes the baseline survey.

56:41 95% of those assigned to receive a cash grant

56:44 receive one,

56:46 and

56:47 of those assigned to and invited to come to

56:49 the training program to do the micro franchising,

56:52 a little over 60% start the training and go through some part of it,

56:56 uh,

56:56 and about 40%.

56:58 Uh,

56:58 get to the point of launching a micro franchise,

57:00 but if the training itself is all you need,

57:02 you don't necessarily have to get to the launch point.

57:04 So we're going to show you intention to treat in both cases.

57:08 Uh,

57:08 the cash grant is similar in magnitude to the value of the

57:12 of the micro franchising intervention,

57:14 but it's very hard to price some of the elements of that intervention,

57:17 so I'm not going to say that they're exactly the same.

57:19 I'm sure they're not.

57:21 OK,

57:21 first thing,

57:22 we followed this group of young women for 6 years

57:26 after the intervention.

57:28 And

57:29 on the dimension of self-employment,

57:31 the results are almost exactly the same

57:34 at each of year 1,

57:36 year 2,

57:37 and year 6,

57:38 and they're also almost exactly the same

57:40 between the cash grant arm

57:42 and the micro franchising arm.

57:44 Whether you are invited to the micro franchising program

57:47 or you receive a cash grant,

57:49 you're a little more than 10% points more likely to have self-employment

57:54 in your portfolio,

57:56 starting that moment and lasting for years to come.

57:59 OK,

58:00 so it's a lasting,

58:00 it's a persistent effect,

58:01 6 years.

58:04 Now,

58:04 how does this

58:06 work out in terms of your portfolio of income generating activities?

58:09 I only talked about self-employment.

58:11 So in the top bar we have

58:13 uh the number

58:15 of income generating activities that you do.

58:17 And in the bottom set of results

58:20 we have

58:20 uh the impact on paid work,

58:23 so

58:23 working for somebody else as opposed to the self-employment.

58:25 In that first year,

58:26 let's look at the bottom left corner first.

58:28 In that first year,

58:29 whether it's the franchise program or the grant,

58:32 you get this thing that helps you do self-employment and you're less likely

58:35 to do wage employment

58:36 because you've got this other thing going on,

58:38 right?

58:39 Uh,

58:40 that doesn't last,

58:42 OK?

58:42 It doesn't

58:43 in a persistent way,

58:45 it does not crowd out wage employment.

58:47 Uh,

58:48 and you can see that pattern,

58:49 uh,

58:49 sort of in the top panel,

58:50 in the top middle bar,

58:51 you can see that by year two,

58:53 particularly in the franchise group,

58:55 people

58:55 have a larger number

58:57 of

58:57 income generating activities in their income generating activity portfolio.

59:02 OK,

59:03 so

59:03 I sell shoes,

59:04 but I also do hairdressing,

59:06 whereas in the comparison group,

59:07 I only sell shoes

59:08 or something like that.

59:11 Um,

59:13 We can talk about whether you're exclusively self-employed.

59:16 That's the top row here,

59:18 and that's

59:19 something that we see an impact on in the first couple of years,

59:21 but that does start to diminish.

59:24 OK,

59:24 so you start out getting excited about this new self-employment thing and that

59:27 that drops off as the only thing that you've got going on,

59:30 as we can see in the bottom right.

59:33 It looks like

59:34 self-employed and working for others is something that gradually

59:37 becomes more likely if you're in the franchise group,

59:39 for example.

59:40 So this is adding to the set of things that you can do,

59:43 and at first you do mainly that and then it's just one of the things that you can do.

59:49 When you get a cash grant,

59:50 are you gonna stop working?

59:52 You're gonna work harder.

59:54 In our population,

59:56 that green bar on the left,

59:57 people who get that cash grant are working hard

1:00:00 to make their new businesses exist.

1:00:03 Uh,

1:00:03 in fact,

1:00:04 they're working more hours than,

1:00:05 uh,

1:00:06 the people who get the micro franchising intervention.

1:00:09 Maybe because there's more work to do.

1:00:11 They have to figure out their business model,

1:00:12 figure out where to source all the capital,

1:00:14 things like that.

1:00:15 Um,

1:00:16 but those effects are not persistent.

1:00:18 Something that I'm sure everyone wants to know about is the effect on income.

1:00:21 I'm showing you cumulative distribution functions,

1:00:24 and

1:00:24 the simplest thing I can say about the

1:00:26 picture on the left is that.

1:00:28 The gray bar being above the blue and green bars at the very beginning,

1:00:32 means that people in the control group are more likely to earn almost nothing.

1:00:37 Than people in either the cash grant

1:00:39 or the micro franchising group,

1:00:40 but that sharp distinction that is there in the first panel

1:00:44 starts to fade away as we get to year 6.

1:00:46 So

1:00:47 in a regression table framework that some of you may like,

1:00:50 uh,

1:00:51 and others of you may not,

1:00:52 um,

1:00:53 we see that

1:00:54 uh

1:00:55 you're earning,

1:00:56 this is in Kenya shillings,

1:00:57 so you divide by 100 to get to dollars,

1:00:59 uh,

1:00:59 people in the franchise and grant groups are earning

1:01:02 a few dollars more per week in the first year

1:01:05 relative to about $5 a week that the control group is earning in that year.

1:01:09 OK,

1:01:09 it's statistically significant.

1:01:10 People are earning more in the first year.

1:01:12 That is no longer true at years 2 and 6.

1:01:17 The income effect goes away.

1:01:19 And

1:01:20 we might think maybe that's just a thing about averages.

1:01:23 Sometimes averages mask variation in the distribution.

1:01:26 We saw some quantile treatment effects.

1:01:28 So we looked around the distribution

1:01:30 here.

1:01:30 This is a difference in CDFs.

1:01:32 And again,

1:01:32 the simplest thing I can say is any place the shaded interval is below the line,

1:01:37 it means that's a place where it has made incomes better for the treatment group.

1:01:41 So

1:01:41 the franchise treatment

1:01:42 really seems to make things better for people with low incomes.

1:01:45 They end up with higher incomes in that first year.

1:01:48 The grant treatment seems to have impacts all across the distribution,

1:01:51 including at the upper right,

1:01:52 the right end of the distribution.

1:01:53 So there's some

1:01:54 pretty successful businesses,

1:01:56 but when we look at years 2 and 6.

1:01:59 There's really no place in the distribution.

1:02:01 You can kind of squint and see one or two little blips,

1:02:04 but there's not really much variation

1:02:06 that the average is masking.

1:02:08 There's not a huge persistent effect on income

1:02:10 somewhere that the average doesn't let you see.

1:02:12 So this is a program.

1:02:13 These are both interventions that help

1:02:15 incomes in the short term,

1:02:16 but that income effect doesn't last.

1:02:19 Um,

1:02:20 one thing that people say about,

1:02:21 uh,

1:02:22 as an example of

1:02:24 women starting businesses rather than men starting businesses,

1:02:26 sometimes

1:02:27 the husband is the problem

1:02:29 and is a constraint on women's business operation,

1:02:32 but most of these women are not married

1:02:34 when they start these businesses.

1:02:35 It's not really the husband's fault.

1:02:37 That's not the mechanism here,

1:02:38 OK?

1:02:39 So women launch businesses,

1:02:41 do they get rid of them entirely?

1:02:42 No,

1:02:43 um,

1:02:44 and then we ask.

1:02:46 Uh,

1:02:47 so if you don't get rid of them entirely and if you are,

1:02:49 you know,

1:02:50 not earning more income,

1:02:51 uh,

1:02:52 what's happening here?

1:02:53 And so

1:02:54 we were brought to ask questions about well-being,

1:02:57 some of which we saw in one of the earlier presentations.

1:03:00 We ask about living conditions,

1:03:02 we ask about food security,

1:03:04 we ask about subjective well-being and happiness now

1:03:06 and what you expect for the future.

1:03:09 Uh,

1:03:10 and what I've got in a table format here

1:03:12 is

1:03:13 a pretty simple idea.

1:03:15 At year 2 we asked this and at year 6 we asked this.

1:03:18 At year 2 we didn't really see any significant effects,

1:03:20 uh,

1:03:21 to speak of on well-being,

1:03:23 but by year 6,

1:03:25 the franchise intervention that had the mentorship and the

1:03:27 life skills training and all of that stuff.

1:03:29 That left people feeling better about their lives,

1:03:32 in some cases in concrete ways and in other cases

1:03:36 in less concrete ways.

1:03:38 So

1:03:38 one lesson we took from this is that it's valuable

1:03:41 to look not only at income and sector of employment

1:03:45 but at how people feel about how they're living in their lives.

1:03:48 It seems like the micro franchising intervention helped people

1:03:52 see themselves

1:03:53 as micro entrepreneurs and know that that's a difficult path to

1:03:56 go on and understand that the struggle that they're on is

1:03:59 is one that is part of the business project that they're

1:04:02 doing and it doesn't mean there's something wrong with them.

1:04:05 Um,

1:04:05 so people see themselves

1:04:07 as micro entrepreneurs more

1:04:09 in the franchise group

1:04:11 and not

1:04:11 in the cash grant group.

1:04:13 So,

1:04:14 uh,

1:04:15 there are a variety of mechanisms you could explore more,

1:04:17 but for those of you doing projects along these lines,

1:04:20 we encourage you to not only look at long term effects

1:04:22 but to look beyond

1:04:23 income and sector of employment.

1:04:25 So I'll leave it there.

1:04:27 Thanks very much.

1:04:32 Great thank you very much for excellent presentations and uh good discipline

1:04:37 on timing so that leaves us a good um 20 minutes for discussion

1:04:43 we heard some

1:04:44 promising interventions we heard some.

1:04:49 Not terribly promising interventions,

1:04:51 at least as far as

1:04:52 labor market outcomes are concerned.

1:04:55 And some concern about scalability.

1:04:57 So

1:04:58 let's open it up to questions,

1:05:01 comments from the floor.

1:05:04 Show of hands,

1:05:05 I have one there.

1:05:07 And then I'll take,

1:05:09 um,

1:05:09 we'll do a second round,

1:05:10 234 right

1:05:12 behind each other there.

1:05:14 Thank you,

1:05:15 uh,

1:05:15 great,

1:05:15 another great session,

1:05:16 uh,

1:05:17 thanks for,

1:05:18 for these insights.

1:05:18 Uh,

1:05:19 question on both the last and the first paper on,

1:05:21 on Owen,

1:05:22 uh,

1:05:23 do the,

1:05:23 how much of the franchises still exist,

1:05:25 uh,

1:05:26 at year 6,

1:05:28 since that's part of your interpretation there and then.

1:05:30 Any,

1:05:32 since this is about income diversification,

1:05:34 any way that you can look at whether this helps cope with shocks or manage shocks,

1:05:39 and that that's where some of the benefits are coming from,

1:05:41 not on average,

1:05:42 but on in this moving part.

1:05:44 Um,

1:05:44 and then,

1:05:45 but on,

1:05:45 on,

1:05:46 on the first one,

1:05:46 super interesting,

1:05:47 um,

1:05:48 it's

1:05:49 really more a question of interpretation.

1:05:50 So the,

1:05:51 the,

1:05:53 do you know why they kind of allocated to the,

1:05:55 to the youth and not to the husbands,

1:05:58 uh,

1:05:58 and then,

1:05:59 uh.

1:06:00 Uh,

1:06:00 uh,

1:06:01 you know,

1:06:01 and,

1:06:02 and,

1:06:02 but,

1:06:02 so,

1:06:02 so,

1:06:03 so,

1:06:03 so you said so optimistic.

1:06:04 So,

1:06:04 it sounds like it's,

1:06:05 you know,

1:06:06 maybe we should wait

1:06:07 kind of in light of,

1:06:08 of what we heard,

1:06:09 but it is like,

1:06:10 uh,

1:06:10 but kind of you are optimistic about it,

1:06:12 so maybe tell us a little bit why you think that was a,

1:06:15 you know,

1:06:15 why that's a positive outcome.

1:06:18 Thank you.

1:06:19 Let's go over there.

1:06:21 Thank you all for these really

1:06:23 interesting presentations.

1:06:24 I have questions for

1:06:26 Owen and Isaac on

1:06:27 sector

1:06:28 and um

1:06:31 I guess speculation about generalizability,

1:06:33 so.

1:06:34 Um,

1:06:35 I know you don't have

1:06:36 data on what this,

1:06:38 what would have happened if you had done this in different sectors,

1:06:40 but I'm wondering if you

1:06:42 know based on how these,

1:06:44 the hairdressing sector,

1:06:45 the tailoring sector compares

1:06:47 to other sectors that,

1:06:49 um,

1:06:50 young people and women in particular are involved in,

1:06:52 whether you think,

1:06:53 how much of this do you think is sector specific

1:06:56 and how much do you think might generalize to other sectors

1:06:59 beyond those?

1:07:03 Hi there thank you so much for your presentation.

1:07:05 I'm Rachel,

1:07:06 a recent

1:07:07 master's graduate at Georgetown University and the

1:07:11 School of Foreign Service in their global human development program

1:07:14 so I was curious about,

1:07:15 so for micro entrepreneurship,

1:07:17 um,

1:07:18 from your research and

1:07:20 things maybe even working on what's the idea in terms of scaling up

1:07:24 so we are talking about providing jobs for the unemployed and the youth,

1:07:27 but how do we move these micro entrepreneurs to become SMEs or even larger.

1:07:32 And where do investors come in?

1:07:33 How do we attract investment,

1:07:34 you know,

1:07:35 are there any angel investor net investment networks you've discovered in Africa?

1:07:39 So just thinking about that scaling up so that we can try to solve this problem of,

1:07:43 um,

1:07:43 youth employment.

1:07:47 Hi,

1:07:47 thank you so much,

1:07:48 uh,

1:07:48 Farhan Majeed from USA um and you know.

1:07:52 Pennsylvania.

1:07:53 Um,

1:07:54 I found all of the presentations really fascinating,

1:07:56 but I just want to comment on the first and the last right now.

1:08:00 So for the first presentation,

1:08:01 I was thinking if

1:08:02 you've thought about the abilities

1:08:04 of the household members,

1:08:06 and,

1:08:07 you know,

1:08:07 like you,

1:08:08 you can think of this as a response,

1:08:10 a household response to abilities and thinking about

1:08:14 what are the efficiency versus equity motives here.

1:08:17 So you're just thinking from a youth versus non- youth,

1:08:19 but I'm just curious about how is that playing out.

1:08:21 And if there's a way for you to measure

1:08:23 preferences and

1:08:25 um

1:08:26 Work with that.

1:08:27 Um,

1:08:27 and for the

1:08:29 last presentation,

1:08:30 I was just thinking,

1:08:30 given the age group of 18 and 19,

1:08:33 and the time period,

1:08:34 if there's um

1:08:36 Um,

1:08:37 delays in like childbirth or like marriage patterns as well,

1:08:41 which you noticed.

1:08:42 Thank you.

1:08:44 Thank you.

1:08:45 Why don't we start with you.

1:08:47 So the first question on

1:08:50 why people switch to the younger people as opposed to the household head,

1:08:54 one of the things.

1:08:56 That was built into the program was

1:08:59 given this age brackets

1:09:02 for it,

1:09:02 so 15 to 45,

1:09:04 and I think while that still allowed many of the caregivers to be in,

1:09:08 um,

1:09:09 it excluded a lot of their spouses because they're typically a large

1:09:12 age differential between the spouses so I think they were just not eligible

1:09:16 so that's probably something to keep in mind for other types of programs if we're

1:09:20 if we're concerned about kind of capture in them yeah

1:09:24 and the second question about.

1:09:26 Um,

1:09:28 the ability of different household members,

1:09:30 I think this is important because.

1:09:32 I guess there are two ways to think about it.

1:09:33 One is

1:09:35 if you have the high ability individual,

1:09:37 then maybe you do want them to be the one

1:09:39 to have the opportunity so that they can produce,

1:09:41 um,

1:09:42 they,

1:09:42 they're more productive,

1:09:43 and then the household can benefit from there.

1:09:46 And then there's equity concern that if they're

1:09:48 the ones that get all the opportunities,

1:09:50 then what about the

1:09:51 less,

1:09:51 um,

1:09:52 able household members?

1:09:54 I think that's,

1:09:55 I mean one of the things that we

1:09:57 want to dig more into both,

1:09:59 um.

1:10:00 Some of the data we have now and in future is to

1:10:03 look more at the specific characteristics of these

1:10:06 household members so

1:10:08 wanting to learn more about

1:10:10 if they've previously

1:10:11 run any kind of enterprise with the education

1:10:14 levels or other measures that we have of what

1:10:17 features of the ability might be predicting their performance.

1:10:20 I think that would guide us a little bit,

1:10:22 but I think at the end,

1:10:23 um.

1:10:24 For

1:10:26 this type of setting I think there tends to be a

1:10:30 shift in in prioritizing the

1:10:32 efficiency

1:10:34 argument in the hopes that

1:10:36 this then will be redistributed

1:10:38 across the household members.

1:10:42 I'll just speak to the question of scale that uh or scaling up that that Rachel asked.

1:10:46 I think that's an,

1:10:47 uh,

1:10:47 an interesting challenge,

1:10:48 and I think it's,

1:10:49 uh,

1:10:49 others might have different views,

1:10:51 but I think

1:10:52 most of the types of interventions that we're talking about here have

1:10:55 modest or no effect on people's ability

1:10:57 to employ others outside their own household.

1:11:00 So

1:11:00 this is,

1:11:01 these are,

1:11:01 you know,

1:11:01 at best moving people from

1:11:04 self-employment businesses that don't exist or

1:11:06 doing very little for their household income

1:11:08 to things that are doing a bit more for their household income,

1:11:10 um.

1:11:11 And I think the question of where might

1:11:14 demand for wage for wage employment come from,

1:11:17 where where

1:11:18 where might that sit is a really crucial one.

1:11:21 The answer might not be that it comes from

1:11:23 starting from really small businesses and growing them into bigger ones,

1:11:26 but thinking about businesses that start on a different level in the first instance

1:11:31 with a level of capital intensity and

1:11:32 things like that that these businesses don't have

1:11:34 and think of that as a place where wage demand might

1:11:37 be or demand for wage labor might be generated instead.

1:11:42 So,

1:11:42 um,

1:11:43 I had a question about,

1:11:44 uh,

1:11:44 the general

1:11:46 generalizability of,

1:11:47 uh,

1:11:47 some of the findings beyond sort of hairdressing and tailoring,

1:11:50 I think from Kehindi,

1:11:51 um.

1:11:53 I,

1:11:53 yeah,

1:11:53 I mean I think as long as uh one can develop a good test,

1:11:57 at least,

1:11:57 uh,

1:11:58 you know,

1:11:58 if I can develop a masonry test and a carpentry test,

1:12:02 I think um the sort of theory of change would be

1:12:05 um

1:12:06 applicable and um again we'll have to test to see whether it works but

1:12:10 I

1:12:11 don't see any issues there particularly that I can think of but

1:12:15 I think one would wanna test it first,

1:12:17 um,

1:12:17 and create a good test.

1:12:19 Um,

1:12:19 and just to sort of piggyback on what Andy was saying about the,

1:12:23 the scaling,

1:12:23 I think the other thing that I think is challenging is that

1:12:27 many of these,

1:12:28 many people in self-employment are not there by choice.

1:12:32 And I think

1:12:33 if you were to find people who have great ideas,

1:12:36 brilliant ideas,

1:12:37 who want to be entrepreneurs,

1:12:40 identifying these types of people that can actually grow and sort of,

1:12:43 you know,

1:12:44 change,

1:12:45 you know,

1:12:45 bring in new ideas,

1:12:46 bring in new products that can actually people want to buy and can then,

1:12:50 you know,

1:12:51 that that's,

1:12:51 that's,

1:12:52 that's really hard to find those types of people,

1:12:54 um,

1:12:55 so,

1:12:55 um,

1:12:56 yeah,

1:12:56 I think that's one challenge

1:12:57 is

1:12:58 who goes into self employment.

1:13:01 Uh,

1:13:01 exactly.

1:13:02 So,

1:13:03 uh,

1:13:03 let me say,

1:13:04 I'll,

1:13:05 I'll answer the question that Tamar didn't ask.

1:13:07 We heard some,

1:13:08 some good news and some not so good news,

1:13:10 I think.

1:13:11 But what's funny about the not so good news

1:13:13 is it seemed like better news the first year.

1:13:17 OK,

1:13:17 there were income effects in the program that I looked at and

1:13:19 in the grants that I looked at in the first year.

1:13:22 And so for a lot of interventions,

1:13:24 including some here,

1:13:25 we so far only have evidence on the good news part of the program.

1:13:28 So

1:13:29 how these things evolve,

1:13:30 and I think how many years is the

1:13:32 tailoring stuff that you've done?

1:13:34 You're participant this long.

1:13:35 How many years is that now?

1:13:36 Yeah,

1:13:36 I mean,

1:13:36 we,

1:13:36 this,

1:13:37 the 20,

1:13:38 yeah,

1:13:38 we started 2013,

1:13:39 yeah,

1:13:39 to 2022.

1:13:40 So some of this is quite long,

1:13:42 but I think these things do have dynamics.

1:13:44 So that's the first thing I wanted to point out.

1:13:46 Um,

1:13:46 I agree completely with,

1:13:47 uh,

1:13:48 Andrew and Isaac's answers to the question of scale

1:13:50 and how to find ways of scaling this in the

1:13:53 project that I've done as well.

1:13:54 Very few of these young women go on to employ someone else.

1:13:57 So then if you're asking where can you find

1:13:59 the partners to do these kinds of things with,

1:14:01 that's a difficult question.

1:14:02 I agree completely with what you said.

1:14:04 Uh,

1:14:05 to Karen's question about,

1:14:06 uh,

1:14:07 are the micro franchises still around,

1:14:08 one way of answering that is to say,

1:14:10 what is it that the young women who are still running self-employment businesses,

1:14:15 what are the sectors in,

1:14:16 and the ones who started in the

1:14:17 mic in the hair salon micro franchise.

1:14:20 are much more likely to still have a hair salon as one of their businesses,

1:14:23 so to some extent

1:14:25 they

1:14:25 don't close,

1:14:26 but they decrease the intensity of their work.

1:14:28 It's something that they've got as an option in their portfolio.

1:14:30 I think it must help them cope with shocks,

1:14:32 though I don't have a sort of specific measure of

1:14:36 shock coping,

1:14:37 but their living conditions are a little bit better,

1:14:40 and they report greater food security.

1:14:42 So in some way that suggests

1:14:43 that yes,

1:14:44 it must.

1:14:47 On generalizability to Kehinde's question,

1:14:49 so,

1:14:50 uh,

1:14:50 which by the way

1:14:51 is almost exactly an interview question I was once asked

1:14:54 when I was applying for a job as an economist.

1:14:57 Anyway,

1:14:58 that question we have a little bit of evidence on because we have at least two sectors

1:15:02 and

1:15:02 though it's not randomized which of the two,

1:15:06 micro franchise models you're in,

1:15:08 uh.

1:15:09 There's variation both in what preferences people express,

1:15:12 so we can exploit that heterogeneity,

1:15:14 and there was differential availability of the different models across

1:15:17 some rounds of rollout

1:15:19 and across either of those ways of looking at it,

1:15:21 a kind of more exogenous way or a more endogenous way.

1:15:24 We don't see.

1:15:25 Consistent evidence of differential impacts.

1:15:28 So it seems like the two models were very similar in

1:15:31 what limited evidence we have.

1:15:33 So that suggests that we could go further.

1:15:36 The theory of change isn't specific

1:15:38 to either of those things,

1:15:39 but

1:15:40 what sector would you come to next?

1:15:41 I don't know.

1:15:44 Uh,

1:15:45 Farhan asked about,

1:15:46 uh,

1:15:46 marriage and childbirth patterns.

1:15:47 I don't think there's any impact that we find on eventual household size

1:15:51 or,

1:15:52 uh,

1:15:53 which is,

1:15:54 or household configuration,

1:15:55 so

1:15:56 no effects

1:15:57 there,

1:15:59 um,

1:15:59 yeah.

1:16:00 Let's take a few more.

1:16:02 I saw a hand up there before.

1:16:04 Um,

1:16:05 OK,

1:16:06 there's

1:16:07 1234.

1:16:09 Um

1:16:13 Hello,

1:16:14 uh,

1:16:14 I'm Priya.

1:16:15 So I'm a nonprofit and we actually work on the

1:16:19 micro entrepreneurship in India.

1:16:21 So,

1:16:22 uh,

1:16:23 from the last 10 years we're working with the women and the youth.

1:16:26 So we understood that when the first year

1:16:29 when you start the business,

1:16:31 everyone,

1:16:32 when we do the

1:16:33 as a cash,

1:16:34 you give a grant.

1:16:35 Sometimes they just come

1:16:36 to get the cash grant,

1:16:38 that's why they attend the training.

1:16:40 And after the attending the training,

1:16:41 many times it happened that

1:16:43 just,

1:16:43 you know,

1:16:44 uh,

1:16:44 that incentive towards the getting the cash

1:16:47 because of that they continue with the,

1:16:49 you know,

1:16:49 training part.

1:16:50 OK,

1:16:50 I want to do the business,

1:16:52 but,

1:16:52 you know,

1:16:53 at the next,

1:16:53 at the last step,

1:16:54 we understood that

1:16:55 many times this woman

1:16:57 actually

1:16:58 don't want to be an entrepreneur,

1:17:00 they just want to be,

1:17:01 have that cash grant,

1:17:03 and because of that they get involved.

1:17:06 So that is one thing which we face the challenge.

1:17:09 Second challenge was that,

1:17:11 you know,

1:17:12 many times it happened that.

1:17:14 The women and the youth,

1:17:15 when we talk about the both ways,

1:17:17 youth

1:17:18 start a business,

1:17:19 we help them to start,

1:17:20 and once they get a good job.

1:17:23 They switch to the job.

1:17:24 Once they're laid off from the job,

1:17:26 they can come to the,

1:17:28 you know,

1:17:28 the livelihood opportunity.

1:17:29 So that is one of the challenges with the youth.

1:17:31 So where we have to

1:17:32 work on the model,

1:17:33 what are the models

1:17:35 are good for the youth

1:17:36 because women and the youth,

1:17:38 we can't keep it on the together.

1:17:40 So I think

1:17:41 there needs to be a different model.

1:17:43 So,

1:17:43 yeah,

1:17:43 what is your point of view on that?

1:17:47 Thank you.

1:17:49 Hi,

1:17:49 um,

1:17:50 thank you so much for the presentation.

1:17:51 My name is Sonia,

1:17:52 and I am an undergrad student from the University of Notre Dame.

1:17:56 So my question is more about like the landscape about,

1:17:59 uh,

1:17:59 financial incentives like right now.

1:18:01 I was wondering like when we were talking about like financial incentives,

1:18:04 are we talking more about like grant or

1:18:06 we're talking more about like franchisement or are.

1:18:08 Are we moving like more towards like microfinance,

1:18:11 like micro loans,

1:18:12 something like that.

1:18:13 And also when we're talking about like offering those incentives to people,

1:18:16 are we,

1:18:17 like,

1:18:17 how do we make sure,

1:18:18 like how do we determine like how much are we offering people

1:18:21 and also like

1:18:22 are we offering to like individual like household or like who,

1:18:25 like who in the household.

1:18:27 So thank you.

1:18:31 Over there and then behind him.

1:18:34 Hello and uh thank you for your presentation.

1:18:37 Uh,

1:18:37 my name is Frederic Aubury.

1:18:38 Uh,

1:18:39 I'm working at the World Bank.

1:18:40 Uh,

1:18:41 I have uh,

1:18:41 a couple of questions for uh Isaac and BT about apprenticeships.

1:18:45 Uh,

1:18:45 the,

1:18:45 the first is

1:18:47 about the,

1:18:48 your opinion on other.

1:18:51 The advice you would give to a policymaker

1:18:53 in your uh experiment you are targeting

1:18:56 the intervention is targeting the masters

1:18:58 to indirectly,

1:19:00 uh,

1:19:00 target the apprentices,

1:19:02 and I was wondering

1:19:03 if you have any opinion about

1:19:05 policies that target the masters or

1:19:07 policies that would target apprentices directly.

1:19:10 Uh,

1:19:10 the reason why I'm,

1:19:11 I'm asking that is

1:19:13 because I'm,

1:19:14 I'm wondering to,

1:19:15 to what extent

1:19:17 the fact that the master is retaining apprentices.

1:19:20 Is telling us about they're

1:19:22 willing to retain

1:19:24 this productive employee,

1:19:26 or does that mean that

1:19:28 apprentices do not have outside option,

1:19:31 uh,

1:19:32 outside option outside of the workshop?

1:19:34 They,

1:19:34 they are lacking of opportunity to use their skills outside of the workshop.

1:19:39 So I was just

1:19:41 wondering whether it might make sense to target also apprentices in some way

1:19:45 to help them in their transition outside of the workshop.

1:19:48 Thank you.

1:19:49 And

1:19:49 last one there.

1:19:53 Thank you for your presentation.

1:19:55 My name is Aida Martinez.

1:19:56 I'm an intern in the World Bank and I'm a student in the Master of

1:20:01 Public International Development Policy at Duke University.

1:20:04 And well this is more common for Owen.

1:20:07 I was thinking that as well

1:20:08 that uh maybe the lack of willingness of

1:20:12 being an entrepreneur

1:20:13 uh

1:20:14 it was the reason to

1:20:16 look those uh fates and the effects of

1:20:20 both franchise and the grants

1:20:22 and

1:20:23 the second thing it was that,

1:20:25 uh,

1:20:25 women.

1:20:26 Uh tend to

1:20:28 like dusting more a larger share of their incomes to

1:20:33 uh household

1:20:34 expenses that could be another reason

1:20:37 and when you mention that uh

1:20:40 the well like the grand beneficiaries uh

1:20:44 were

1:20:45 not married,

1:20:47 uh,

1:20:47 I was thinking that

1:20:49 maybe is not the case for all of the participants of this uh

1:20:53 like RTT

1:20:54 but

1:20:55 um.

1:20:57 Economic violence is not only perpetrated by uh husbands but also for parents,

1:21:03 for sisters,

1:21:04 for

1:21:05 uh siblings in general

1:21:07 so I was thinking about that like maybe uh looking at that

1:21:11 uh could help to understand

1:21:13 a little bit more

1:21:15 why

1:21:16 those effects change.

1:21:18 Thank you,

1:21:18 thank you very much,

1:21:19 Isaac.

1:21:19 Why don't we start with you because there was a specific question to you.

1:21:22 Yes,

1:21:23 um,

1:21:23 so thank you so much for the question.

1:21:24 Um,

1:21:25 so,

1:21:25 um,

1:21:26 in terms of,

1:21:27 um,

1:21:28 you know,

1:21:28 how do you target

1:21:30 masters or train or apprentices and sort of how you think about the sort of the,

1:21:33 I guess the outside options,

1:21:35 what I would say is,

1:21:36 um.

1:21:38 There's a very complex sort of

1:21:40 uh

1:21:42 tradition I would say of you know determining when a

1:21:47 uh

1:21:47 against the traditional system right so you you

1:21:49 enter the apprenticeship and basically the you know the

1:21:52 apprent the master trainer basically has a lot of

1:21:56 power over

1:21:58 when you compete right.

1:22:00 Um,

1:22:00 and so,

1:22:02 and this is,

1:22:02 I think,

1:22:02 part of the,

1:22:03 I think the contractual sort of issues and the monitoring issues,

1:22:06 um,

1:22:07 you basically end up,

1:22:09 and again at the end of the day,

1:22:10 basically

1:22:11 they graduate,

1:22:12 so to speak,

1:22:13 and they will move away from the firm,

1:22:14 and there's a question of like,

1:22:16 OK,

1:22:16 when are they

1:22:18 going to graduate,

1:22:19 and the,

1:22:19 and the master is the one who actually determines that,

1:22:21 right,

1:22:22 so

1:22:22 what you see is basically a lot of training that just gets dragged out

1:22:27 for a really,

1:22:27 really long time.

1:22:28 And I think that this is evidence of a

1:22:31 part of this monopsonistic sort of um power that they have

1:22:35 over them to sort of control their lives because basically.

1:22:38 When you sort of say I did an apprentice,

1:22:40 I,

1:22:40 you need to have like like a reference letter from them,

1:22:42 right,

1:22:43 so they can't leave until they get that reference letter

1:22:45 and there are these sort of um trade

1:22:47 associations that are very powerful that sort of,

1:22:49 you know,

1:22:49 have these customs about so

1:22:51 so there's just a lot of I think um

1:22:53 um I think norms and sort of societal sort of structures here that I think,

1:22:58 um,

1:22:59 really I think

1:23:01 um.

1:23:02 Put

1:23:03 apprentices at a disadvantage where they end up sort of

1:23:06 staying in these really low paid apprenticeships for a long time

1:23:09 and the master trainer is able to basically extract a lot of rents,

1:23:12 um,

1:23:13 so yeah I think there are,

1:23:14 I think

1:23:15 ways to really speed up training to targeting the um

1:23:20 uh

1:23:20 trainer as we did.

1:23:21 Uh,

1:23:22 and I think,

1:23:23 yeah,

1:23:23 there are also probably complementary ways we can

1:23:25 sort of help the apprentice move faster,

1:23:27 but I think

1:23:28 one of the things we need to do is sort of systemic

1:23:30 where we basically need to sort of put some more regulation and oversight

1:23:34 to make sure we're not getting people in apprentices for 5 years,

1:23:37 right?

1:23:37 So I think

1:23:38 this is,

1:23:38 I think,

1:23:38 a role for government to really come in and sort of,

1:23:40 I think,

1:23:41 regulate the system a little bit more

1:23:43 and sort of put standards and,

1:23:44 um,

1:23:45 and sort of I think really just.

1:23:47 Upskill and upgrade this apprenticeship system a little bit more.

1:23:50 Thanks.

1:23:51 Let's move to

1:23:53 the rest of you to pick up on the questions around.

1:23:57 What if people really don't wanna be an entrepreneur,

1:23:59 um,

1:24:00 and,

1:24:01 and then some of the design elements that were brought up,

1:24:04 uh,

1:24:05 why don't you start Owen and then we'll go.

1:24:07 OK,

1:24:08 uh,

1:24:09 yes,

1:24:09 I'm not sure that people think that entrepreneurship is

1:24:11 the thing that they have always wanted to do,

1:24:13 uh,

1:24:14 but in an environment where there's really low labor demand from the wage sector,

1:24:18 uh,

1:24:19 it's helpful to have at least as a coping strategy if not

1:24:21 if not something more.

1:24:23 So

1:24:24 I think that's an accurate description of a lot of the people here.

1:24:26 I should say we didn't advertise cash grants.

1:24:28 We said if you're interested in starting a business

1:24:30 and then we said,

1:24:31 uh,

1:24:32 good and bad news.

1:24:33 The business training is full,

1:24:34 but

1:24:35 I do have something good,

1:24:36 and people didn't know that that was coming,

1:24:38 so I don't think we faced particularly adverse

1:24:40 selection of the type that you described,

1:24:42 but,

1:24:42 uh,

1:24:43 completely

1:24:43 fair point,

1:24:44 um.

1:24:46 With respect to the point about uh

1:24:49 the the roots of economic violence may not

1:24:50 exclusively originate in the husband and things like that

1:24:53 uh.

1:24:54 Completely agree and one of the things we did

1:24:56 that we were frustrated by is that we looked for

1:24:59 the

1:25:00 predictable heterogeneity.

1:25:01 We looked for baseline characteristics,

1:25:03 who the woman was residing with,

1:25:05 uh,

1:25:05 how much education she had,

1:25:07 a host of background characteristics,

1:25:08 and we used a variety of machine learning methods to ask,

1:25:12 does any of those background characteristics predict

1:25:14 whether this is gonna work out for her in the long run or not,

1:25:17 and we were unable to find a reliable predictor along those lines.

1:25:21 Uh,

1:25:21 so

1:25:22 I think it's still open to figure out what to do there,

1:25:25 and I forgot my

1:25:26 best anecdote in answer to

1:25:27 Kehinde's question.

1:25:28 We actually didn't come here originally to study,

1:25:31 uh,

1:25:32 hairdressing

1:25:32 and,

1:25:33 uh,

1:25:33 food carts.

1:25:34 The original micro franchising model that the IRC had

1:25:37 done in West Africa and was really happy about

1:25:39 was a refrigerated goods micro franchise called

1:25:43 Ice Ice Baby.

1:25:45 So I think there's at least one more sector out there,

1:25:47 Andy.

1:25:49 Um,

1:25:49 very quickly,

1:25:50 yeah,

1:25:50 very briefly,

1:25:50 I think the,

1:25:51 the

1:25:52 Priya,

1:25:52 your,

1:25:52 your anecdote about,

1:25:54 um,

1:25:55 conditionality,

1:25:55 I think is an interesting one.

1:25:56 I think speaks to,

1:25:58 uh,

1:25:58 an important challenge.

1:25:58 So one thing is,

1:25:59 you know,

1:26:00 do people wanna become an entrepreneur,

1:26:01 and then the other thing is kind of once you're on a path toward that,

1:26:04 you know,

1:26:05 should you

1:26:06 force them or,

1:26:07 you know,

1:26:08 economically incentivize them in strong ways to go through particular,

1:26:11 you know,

1:26:11 trainings targeted at particular things,

1:26:13 um,

1:26:13 I think

1:26:14 that

1:26:15 model is predicated on the idea that there's some market.

1:26:18 Market failure out there that people don't know that this

1:26:19 training is really going to be great for them,

1:26:20 don't realize the value of it.

1:26:22 That may be true in some circumstances.

1:26:24 I think the evidence we're seeing in a variety of these

1:26:27 studies is that those benefits are modest

1:26:30 and that has

1:26:31 therefore a double tax,

1:26:32 right?

1:26:32 You're both spending money

1:26:34 from the organization's perspective on things that aren't,

1:26:36 you know,

1:26:37 that could be cash that does good things for somebody else

1:26:39 and you're taking up people's time that they could be putting into the

1:26:42 enterprise itself or other things that they might enjoy in their life.

1:26:45 So I think that's worth thinking really carefully about.

1:26:48 Right,

1:26:50 so

1:26:50 there was a question about how the financial amounts are set,

1:26:53 how much to offer.

1:26:54 I think

1:26:55 I was looking at the range of amounts.

1:26:57 It's somewhere between $200 to $300 tends to be the amount,

1:27:00 and I think it's

1:27:01 based on what

1:27:02 could be used to start up a micro enterprise in the setting,

1:27:07 um.

1:27:08 It's very I completely agree with

1:27:12 the other panelists that

1:27:14 this is really a stopgap measure in many cases

1:27:16 when I was making the title of my slide when I wrote the choice of entrepreneur,

1:27:20 I chuckled to myself that

1:27:22 no really these are not entrepreneurs,

1:27:24 these are,

1:27:25 you know,

1:27:26 some kind of sustain like sustenance subsistence,

1:27:29 um,

1:27:30 livelihood activities

1:27:31 and

1:27:32 you know I when I always think of this big problems of youth unemployment

1:27:36 I.

1:27:37 The questions are much bigger than I feel like these are

1:27:40 all micro solutions to

1:27:42 a macro problem.

1:27:44 There are no jobs.

1:27:45 That's why we have all of these programs

1:27:47 that are that are at play and

1:27:49 you know we're talking about the Bangladesh

1:27:52 story early and I feel like if all of us could get that.

1:27:55 In our countries like

1:27:57 people that can employ en mass young people

1:28:00 and um really find ways to

1:28:02 to address these you know this pyramid that all of us are looking at and

1:28:06 and very worried about so

1:28:09 um that's a very good place to

1:28:11 end this discussion.

1:28:12 Thanks to all of you and thanks very much to the panel.

1:28:15 A round of applause please.

1:28:43 OK.

1:28:44 I

1:28:46 Oh,

1:28:47 OK,

1:28:47 they don't have to.

1:28:49 All right.

1:28:51 Should we?

1:28:53 Uh,

1:28:54 Uh,

1:28:55 I was told that would be an issue,

1:28:56 so I'm fine.

1:28:57 I'll just sit here.

1:28:59 OK,

1:29:00 all right,

1:29:01 um,

1:29:01 I wanna ask how many of you,

1:29:04 uh,

1:29:05 participated in the discussions yesterday as well?

1:29:09 OK,

1:29:10 quite a,

1:29:11 quite a few,

1:29:12 all right,

1:29:13 um.

1:29:14 I wanna thank uh the Center for Global

1:29:17 Development in particular for giving me this opportunity

1:29:20 not just to moderate this wonderful panel but also to give uh closing remarks.

1:29:26 And I was really wondering yesterday

1:29:30 after hearing the discussions,

1:29:32 what I can bring

1:29:33 to this audience by way of closing remarks

1:29:36 at the end of the two full days of

1:29:40 amazing presentations,

1:29:41 big ideas,

1:29:42 fruitful discussions,

1:29:45 and I thought I would lean on my comparative advantage.

1:29:49 And you might ask what is that?

1:29:51 And I would respond by saying I am retired,

1:29:55 so

1:29:56 I have been away from these conversations for almost 10 years

1:30:01 and what that does for me is that it gives me some perspective,

1:30:05 um,

1:30:06 it also allows me to plead ignorance because I will say

1:30:10 some things that I don't know a great deal about.

1:30:13 So what I want to do is I want to talk about what I heard

1:30:17 during these two days,

1:30:20 uh,

1:30:20 and especially today

1:30:22 and reflect on

1:30:23 how different.

1:30:25 What I heard was from what I knew,

1:30:28 what the discussions were 10 years ago,

1:30:31 where I was in the midst of these conversations.

1:30:34 So there was a lot of discussions on challenges and what I heard on that side

1:30:40 is that uh many of the issues facing the global community in

1:30:45 the developing world today

1:30:47 are not entirely new,

1:30:49 but they are definitely qualitatively different.

1:30:53 Yesterday's presentations focused on these big issues.

1:30:58 Digital technology

1:31:00 that was definitely around 10 years ago,

1:31:03 but it is much more ubiquitous today

1:31:06 and AI in particular

1:31:08 promises to transform

1:31:10 our lives,

1:31:11 our societies,

1:31:12 and our economies in ways

1:31:14 that could not be imagined 10 years ago.

1:31:18 Second,

1:31:19 the climate agenda

1:31:20 again,

1:31:21 we have had warnings about the climate crisis for many years

1:31:26 unheeded,

1:31:27 but I have a sense that the urgency of the climate agenda now

1:31:32 has become

1:31:33 a mainstream topic,

1:31:35 and there are

1:31:36 constant calls for action on raising

1:31:40 trillions of dollars

1:31:41 and changing the trajectory of economic production.

1:31:46 Third is that all these,

1:31:48 this is happening,

1:31:49 these two major drivers of change,

1:31:53 uh,

1:31:53 they're happening at a time when we have historically

1:31:57 low growth rates globally.

1:32:00 And in a geopolitical context

1:32:03 where many OECD countries have turned inward.

1:32:08 And both those things severely curtail

1:32:11 the opportunities for low-income countries to grow

1:32:14 through trade.

1:32:15 That has been the

1:32:16 traditional path as demonstrated by East Asian countries

1:32:21 to growth into middle-income and higher-income status,

1:32:24 and also to take advantage of migration because the

1:32:27 same inward-looking policies on the part of wealthy countries

1:32:30 are also closing off those opportunities.

1:32:34 On top of this is the accumulating debt burden

1:32:38 for many developing countries that we heard about yesterday.

1:32:42 Which ends up reducing the fiscal space for investment and infrastructure

1:32:47 and the social sectors.

1:32:50 So that's for the challenges.

1:32:53 What did I hear on the side of solutions,

1:32:56 and I'll focus on today.

1:32:59 Um,

1:33:00 and today was meant to focus on,

1:33:02 as I said,

1:33:02 on the economic inclusion.

1:33:05 Of

1:33:05 especially women

1:33:07 and youth.

1:33:10 A lot of the conversations that we had today,

1:33:12 the discussions at the panels

1:33:14 are topics that I was very familiar with in many ways the same,

1:33:18 including creating jobs through beauty salons and cosmetology.

1:33:23 Um,

1:33:24 And

1:33:26 many

1:33:26 of the discussions

1:33:29 focused on

1:33:30 reducing the constraints

1:33:33 to the labor market participation of women

1:33:37 through various means,

1:33:39 uh,

1:33:39 including

1:33:40 changes in societal norms

1:33:43 and,

1:33:44 um,

1:33:45 through safety nets,

1:33:46 cash transfer programs,

1:33:49 uh,

1:33:49 credits.

1:33:52 And

1:33:54 outside

1:33:54 encouraging labor,

1:33:56 greater labor force participation on the part

1:33:58 of women by reducing these constraints,

1:34:01 we had the discussion most notably in my panel today

1:34:04 on how to encourage our programs to foster more entrepreneurship.

1:34:11 Entrepreneurship,

1:34:13 as we all ended up saying may be

1:34:16 too big of a board for what's going on.

1:34:18 It's really,

1:34:20 uh,

1:34:20 self-employment and household businesses that we're talking about

1:34:24 and

1:34:25 both of these types of interventions,

1:34:27 removal of barriers as well as

1:34:30 working

1:34:31 on

1:34:32 the small micro enterprises

1:34:35 really take labor demand

1:34:37 as a given.

1:34:39 And in the constant in the context of a given labor demand,

1:34:44 what can we do to encourage the economic inclusion

1:34:49 of women

1:34:50 and youth

1:34:50 and these are the

1:34:52 conversations that we had today,

1:34:54 some incredible creative experimentation.

1:34:59 That ultimately.

1:35:02 I feel may not move the needle.

1:35:06 On the discussion around women,

1:35:08 I think there are some big ideas there.

1:35:10 If we're able to change societal norms,

1:35:14 if we're able to remove some of the barriers as in providing

1:35:18 childcare services,

1:35:19 for example,

1:35:20 to women that we heard

1:35:21 that would encourage them,

1:35:23 enable them to participate more,

1:35:25 I can see the potential for scale and economy-wide impacts.

1:35:30 I have a difficult time doing the same for the

1:35:32 types of interventions that we talked about in this panel.

1:35:36 They are important,

1:35:37 as we said,

1:35:38 because for the moment at least

1:35:40 these are the only things that are available

1:35:42 to poor people that don't have access to

1:35:45 many opportunities and certainly not wage labor.

1:35:49 So,

1:35:51 Um,

1:35:52 two main takeaways for me.

1:35:55 I think we need to think more boldly

1:35:57 and more creatively

1:35:59 about addressing

1:36:01 the challenges of the day

1:36:03 which are the ones that I talked about in terms of the.

1:36:06 Uh,

1:36:07 technological improvements and the climate agenda

1:36:12 we have learned tremendous amount through a portfolio of impact evaluations

1:36:17 we have learned how to think about value for money,

1:36:21 which is,

1:36:22 um,

1:36:22 what,

1:36:23 uh,

1:36:23 uh,

1:36:24 Rachel has been emphasizing.

1:36:26 And these are all very important learnings,

1:36:29 but I think the challenges of today

1:36:32 require us

1:36:33 to get outside our comfort zones and think more boldly.

1:36:41 I think there's significant room for enhancing our knowledge

1:36:46 about what economic inclusion will look like

1:36:50 in a world that's dealing with the climate change

1:36:53 and the promise and threats of digital technology.

1:36:57 How

1:36:58 can AI technology and innovation more generally

1:37:02 be harnessed to improve labor productivity,

1:37:06 especially for low-skilled workers?

1:37:09 Can we think about experiments that demonstrate how AI can complement

1:37:14 rather than supplant current workers

1:37:18 in the education sector,

1:37:19 in agriculture,

1:37:20 in the health sector,

1:37:21 in social work.

1:37:24 In the area of the green economy

1:37:27 and.

1:37:28 I really don't know what that means,

1:37:29 and I found out that

1:37:32 many people don't know what that really means,

1:37:34 but it is a thing.

1:37:37 What scope is there for low income countries to

1:37:40 be part of the supply chain for some products?

1:37:45 If not solar panels and batteries,

1:37:47 which I was disabused of yesterday,

1:37:50 something else perhaps,

1:37:51 I don't know exactly what.

1:37:54 What about um investments in adaptation

1:37:57 to reduce the costs of climate induced disasters in low income countries,

1:38:03 same countries that are paying for the costs

1:38:06 of a problem

1:38:07 that they really had no part in producing.

1:38:11 Experiments to produce

1:38:14 heat resistant crops,

1:38:15 I think Rachel mentioned this yesterday,

1:38:18 adapt infrastructure to become more resilient to climate change.

1:38:24 There was also an interesting panel yesterday on the

1:38:27 revival that industrial policy seems to be experiencing.

1:38:33 And this exploring the scope for more activist government policies to spur growth.

1:38:39 What I did not hear

1:38:41 in that panel in that discussion

1:38:44 was any attention to distributional concerns.

1:38:48 So if innovation,

1:38:50 the green economy,

1:38:52 and

1:38:53 some new form of industrial policy

1:38:55 are to produce gains for the poorest segments of society,

1:39:00 for youth and for women.

1:39:02 I think that people that worry about these concerns have to raise these issues

1:39:08 and be part of those conversations.

1:39:12 So my second point is that I believe this agenda,

1:39:15 distributionally sensitive innovation,

1:39:18 climate transition,

1:39:18 and new industrial policies

1:39:20 need to form a big part of future research.

1:39:25 For people who

1:39:27 worry about poverty reduction,

1:39:29 jobs,

1:39:29 and human capital.

1:39:31 I think there's a significant need for generating evidence on what works

1:39:36 and learning through experimentation.

1:39:39 And for this to happen,

1:39:41 I really think there needs to be a much closer collaboration between

1:39:45 macro and microeconomists.

1:39:48 In other words,

1:39:49 more coherence,

1:39:51 the word of the conference

1:39:53 between day one and day two of this conference.

1:39:57 I was expecting

1:39:59 to see more of a connection between the issues raised on day one.

1:40:04 And what we talked about today,

1:40:07 I mean,

1:40:07 Rachel touched upon

1:40:09 why it is that we're talking about

1:40:11 human capital

1:40:13 today,

1:40:13 which is an intributor contribute uh important contributor to growth,

1:40:17 but I think a more

1:40:19 explicit link I would have found more helpful.

1:40:23 I was very happy to hear Michael Kramer's thoughts yesterday

1:40:27 on how to encourage innovation that delivers

1:40:31 social benefits.

1:40:33 More government and private sector collaboration

1:40:36 through appropriate incentives

1:40:39 potentially mediated

1:40:41 through new institutional arrangements.

1:40:44 There is clearly room for more creative thinking along those lines,

1:40:50 and I trust that you all will rise to the challenge

1:40:54 because I'm gonna go off

1:40:56 into my retirement again.

1:40:58 Thank you very much for listening to me.

1:41:01 I.

showAllTimestamps

no

transcript

Welcome to your last session of the day. In fact, the last session of the conference, and it's going to be a good one. Today was meant to focus on the economic inclusion of women and youth. You will have noticed that the 1st 3 panels we had all talked about women, clearly close to my heart, but this is the panel that's gonna be talking about the economic inclusion of youth. So please stay tuned. My name is Tamar Manuel Yan Atten. I'm a senior non-resident fellow at the Brookings Institution after a long career at the World Bank, um, and we have a wonderful group of presenters here who, I, as I said, are gonna be talking about the topic of economic inclusion of youth. Now I don't really think this topic needs much of a motivation. But I'll give you one in any case, uh, I believe you all know that youth employment is a challenging issue everywhere around the world and it is particularly acute in some regions of the world, Middle East and sub-Saharan Africa in particular. According to World Bank estimates, the working age population in sub-Saharan Africa will increase by 740 million by 2050, more than doubling its current size, which is at 630 million today. If unemployment and underemployment in general and for youth are a problem now, they will be an even greater drag on growth, source of frustration, and a source of instability in the future. Rachel touched upon that point this morning. What programs may help foster the economic inclusion of youth is the topic of this panel. So let me introduce the speakers who are all gonna be talking, presenting papers that all are in the African continent, um, appropriately. Our first speaker is Oyebola Okunogbe. She's an economist in the Development Research Group of the World Bank. She was born and raised in Nigeria. Next will be Andy Zeitlin. He is an associate professor at the McCourt School of Public Policy at Georgetown University. He was born and raised in Stow, Massachusetts. I don't think I know where Stow, Massachusetts is. He's also a non-resident fellow at the Center for Global Development. Next will be Isaac Mbiti, who's an associate professor of public policy and economics at the Frank Batten School of Leadership at Public Policy. Prior to this, he was an assistant professor of economics at Southern Methodist University and a Martin Luther King visiting assistant professor at the, at MIT. He's a JPA bred MBR and ISA affiliate. He was born and raised in Kenya. And our final speaker will be Owen Ozi who's an associate professor in the department of economics at Williams College. Um, you will remember one of our previous speakers is also Williams College. It made me wonder about the in-household bargaining that took place for that outcome to happen, but we'll leave that aside. So Owen was previously a senior economist in the World Bank's development research group. He's a JPal Breed and ISA affiliate. So why don't we go ahead and start? Oyebola? Great, thank you so much for the introduction and thank you to the organizers for inviting me to this great event and thank you to you all for staying. 1 2nd to get the amount of time you have straight. I was told 12 minutes and I think so were all the presenters, but the last session got 14, so I'm. Going to ask for equity and treatment and ask for 14. Thank you. Great, yeah, I'm saying thank you to you all for staying till the glorious end. I'm looking forward to all the discussion we'll have. So today I'm going to be talking to you about a livelihood program in Nigeria, uh, one of the. Comments we got in one of the earlier sessions on cash transfers was what is it why you know instead of teaching people how to instead of giving them fish, how about you teach them how to fish and many governments apparently think in a similar way and many there's a lot of policy interest in helping people to graduate out of poverty. So many programs have a basic targeted cash transfer program to provide consumption support and then there's now um light layering on top of that a livelihood intervention to help households develop a sustainable income generating activity. Now one can think about how this should be targeted within the household so who in the household should be the recipients and then how could one promote positive household dynamics and this is where the focus on youth employment is going to become apparent as you'll see later on because what we find in our studies is that it ends up once the households get the choice of who to target young people in the household tend to be the ones that are selected. Let me walk you through this. So this is the Nigeria National Safe Social Safety Net Program. We call it NASP, and it's a very large, uh, program in Nigeria. It's reaching over 2 million people now. It started in 2017 and has grown rapidly. There's much I can say about this program, but I'm going to focus on a key element, on some key elements. And before I go further, I'm just going to go back and introduce my co-authors because one of them is in the room. OK. I was just excited to get into it. So Kindi and Jay is here, um, Robin, Tama, Ayoilli, Naira, um, are all at the World Bank. OK, let's continue. OK, so in the NAS program, the first thing to point out is that there is a basic package where, um, households get 5000 naira, uh, for each month. This is about equivalent of $12 and depends on which exchange rate you use. I know we're sticking with the one from a few years back. OK, so they also get um savings group mobilization which is where um the recipients in. Communities, they get a lot of training on saving some of the the importance of saving even some of this cash transfer so that they can channel it into productive activities and the recipient of this is the caregiver who is typically the senior woman in the house. So if it's a male headed household, which is the majority, usually the wife of the household head or um someone similar category, and there's usually an alternate who can receive the cash transfers if the caregiver is not available. Um, we worked with the government to develop a randomized, um, experiment to be able to look at other components that are layered on top of this basic package. So there was a top-up package which, in addition to these two elements I described also got a core responsibility. Support which involves training around things like early childhood development, nutrition, sanitation, and in for this the households get an additional 5000 so they get double the amount again here the recipient of all this training is a caregiver. And there's a 3rd arm which is the livelihood package. That has life skills training components, business skills training, mentoring and coaching, and the goal of all of this is to help the recipients to think through what they would do when they eventually get the last component, which is the live productive grant so it involves things like goal setting, um, budgeting, management, time management, how do you develop a business plan. How do you market your products? Basically, all kinds of skills that they will need along these three dimensions and the mentoring and coaching also involve one on one component where you actually walk people through developing a business plan and what exactly they will do once they get this grant. And the goal of this is that you didn't want people to get the grants and then not have a plan in place for all of this. So, all the work I'm going to be describing today happens within this livelihood package. We then um among the group that are randomized to receive the livelihood package, we randomize some elements of the way it's delivered. And the key questions that we're trying to answer, number one is who within the household should participate in the Livelihoods program. The um. In some programs, the default is the caregiver. So this is the person that the government has been investing in so far. They've been receiving the savings group training. If there was a lively, there was a core responsibility training. This person has been the one that the government has really been engaging with. And so it seems like the natural thing is to continue with giving it to them, and many programs do this. But we might think that the qualities that we might be interested in for a caregiver. Who's the recipient of the cash transfer may be quite different from who we want to be the recipient of the livelihoods training program so for the caregiver maybe we want to be the rationale like many programs is um you want to give it to the woman in the house thinking about the welfare of children and spending the finances on household needs household consumption. Or for the productive members where we care about the person being able to convert this um investment into profitable activities maybe we want to have people who are more um business oriented who have ideas who are creative or entrepreneurial there might be other activities qualities that we seek in them. And now we can think that the households might also be the people who have the best information about their members. They know their characteristics. They know their strengths and weaknesses and so might think that they probably have the best information on who should be the productive member. But at the same time there's the risk of capture if the household head just says, OK, then it's going to be me because I want the money to go to myself or um or is making the decision on other criteria that might not be the most um productive ones. And so we might think that it might be important to also foster positive household dynamics in the setting to ensure that the most productive members chosen. Let me now tell you specifically what we do in this in the intervention. So the first thing is we want to examine the impact of allowing households the choice of who should be the productive member who is going to receive the train all the trainings and also receive the productive grants once it's given. And the instructions that they're given is that they choose someone who is between that should be 1818 to 45 years old and able bodied. And so this is the same instruction that's given in the um. So when it's the default arm it's the caregiver if they fit this criteria if they don't then they um choose the alternate if they fit it and then if they don't then they can move to anybody else in the household that fits it and if it's in the treatment um they're free to choose anyone that meets the criteria. Second thing we did was to do a household sensitization to encourage collaborative decision making so we had a facilitator go to the household and to just help them to have a conversation on the productive potential of all the household members, so. All everyone, you know what, let's think about person A, what's the education? What's their experience, what would they be good at? What are they currently doing? who was basically trying to get a profile of each person as a way of helping the household to really be proactive and thinking about this. We also had a video to illustrate positive models of collaborative household decision making and the goal of all of this was especially interacting with the choice was to see if it will lead to uh the choice of a more productive member. So this is what the design looked like there were 2000. Households and then we had on top is the. Um, so the number one block, uh, the default, um, caregiver is selected and there's no sensitization and then we have the sensitization arm, the household choice arm with no sensitization, and then the combined. I'm Before going I just tell you a little bit more about the sample for those of you who know Nigeria, we have 6 states, so these are in each of the 6 geopolitical zones, and we're in 12 rural LGAs. They wanted to see the government was particularly interested in seeing how this type of live loop program would work in the poorest of the poor. Um, we have the average of 7.4 household members. 90% of our caregivers are women. Average of 41 years old. Less than half of them have ever attended school. Um, 90% of the households are farming, and this is important, the 63% of them already have some kind of non-farm business. I'll explain that more. Some pictures to see on some of our recipients. OK, so just pulling um the timeline. So starting in 2018 we had our baseline survey which was in a sample of people who are all already receiving the basic cash transfer. So the cash transfer line goes the entire time because the whole everyone in the sample is a cash transfer recipient. Um, Implementation is delayed by some time and then COVID hit so we don't actually implement the interventions until early 2021 where we have the household choice. um, households can make their choice and then we have the sensitization visits and then that's very quickly followed with the trainings that households are receiving. And then the cash transfer is bursed. While the cash grants are being disbursed is when we need to conduct our midline survey for World Bank people for logistical reasons, the project and you know, so we had to have this time in which leads to really interesting results on our show in a minute. So we have. About a third of our recipients have received, but it's really right as they're receiving, so all the results I'll be showing you today are really coming from people having all of this training and the anticipation of the grant and then any savings that they have or whatever activities that they've started doing, it's all coming from there as opposed to the, so it's all pre-grant training and we're going back this fall to collect more data, um, now that people have actually had the grant for a couple of years. OK, so now finally let me show you some results. So the first thing is on the selection of the um. Of the beneficiary of the productive member, so what we find is that when households have the choice, a third of the time they shift away from the caregiver when they have the option, and we find that the shift is equally to men and women. So in the first paragraph there about 80% of people. Um, in the default, the caregiver is in, in the control group, the caregiver is selected. This reduces by 25% points, and this is about equal split between a shift to women and to men. So when it's no longer the caregiver, half the time it's a woman and half the time it's a man. Um, it's less likely to be the household head's wife. That's just a reflection of the fact that many household heads wives are the caregiver, and it is more likely now to be the child of the household head, and it's also more likely to be a young person, someone under the age of 40. And moving now just to summarize what happens with the. Very early results that we're seeing again this is all pre-grants but we're seeing an increase in household employment um and we think this is coming, um, primarily from the fact that the caregiver is a lot of them just from the cash transfers the savings groups they're already mobilizing into already working in a, um, productive, um, household enterprise so they're doing little petty trading they're doing something and so there are other members who are not. Economically active who are not engaged and so we find higher employment now for these other members we see an increase in household earnings as well and we see some um increases not statistically significant but once we um uh combine all of these different sources of income we see increases in the overall household earnings. On the household sensitization we don't see any effects on the choice of the productive member. We see some improvements in women's mental health and their participation in paid work and in children's nutrition, um, but we don't see any effects on some of these outcomes I just showed about economic activity and consumption. And interestingly we also don't see anything in this intersection of the two, so our ideas that helping them to discuss would help apparently they can make those decisions on their own. OK, so just um wrapping up, these are all very early results like I've pointed out we're um going to get more data but what we see so far is that it seems that private information that households have among themselves is very important and it's and we need to think about how we can leverage them in design of this program. Yesterday and even part today there was a lot of talk about how governments are in a very tight fiscal space and you know these investments in social protection programs uh uh fiscally demanding and here we see really promising evidence of how a simple tweak in the within household program targeting can lead to significant improvements in outcomes. And here we see specifically that it's moving towards the younger members of the household which then deals with this big big problem that we've discussed many times here about the youth unemployment so um we're cautiously excited and look forward to um telling you more about what we find in a few months. Thank you. Ask the baton. Uh, OK, great, um, so, uh, it's gonna take 10 seconds. That's fine. Uh, so I'm standing in for Thomas Gin, uh, who's, uh, based at CGD, and my fantastic co-author, uh, and this is joint work of ours, uh, together with Travis Bassler at Rochester, uh, Belinda Muya at the, uh, IRC, uh, and Ibrahim Kassiri at EPRC in Uganda, um, and so we're gonna be talking about work that is done, waiting. We're gonna talk about some work that we've done, uh, focusing on, um, uh, programs to sort of, uh, provide both cash assistance and mentorship, uh, to youth, um, in the, in urban. Did you click. There's nothing. It's not me. There's nothing, uh, in, in Kampala in Uganda. There we go, come back, um, OK, so we'll be talking about work, uh, that is, uh, uh, taking place in Kampala in Uganda, um, and I think with a special eye, I think we're, we're still allowed to talk about women in this session. So with an eye towards, uh, sort of gender cleavages and then in particular, uh, cleavages acro uh. Sort of another dimension of potential economic exclusion which is between host country nationals and refugees in this urban context as we think about what are the particular barriers that might be there for these particular populations that might impede economic integration for them. So to think about where we're coming from this and and it's nice to kind of follow Oyebola's uh presentation because I think there's some features that I think you'll you'll hear across several presentations about the types of livelihood interventions that are that are there. So here I think you know the starting point is that this is a population that is engaged in informal micro entrepreneurship to a large extent, uh, but in a way that has relatively low profits, uh, for most, um, and we hypothesize that some of the constraints that they're up against are both financial constraints. So there are liquidity constraints stopping people from expanding their businesses, um, but also maybe managerial capital that would impede them or or the presence of which would enable them to sort of better put those resources to use. um we've seen lots of cash interventions with some positive effects, at least in the short run, sometimes these are lower for women and so we want to think about kind of what are the potential design features that you could build around that to make sure that those returns were there for all and. When we think about managerial capital training interventions that might sort of provide formal business training um have varied and I think sort of modest effects in general relative to the relative high costs and so thinking about on uh mentorship models that might provide an alternative to that that sort of provide the decentralized delivery of information at lower cost to the uh to the NGOs that are trying to support these things or the governments that are trying to support these things may be a viable and interesting alternative. Uh, which raises, I think, potentially interesting questions about how these things get tailored to the particular recipient, how it matters who is mentoring whom for the types of constraints that they might unlock, um, so that's part of what we're going to explore in this study. Um, and so we're gonna think about kind of how mentorship might improve business practices, uh, the networks that people have access to, um, and also knowledge, knowledge exchange in ways that might be important, uh, for these populations in particular. OK, so we're working with the IRC to study uh an, an aspect of their rebuild programming. Uh, we do so, um, with a sample of Ugandans, uh, who we'll, we'll describe as the host country nationals and then refugees and also purposefully sampling across gender cleavages, um, to, to look at both impacts on men and impacts on women. Um, and in the large part we're gonna be focusing on a, uh, cohort of 2000 relatively inexperienced perspective entrepreneurs who fall into this kind of broadly defined youth category, um, and then, uh, they are in some instances gonna be paired with mentors who come from a slightly more experienced cohort and tend to be a little bit older. At the core of the interventions that everybody's going to get are a set of cash grants. These are $450 equivalent, and the control group is getting these sort of at the end of the study after, after all the measurement is wrapped up. And then this is in some instances paired with this mentorship intervention that puts people not in one on one mentorship but into groups of three mentees together with a single mentor. They met weekly for a period of six months, or that was the program to which they were assigned. Um, this is about 2/3, almost 2/3 of our sample, um, and consists of a couple of pieces. Uh, at the core of it is some actual kind of structured content to give a conversation starter and maybe to sort of, uh, provide a framework where, where, um, uh, where, uh, for the kind of exploration of some formal skills to the extent that those kind of managerial skills are important and standardizable in some way. But we're also going to think about how the particular pairings might matter for the types of information that people can give each other, so we experimentally vary whether the groups are homogeneous or heterogeneous with respect to gender and nationality, and we'll sort of think about that. And then following the literature that suggests that some of this idiosyncratic advice that can be particularly helpful but that fades away as pairings fall apart, we also have a kind of layer in this study that thinks about whether. Whether part of the constraint to the success of these things might not be might be in the in the incentives within the mentorship models and whether some kind of shared fate might be important to getting to perform better, so there's some group level incentives in some of these groups and what I'll share with you here are preliminary results from a baseline and then quarterly surveys over a year of exposure. We've got one last study round coming up behind that. In broad terms, what we're going to find is positive and persistent both over time and over groups, I think to a surprising extent effects of cash, and I'll show you how that holds up. And then on average, essentially a null effect of the mentorship layer put on top of that, but a null effect that masks some, I think, quite interesting. Heterogeneity that's central to the design here and in particular we'll see positive effects for some groups, for men in particular, and especially when they are matched with men, and some negative effects for women, particularly when they're matched with women, and that challenge, I think, will be in terms of the design is a little surprising to all of us and something we want to unpack in this and further work. OK, so, uh, to be brief, uh, with background, so Uganda's a, a, a progressive, uh, uh, refugee hosting country with lots of people there and with, uh, pretty broad rights to work and to move. Um, this is a population in this urban setting that they are not, um, because they are registered refugees, they are not refugees who have just arrived. the previous month or anything like that, many of them have businesses. 70% have businesses at baseline. That's a little higher among the Ugandans and a little higher among the women in our sample. Their profits are low from these businesses, are modest from these businesses at $28 a month as we measure it, but similar. Between the Ugandan and refugee populations, bearing in mind that this is a host population that's selected to be sort of comparable in in sort of the economic barriers that they face, uh, a little higher for men and significantly higher for the mentors who are brought in precisely because of having established businesses along those lines but not sort of a whole world apart you might say. Um And one of the key questions here will be what might be the importance of the business context that people can be introduced to, so, so it's notable then that Ugandan men have the most business contacts at baseline, um, but not a lot of cross-nationality or cross gender contact, which suggests that that might be, there may be frictions there if there if there's useful information exchange to be had that might have some positive effects. So let me show you then some of our our core results. So as I said, I'm I'm showing you for brevity results that are just presenting the kind of pooled effects across all of the subject groups for the time being and across all of the rounds, but these are these are temporarily persistent and persistent across the host and refugee populations to to a surprising extent. So we see quite substantial effects on profits. These are moving from about $40 equivalent in the post-intervention period to close to $80. We see a near doubling of the capital stock, so a lot of the cash that people are getting is showing up in increases in the capital stock. We see against even that relatively high base some increase in the prevalence of business ownership. And then these things filter through into better mental health, so people report being sad most of the time to to a lesser extent. We see that these don't seem to be these sort of related to Oyebola's question in a way these targeted interventions don't seem to be crowding out income elsewhere in the household, so it's not the case that the household earning effects are substantially less than the beneficiary business effects, and we see this sort of passing through into. Measures of improved consumption stability and the like as the skipping of meals decreases substantially in this population. So I think the first thing as we think about the sort of fundamental constraints in this population is that liquidity constraint is substantial. And while there's a broad literature here, I think we contribute something to the sort of sense of stability and persistence over time and across populations. Turning to the mentorship dimension of this, um, we see that these effects are really quite modest. We look at this across subgroups and thinking about who people whom people are paired with when we want to think about what the consequences of mentorship might be. Um, so showing you first in the panel on the left, uh, the benefits of mentorship for Ugandans and refugees again across both genders in our sample here. So relative to cash, you see that Ugandans are in point estimate terms slightly worse off in the mentorship program than without it, and the refugees are sort of, you know, more or less the same to a large extent with some hint of something positive when they're mentored by a refugee rather than by Ugandan. The gender dimension is where this becomes particularly interesting. So as you can see, so at the 10% level, men being mentored by men have statistically significant profit effects in this sample over the study period. And these are sharp, especially in the early early rounds, and then mentorship seems to, on the other hand, have negative effects for women, particularly when women are mentored by other women, contrary to something of our expectation here and the expectation of our implementing partners. Um, so thinking about how the decisions that are made by people are shaped by the experience of mentorship and, and how the access to information that they get may sort of be traded off against the ways that they, they, they make slightly different decisions or or or or decisions that exposes seem not to have worked out I think is an interesting and key question for us. We don't see. These effects driven by changes in the sectors to which people enter into, we don't see them driven by the likelihood of opening a business, so, so some of the obvious channels that you might have expected are not there, um but uh but I think this is potentially potentially important where we do have some sense of what might be going on is when we look at how this works across the distribution of business profitability. Um, and so when we, so what I'm showing you here are quantile treatment effects looking across the range of, of, uh, business profitability in the, in the after intervention period, um, and first showing you in the upper panels for men and for women what the impacts of the cash grants are by quantile. So unsurprisingly, you know, we see these big effects at the top of the distribution um that are that are potentially important here. Strikingly then, when we look at the effects of mentorship, and so the bottom graphs are going to show you the effect of mentorship relative to cash alone across the quantile of the of the profit distribution. We see that the negative female effect seems to be coming particularly at the top, so it seems like some high return things that might have happened in the absence of mentorship don't get invested in because of the conversations that are happening in these groups, particularly when women are being mentored by other women in this context. But exactly how that's being driven and what trade-offs, what other benefits might be there that are being traded off against, I think are questions we want to explore a little bit more. OK, so then just to, to wrap up, um. We see in this work substantial effects of the cash grants on business and household outcomes. These effects are persistent over time and across populations, and, and I think one contribution here is to sort of line up in time and at the same time. Time and space, uh, a trial studying both the refugee and host population impacts and to show that we have some evidence of the generalizability of the sort of interventions to alleviate liquidity constraints across that that barrier or across that division. We have these marginally positive effects of mentorship among men, particularly when they're being mentored by other men, and these marginally negative effects of mentorship among women, particularly when they're being mentored by other women and seemingly concentrated among the upper quartile of the profit distribution, whereas we don't see this being driven by substitution of other economic activities, we don't see this being driven by differences in the sectors that they choose, and we don't see this being driven by, for instance. The taxation by their mentors or something like that, that there might be some bargaining going on within the mentorship group that that results in funds not going to them. So some evidence on those constraints and some mysteries about where why why this mentorship model has fallen down in some instances for some people. Perfect, thank you. So thank you so much. Uh, it's really great to be here. Um, and thank you for being here. I know it's where I'm keeping you from, or we're keeping you from dinner, so, uh, we'll try and be, uh, engaging and interesting, and, uh, but yeah, we'll try. So, uh, I'm gonna give a um Uh, a presentation on, uh, some work I've been doing in Ghana with co-authors, uh, Gabriel, Morgan, Jamie, and Isabel, and we've had some outstanding, uh, RA support from, uh, DT and Michelle as well. So I wanna just give them a shout out to people who are behind the scenes who do great work in helping us move things along, um, so we've been working in Ghana for some time and. Um, you know, I think, and I, I, I'll probably skip a lot on, on, on this motivation here and this sort of framing, but essentially, you know, we're all thinking about youth unemployment and underemployment, and, you know, uh, it's not good, right, though I think we can, uh, all agree on that, and we're all trying to find ways on, uh, to, to address this problem, right? And so what we've been doing in Ghana for some time is looking at apprenticeships and, uh, in many, you know, uh, low and middle income countries we have apprenticeship systems they're pretty common. Um, I think often when people think of apprenticeships they think of sort of the sort of German model in some sense of, you know, well established big firms, etc. but in these contexts that we're looking at we're often thinking about or, you know, speaking about sort of, um, working in very small scale informal sectors, right, where the firm owner, you know, has maybe like 2-3 employees and a couple apprentices, and they're the one that's providing the, the training, and that's kind of the model of training that we're looking at and studying in, in, in we've been studying in our work here. And so why do people sort of get interested about apprenticeships in general? I think they're seen as this promising way of sort of providing skills, um, but also and with the idea that you know I can boost the employment prospect uh at the end of the day, um, and we, it, it's seen as maybe practical right because again you're learning sort of on the job you're learning from someone who actually is active in the trade, right? and in some cases this is paired with on you know classroom instruction sometimes not um. And you know there's sort of some thought that you know what ends up happening really is that at least in the sort of the systems that we're studying is that the apprentice basically learns from the trainer and then basically replicates exactly what the trainer does and sort of copies their business at the end of the day and goes sets up their own firm right so. Again, unlike sort of what you see in developing countries, there's no real retention of like the firm hiring the apprentice at the end of the day where this is sort of a screening mechanism for for your workers, but this is really just a uh a training um system right that then sort of trains other folks that enter um and start new businesses, right? so. Uh, you know, this arguably could be more relevant for sort of the context we're talking about where this is a large informal sector, right, um, you know, in Ghana, you know, 88% of men, 95% of women are in this sort of informal sector that's, you know, arguably low productivity, right, and so. You can imagine that this is, you know, you're learning from someone who's actively in the space who's been experiencing in this space and so this might actually be a good um um form of training, a relevant form of training for that relative to let's say a vocational training uh institution which is very classroom, uh, based right and so um. You know, and you know, it's, it's somewhat surprising, or at least you know one surprise people often get surprised, but just how common these, these sort of types of training are and, you know, in Ghana for example, like, uh, this type of training in formal sector apprenticeships has trained almost 4 times as many individuals as all other sort of formal training of, um, alternatives, right? So if you look at sort of people who went to sort of vocational schools and all these other forms of schooling, you know, apprenticeships basically dominates it yet. The actual academic literature looking at apprenticeships like the economics literature is actually pretty thin relative to the importance um. Now, um, well, again it could be relevant in terms of, you know, actually training you for, uh, the, the, the sectors, um, I think a lot of, uh, observers have raised quality concerns, right? And so things that they've raised are some factors have sort of pointed to specifically are, you know, many of these firms are using outdated technology. There's a lack of monitoring that's going on there's a lack of formal contracting the standards in the curriculum are often sort of not there, right? They're just sort of everyone's kind of doing their own thing in some sense, um, there's no sort of formal certification or it's limited in many contexts, um. And there's also the strategic concern right where basically because again I'm not unlike you know sort of like a German system where the firm sort of hires a person at the end of the day here you're basically after the training they go off and start their own business they replicate your own business, right? And so there's this concern that you might be like well I'm actually I don't wanna train Andy Owen Oyobola because they're gonna start competing with me so I better maybe slow the pace a little bit, maybe withhold some things so there are these strategic concerns that come into play as well. So How could we address this concern, and that's what we're going to look at, um, or that's what I'm gonna talk to you about today is basically one potential way to address, uh, some of these concerns, these quality concerns, and one way to do that is by introducing financial incentives very much in the spirit of what Andy was talking about or just sort of getting everyone's skin into the game, right, and sort of a shared, um. Uh, skin in the game, so to speak, right, so the idea here is that we will, again, economists love incentives, right? That's any, that's always a solution to everything that economists sort of come up with, um, but you know, in this case we thought, you know, um, the trainer remuneration or, you know what they actually get right in terms of cash, um, and what they apprentice, you know, if we can strengthen that link, right, that could potentially increase, um. Um, you know, the sort of skills, uh, uh, attainment right amongst, among apprentices, right, and so we thought, well, this could encourage training effort and improve skill acquisition and maybe, uh, sorry, some of these things didn't render very quick uh correctly, um, better skills could lead to better. The labor market outcomes right now again this was I think motivated a lot from I think, uh, the academic literature in, in, in schoolings right where you see in schoolings where you see basically financial incentives for school teachers in K-12 education can be quite effective. There's a lot of literature on this, um, that's, that's sort of done shown that. Now Well that's true in the sort of schooling literature, it's unclear if sort of financial incentives will be effective in the apprenticeship system, right? So again, when I come to a trainer, right, there's a large opportunity cost of training, right? So like if I'm training someone, right, if I spend a little bit more time training or Ebola Andy and Owen, right, I'm potentially not spending time making stuff, dealing with customers, right? So, so there's a large opportunity cost of training, right? And there's also these strategic considerations like, you know, if I do put more effort they're gonna be better at competing against me, right? So, so it's not clear this will automatically just translate from the education sector or, you know, the formal K-12 education sector into this apprenticeship system, um. Now, on the other hand, um, what you do see if you just to sort of survey sort of a lot of training programs that do exist is you see many of them sort of feature these sort of, um, outcome-based contracts for trainers where basically the trainer payment is linked to training outcomes. So you've seen this in, you know, some work that the World Bank did in Liberia by I think lots of folks, um. And other places, right, so that's become a common feature of, of, of many training programs but again that's we actually don't have limited evidence of the effectiveness of that sort of contractual structure so. So, so in this experiment what we're gonna do is, uh, now it's embedded within a larger experiment, uh, that we're doing, uh, that's, you know, still sort of working on it, on that's trying to look at the returns to apprenticeships in, in Ghana and which was that RCT was in collaboration with the, the Council for Technical Educational Education and Training in Ghana, uh, it's a government agency. And so what we're doing here, what I'm going to talk about today here is basically a sub experiment within that larger experiment where we're basically looking at, you know. What is the effect of giving incentives to training providers within this, uh, larger RCT? OK. The apprenticeship training that we're, uh, looking at is gonna focus on construction trades, uh, cosmetology and hairdressing and garments and tailoring, right? So those are the sort of the trades that were offered. We didn't pick those as, as researchers. This is what the, the, the government sort of had picked. Um, and, uh, due to logistical considerations and just constraints that we had, um, basically there were no syllabi, right, that we could use, um, we could only sort of implement this, uh, for cosmetology and hairdressing and tailoring and garments. So this is just a broader um research design uh sketch just to sort of give you an idea of what's going on. So in 2012, so yeah we, this has been going on for a while, um, we sort of did the baseline for the larger RCT, you know, recruited us applicants, um. You know, we sort of put people into different buckets of control treatment. Uh, priority you'll see there is basically we allowed, uh, district officials to pick, um, oh, only 5 minutes. Alright, uh, pick, uh, people, so I'll keep going. I'll talk a little bit faster and then we moved to match apprentices with trainers, uh, that was sort of 2013, and then we, uh, layered on the incentives program just among those that were in training, OK, and so we have about, uh, you know, 225 firms in each, in each, um, in each treatment arm. Um, what are we doing basically in the treatment group, um, trainers are gonna give a, uh, get a financial bonus based on a skills test that we're going to administer to their apprentice, right? So basically if they, we, we come in, we're gonna test the apprentice, um, see how, you know, how. Well they've learned hairdressing or tailoring and then basically the better the apprentice does within the district, right, so if they're the top ranked, uh, tailor of this test within the district, the apprentice gets a bigger bonus. So it's a rank order tournament structure, um, within the district and in the control group they're going, the trainers are going to get a fixed payment of 100 Ghana CDs this, uh, if the if the apprentice takes the skills test, right? So, um, basically there's no incentive there to, you know, increase your effort. Uh, and the average payment is equalized across, across both groups. Um, here's a bonus schedule. I'll skip it, but, um, and I'll skip all of this. So we have lots of data, uh, basically we have our baseline, we have our skills data, we have, uh, tracking data, so we're going through, um, Um, the intervention we have, and we have two follow up data, uh, we have one, sort of what we'll call a midline, so to speak, in 2017-2018, which is like two years after, um, the intervention of the incentives, and then we have a 2022 follow up, uh, as well. So I'm actually gonna present both and the 2022 follow up with 2022, 2023, that's fresh off the this data, so it's very exciting. Um, so who are these folks, um, you know, apprentices, they're young people, they're about, they're about 23 at, at the time of enrollment into the program, about 7 years of schooling, um, uh, because again, we're focusing primarily on cosmetology and tailoring, they're basically all women, OK, um. So skills assessment really hard to do um we basically had to hire people who were external experts to go around and conduct these assessments uh so folks had to take a theoretical portion like there was a test that was developed uh on the theory of hairdressing, the theory of garment making, and then we actually had to do a practical test. Basically, and you can see there on the pictures, like, you know, how well can you braid hair, how well can you, you know, style the hair, and then, um, for the garments, you know, we had they had to like, you know, sew zigzag scallops, straight edge, and then they also had to like you know, actually make a dress, uh, and we graded all that, uh, about 2/3 of the folks that, um, uh, were assigned treatment actually took participated in the assessment, and that's balanced across treatment and control. And what do we find um in terms of skills in the treatment group practical uh practical assessment, right, so how well can they sew, how well can they do hair, uh, we had about a 0.13 standard deviation increase uh in um in, in, in learning outcomes, right? So they actually did a lot better in, in skills. Assessment, um, in a theory it wasn't really much that was stati statistically significant, and they, and then they could get a certificate if they passed a, um, a threshold, and again there was no real difference between the treatment and control in terms of passing this, this test, um. We come back sort of in 2017 we give him a quiz, um, that basically says, you know, if you were to, you know, and that was sort of very practically, um, very practical questions about like, you know, how would you do this hair, how would you sort of, uh, um, you know, do this stitch, etc. and again still this is again. Two years after the the incentives they were better on they did better on that quiz and they also did better on sales on the sales skills index, right, basically we asked them how would you actually you know sell this stuff right? so basically skills went up as a result of the treatment and I think we're pretty convinced about that. In terms of labor market outcomes, um, we don't see much in terms of a change in labor supply like we know which sector they worked in, whether they worked for self-employment or wage work, um, but we see increases in earnings, right, so we're about, uh, you know, 25% essentially increase in earnings in total earnings over the past 12 months we're seeing sort of, uh, and that's driven in, in, you know, in 2017 by self-employment earnings, right? So they're basically much more productive in self-employment, um, as a result of that, right? And then. Now, why is that? Well, obviously skills went up. They also worked more hours during the apprenticeship, right? So they got more training, so to speak, um. Their trainers used more syllabi, right, so there was a standardization that sort of came into play and there was an increase in the pace of training, right, so basically there were actually we see that apprentices were more likely to complete their training and um quicker. What was surprising to us was these results were persistent through 2022. Right, so. Couple of years ago we come back we measure their labor supply and earnings we see that uh in total earnings is still up by, you know, 25, 28%. Wage earnings is what's driving this now, which is sort of surprising, um, so it's actually being driven by wage earnings and there's no real increase in self-employment earnings, um. And uh again it's it's all just coming through wage and and now we're seeing in terms of labor supply they were more likely to work for a wage. So we were asked to do some policy sort of thoughts too which I'll, I'll I'll, I'll do my minus 1 minute I know, um, but essentially what I'll say is well designed incentives could improve apprentice learning and uh in a cost effective manner, right? So it's hard to do. You have to really pay attention to the design details. Uh, and in our setting the program paid for itself in, you know, about 15 months, right, in terms of apprentice earnings, right, so it's actually pretty cost effective to do this. Unclear if this will scale up, right, because again, as you scale this up, people can learn how to game and and sort of, um, game the system. But on the other hand, as I think, uh, Owen and co-authors work and Andy's work actually, um, has shown, the incentive structure could attract a different or better trainer to the system. OK, so we just need to do more research on the scale of potential, uh, and I'll stop there. Thank you. OK, you've stuck around to the end. You've heard about, uh, some long term follow up. You've heard about a cash grant program. You've heard about a program that built livelihood skills, uh, teaching people to fish and not merely giving them a fish, and in my presentation, I hope to talk about all of those things. Uh, this is a firm of one's own. It's joint work with Andrew Bruteval Newman and Madalena Honorati at the World Bank. Gerald E Papa, who's just finished his PhD at University of Delaware, and Pamela Jaquila, who along with me, we're at, uh, Williams College, so. Uh, I want to show a picture, a couple of pictures that illustrate things that Tamar said right at the beginning. This is a picture taken from a report that, uh, Dion Filmer in the room and Louise Fox wrote. What you can see is that, uh, on the left side, sub-Saharan Africa's population pyramid has a tremendous amount of youth, OK, compared to almost anyplace else in the world. And so when we talk about employment, we need to think about what are young people going to do. That's really the focus of policy and you pointed that out earlier and this is a picture that illustrates it. Uh, when we talk about sub-Saharan Africa, this is a, a graphic taken from Bandiera et al. Uh, look at the bottom thing with the pink bar and the green bars, OK? The green bars are self-employment, the pink bars are salaried employment, wage employment. The top one is Africa, Sub-Saharan Africa, 28 countries I think, and the bottom bar is other countries in the world, and the stylized fact here is that self-employment is really big in sub-Saharan Africa. Labor demand in wage employment is quite low, and there are factors that you can think about changing that. We've heard talks today and yesterday about different ways that, uh, industrialization and a revolution in services may change labor demand, uh, but this is the situation as it is now. This is another picture of the same idea. The pie chart on the right side shows that there's this little quarter in the upper left that is wage sector and everything else is not in sub-Saharan Africa. Also from Filmer and Fox. So the project I'm gonna talk about today happens in, uh, neighborhoods on the sort of east side of Nairobi, Babadogo, Dandora, and Lunga Lunga. Uh, people who live and work in these neighborhoods live and work with varying degrees of formality and informality, uh, and this is going to be an intervention a lot like the ones that you've heard about. The intervention uh is run by the International Rescue Committee which was working on the project in uh Uganda as well. It's a micro franchising program, so it's going to bring together a bunch of the elements of things that you've heard about. The target population and the population in this study is young women aged 18 to 19. They're living in these neighborhoods and they express an interest in starting a business. The idea, the distinction between micro franchising and other multifaceted interventions is that in addition to providing a kind of life skills training, which you heard about in Oyebola's presentation, uh, and a specific business skills training, in addition to providing capital so that you can start your business to overcome any capital constraint, credit constraint you may face, in addition to all of those things, micro franchising is going to attach. You and your small business to a nationally recognized brand. You're gonna have a business model to do, OK? Now it's self-employment, so you can decide whether you want to do it or not, but there's gonna be a business model all worked out and recognized branding. So that's like we're trying to overcome the IRC is trying to overcome with this program as many of the possible constraints that may inhibit young people from launching their businesses as possible along the lines that we've heard in other presentations today. So you're matched with one of a few different franchise models on the basis of the preferences that you express, and you get all this training, and you even get mentoring, that was a feature of several of these things as well. You get mentoring after your micro franchise is launched. Uh, the two sectors that this ends up being in, the partners that the International Rescue Committee had to work with young women in the east side of Nairobi, uh, were hair salons. So here you can see an academy where people are learning how to do different things with hair, uh, and here's one of the salons that they've started after coming out of the academy. And uh mobile food carts where they sell-prepared, ready to eat foods, uh, and there's like a, a company that they're associated with and a national brand in each case. Another thing I mentioned you heard about was a cash grant. So we're going to run a sort of a horse race of sorts between two interventions that are both meant to help people in this population. So one of them is the micro franchising program. The other is a completely unrestricted cash grant. It's about $230 20,000 Kenya shillings, and there was no particular encouragement to do anything with this money. So you could think a lot of things might happen as a consequence of getting a few 100 extra dollars. Now, what does this do for kind of thinking like economists? Uh, this will relax the credit constraint, so if you're just short a few $100 and you don't have a way to get credit, this will overcome that problem for you. But if you face all kinds of other constraints, this won't necessarily do anything about them. So we think of these as competing models, um, and so if, for example, a variety of skills that you could get in the training program that accompanies the micro franchising intervention, if those are the skills you're really lacking, then it will do much better. So which of these is going to help people get out of poverty better? Um, one thing to notice and one often described advantage of a cash grant program is that it's much easier to implement. It's very simple. You don't have to find a partner organization of the kind who's willing to do training and launch people with a brand. It's much simpler than that. Uh, and if I'm running a training session next Monday starting at 9 o'clock in the morning. It's going to be a logistical constraint to get people to be there. Some people have something already next Monday at 9 o'clock in the morning, but if I just need to get you a couple of $100 we can meet at almost any time to do that. So takeup is going to be very high in the cash grant program relative to the micro franchising program, and that's common. The fact that it's not 100% take up, more like 2/3, 60%, that's common in the literature on active labor market programs. People who are looking to change the way they interact with the labor force. Have other stuff going on and so they aren't all going to show up at a training at a specified time, even if they express an interest in doing so. So for that reason, what I'm going to show you are going to be what we call intention to treat results. So I'm going to say here's who's assigned to this group, here's who's assigned to that group. It's a randomized trial with 3 arms. So People in this randomized trial express interest in the program. The program has limited spots, so in the middle, in the blue, and blue will be a color associated with micro franchising throughout the rest of the presentation. Several 100 people are assigned to the micro franchise group. In the green, you have people who are assigned to the grant group, and in the gray you have the control group. As I said, uh, the takeup is going to be lower in the more complicated program, so we see that we get everybody who completes the baseline survey. 95% of those assigned to receive a cash grant receive one, and of those assigned to and invited to come to the training program to do the micro franchising, a little over 60% start the training and go through some part of it, uh, and about 40%. Uh, get to the point of launching a micro franchise, but if the training itself is all you need, you don't necessarily have to get to the launch point. So we're going to show you intention to treat in both cases. Uh, the cash grant is similar in magnitude to the value of the of the micro franchising intervention, but it's very hard to price some of the elements of that intervention, so I'm not going to say that they're exactly the same. I'm sure they're not. OK, first thing, we followed this group of young women for 6 years after the intervention. And on the dimension of self-employment, the results are almost exactly the same at each of year 1, year 2, and year 6, and they're also almost exactly the same between the cash grant arm and the micro franchising arm. Whether you are invited to the micro franchising program or you receive a cash grant, you're a little more than 10% points more likely to have self-employment in your portfolio, starting that moment and lasting for years to come. OK, so it's a lasting, it's a persistent effect, 6 years. Now, how does this work out in terms of your portfolio of income generating activities? I only talked about self-employment. So in the top bar we have uh the number of income generating activities that you do. And in the bottom set of results we have uh the impact on paid work, so working for somebody else as opposed to the self-employment. In that first year, let's look at the bottom left corner first. In that first year, whether it's the franchise program or the grant, you get this thing that helps you do self-employment and you're less likely to do wage employment because you've got this other thing going on, right? Uh, that doesn't last, OK? It doesn't in a persistent way, it does not crowd out wage employment. Uh, and you can see that pattern, uh, sort of in the top panel, in the top middle bar, you can see that by year two, particularly in the franchise group, people have a larger number of income generating activities in their income generating activity portfolio. OK, so I sell shoes, but I also do hairdressing, whereas in the comparison group, I only sell shoes or something like that. Um, We can talk about whether you're exclusively self-employed. That's the top row here, and that's something that we see an impact on in the first couple of years, but that does start to diminish. OK, so you start out getting excited about this new self-employment thing and that that drops off as the only thing that you've got going on, as we can see in the bottom right. It looks like self-employed and working for others is something that gradually becomes more likely if you're in the franchise group, for example. So this is adding to the set of things that you can do, and at first you do mainly that and then it's just one of the things that you can do. When you get a cash grant, are you gonna stop working? You're gonna work harder. In our population, that green bar on the left, people who get that cash grant are working hard to make their new businesses exist. Uh, in fact, they're working more hours than, uh, the people who get the micro franchising intervention. Maybe because there's more work to do. They have to figure out their business model, figure out where to source all the capital, things like that. Um, but those effects are not persistent. Something that I'm sure everyone wants to know about is the effect on income. I'm showing you cumulative distribution functions, and the simplest thing I can say about the picture on the left is that. The gray bar being above the blue and green bars at the very beginning, means that people in the control group are more likely to earn almost nothing. Than people in either the cash grant or the micro franchising group, but that sharp distinction that is there in the first panel starts to fade away as we get to year 6. So in a regression table framework that some of you may like, uh, and others of you may not, um, we see that uh you're earning, this is in Kenya shillings, so you divide by 100 to get to dollars, uh, people in the franchise and grant groups are earning a few dollars more per week in the first year relative to about $5 a week that the control group is earning in that year. OK, it's statistically significant. People are earning more in the first year. That is no longer true at years 2 and 6. The income effect goes away. And we might think maybe that's just a thing about averages. Sometimes averages mask variation in the distribution. We saw some quantile treatment effects. So we looked around the distribution here. This is a difference in CDFs. And again, the simplest thing I can say is any place the shaded interval is below the line, it means that's a place where it has made incomes better for the treatment group. So the franchise treatment really seems to make things better for people with low incomes. They end up with higher incomes in that first year. The grant treatment seems to have impacts all across the distribution, including at the upper right, the right end of the distribution. So there's some pretty successful businesses, but when we look at years 2 and 6. There's really no place in the distribution. You can kind of squint and see one or two little blips, but there's not really much variation that the average is masking. There's not a huge persistent effect on income somewhere that the average doesn't let you see. So this is a program. These are both interventions that help incomes in the short term, but that income effect doesn't last. Um, one thing that people say about, uh, as an example of women starting businesses rather than men starting businesses, sometimes the husband is the problem and is a constraint on women's business operation, but most of these women are not married when they start these businesses. It's not really the husband's fault. That's not the mechanism here, OK? So women launch businesses, do they get rid of them entirely? No, um, and then we ask. Uh, so if you don't get rid of them entirely and if you are, you know, not earning more income, uh, what's happening here? And so we were brought to ask questions about well-being, some of which we saw in one of the earlier presentations. We ask about living conditions, we ask about food security, we ask about subjective well-being and happiness now and what you expect for the future. Uh, and what I've got in a table format here is a pretty simple idea. At year 2 we asked this and at year 6 we asked this. At year 2 we didn't really see any significant effects, uh, to speak of on well-being, but by year 6, the franchise intervention that had the mentorship and the life skills training and all of that stuff. That left people feeling better about their lives, in some cases in concrete ways and in other cases in less concrete ways. So one lesson we took from this is that it's valuable to look not only at income and sector of employment but at how people feel about how they're living in their lives. It seems like the micro franchising intervention helped people see themselves as micro entrepreneurs and know that that's a difficult path to go on and understand that the struggle that they're on is is one that is part of the business project that they're doing and it doesn't mean there's something wrong with them. Um, so people see themselves as micro entrepreneurs more in the franchise group and not in the cash grant group. So, uh, there are a variety of mechanisms you could explore more, but for those of you doing projects along these lines, we encourage you to not only look at long term effects but to look beyond income and sector of employment. So I'll leave it there. Thanks very much. Great thank you very much for excellent presentations and uh good discipline on timing so that leaves us a good um 20 minutes for discussion we heard some promising interventions we heard some. Not terribly promising interventions, at least as far as labor market outcomes are concerned. And some concern about scalability. So let's open it up to questions, comments from the floor. Show of hands, I have one there. And then I'll take, um, we'll do a second round, 234 right behind each other there. Thank you, uh, great, another great session, uh, thanks for, for these insights. Uh, question on both the last and the first paper on, on Owen, uh, do the, how much of the franchises still exist, uh, at year 6, since that's part of your interpretation there and then. Any, since this is about income diversification, any way that you can look at whether this helps cope with shocks or manage shocks, and that that's where some of the benefits are coming from, not on average, but on in this moving part. Um, and then, but on, on, on the first one, super interesting, um, it's really more a question of interpretation. So the, the, do you know why they kind of allocated to the, to the youth and not to the husbands, uh, and then, uh. Uh, uh, you know, and, and, but, so, so, so, so you said so optimistic. So, it sounds like it's, you know, maybe we should wait kind of in light of, of what we heard, but it is like, uh, but kind of you are optimistic about it, so maybe tell us a little bit why you think that was a, you know, why that's a positive outcome. Thank you. Let's go over there. Thank you all for these really interesting presentations. I have questions for Owen and Isaac on sector and um I guess speculation about generalizability, so. Um, I know you don't have data on what this, what would have happened if you had done this in different sectors, but I'm wondering if you know based on how these, the hairdressing sector, the tailoring sector compares to other sectors that, um, young people and women in particular are involved in, whether you think, how much of this do you think is sector specific and how much do you think might generalize to other sectors beyond those? Hi there thank you so much for your presentation. I'm Rachel, a recent master's graduate at Georgetown University and the School of Foreign Service in their global human development program so I was curious about, so for micro entrepreneurship, um, from your research and things maybe even working on what's the idea in terms of scaling up so we are talking about providing jobs for the unemployed and the youth, but how do we move these micro entrepreneurs to become SMEs or even larger. And where do investors come in? How do we attract investment, you know, are there any angel investor net investment networks you've discovered in Africa? So just thinking about that scaling up so that we can try to solve this problem of, um, youth employment. Hi, thank you so much, uh, Farhan Majeed from USA um and you know. Pennsylvania. Um, I found all of the presentations really fascinating, but I just want to comment on the first and the last right now. So for the first presentation, I was thinking if you've thought about the abilities of the household members, and, you know, like you, you can think of this as a response, a household response to abilities and thinking about what are the efficiency versus equity motives here. So you're just thinking from a youth versus non- youth, but I'm just curious about how is that playing out. And if there's a way for you to measure preferences and um Work with that. Um, and for the last presentation, I was just thinking, given the age group of 18 and 19, and the time period, if there's um Um, delays in like childbirth or like marriage patterns as well, which you noticed. Thank you. Thank you. Why don't we start with you. So the first question on why people switch to the younger people as opposed to the household head, one of the things. That was built into the program was given this age brackets for it, so 15 to 45, and I think while that still allowed many of the caregivers to be in, um, it excluded a lot of their spouses because they're typically a large age differential between the spouses so I think they were just not eligible so that's probably something to keep in mind for other types of programs if we're if we're concerned about kind of capture in them yeah and the second question about. Um, the ability of different household members, I think this is important because. I guess there are two ways to think about it. One is if you have the high ability individual, then maybe you do want them to be the one to have the opportunity so that they can produce, um, they, they're more productive, and then the household can benefit from there. And then there's equity concern that if they're the ones that get all the opportunities, then what about the less, um, able household members? I think that's, I mean one of the things that we want to dig more into both, um. Some of the data we have now and in future is to look more at the specific characteristics of these household members so wanting to learn more about if they've previously run any kind of enterprise with the education levels or other measures that we have of what features of the ability might be predicting their performance. I think that would guide us a little bit, but I think at the end, um. For this type of setting I think there tends to be a shift in in prioritizing the efficiency argument in the hopes that this then will be redistributed across the household members. I'll just speak to the question of scale that uh or scaling up that that Rachel asked. I think that's an, uh, an interesting challenge, and I think it's, uh, others might have different views, but I think most of the types of interventions that we're talking about here have modest or no effect on people's ability to employ others outside their own household. So this is, these are, you know, at best moving people from self-employment businesses that don't exist or doing very little for their household income to things that are doing a bit more for their household income, um. And I think the question of where might demand for wage for wage employment come from, where where where might that sit is a really crucial one. The answer might not be that it comes from starting from really small businesses and growing them into bigger ones, but thinking about businesses that start on a different level in the first instance with a level of capital intensity and things like that that these businesses don't have and think of that as a place where wage demand might be or demand for wage labor might be generated instead. So, um, I had a question about, uh, the general generalizability of, uh, some of the findings beyond sort of hairdressing and tailoring, I think from Kehindi, um. I, yeah, I mean I think as long as uh one can develop a good test, at least, uh, you know, if I can develop a masonry test and a carpentry test, I think um the sort of theory of change would be um applicable and um again we'll have to test to see whether it works but I don't see any issues there particularly that I can think of but I think one would wanna test it first, um, and create a good test. Um, and just to sort of piggyback on what Andy was saying about the, the scaling, I think the other thing that I think is challenging is that many of these, many people in self-employment are not there by choice. And I think if you were to find people who have great ideas, brilliant ideas, who want to be entrepreneurs, identifying these types of people that can actually grow and sort of, you know, change, you know, bring in new ideas, bring in new products that can actually people want to buy and can then, you know, that that's, that's, that's really hard to find those types of people, um, so, um, yeah, I think that's one challenge is who goes into self employment. Uh, exactly. So, uh, let me say, I'll, I'll answer the question that Tamar didn't ask. We heard some, some good news and some not so good news, I think. But what's funny about the not so good news is it seemed like better news the first year. OK, there were income effects in the program that I looked at and in the grants that I looked at in the first year. And so for a lot of interventions, including some here, we so far only have evidence on the good news part of the program. So how these things evolve, and I think how many years is the tailoring stuff that you've done? You're participant this long. How many years is that now? Yeah, I mean, we, this, the 20, yeah, we started 2013, yeah, to 2022. So some of this is quite long, but I think these things do have dynamics. So that's the first thing I wanted to point out. Um, I agree completely with, uh, Andrew and Isaac's answers to the question of scale and how to find ways of scaling this in the project that I've done as well. Very few of these young women go on to employ someone else. So then if you're asking where can you find the partners to do these kinds of things with, that's a difficult question. I agree completely with what you said. Uh, to Karen's question about, uh, are the micro franchises still around, one way of answering that is to say, what is it that the young women who are still running self-employment businesses, what are the sectors in, and the ones who started in the mic in the hair salon micro franchise. are much more likely to still have a hair salon as one of their businesses, so to some extent they don't close, but they decrease the intensity of their work. It's something that they've got as an option in their portfolio. I think it must help them cope with shocks, though I don't have a sort of specific measure of shock coping, but their living conditions are a little bit better, and they report greater food security. So in some way that suggests that yes, it must. On generalizability to Kehinde's question, so, uh, which by the way is almost exactly an interview question I was once asked when I was applying for a job as an economist. Anyway, that question we have a little bit of evidence on because we have at least two sectors and though it's not randomized which of the two, micro franchise models you're in, uh. There's variation both in what preferences people express, so we can exploit that heterogeneity, and there was differential availability of the different models across some rounds of rollout and across either of those ways of looking at it, a kind of more exogenous way or a more endogenous way. We don't see. Consistent evidence of differential impacts. So it seems like the two models were very similar in what limited evidence we have. So that suggests that we could go further. The theory of change isn't specific to either of those things, but what sector would you come to next? I don't know. Uh, Farhan asked about, uh, marriage and childbirth patterns. I don't think there's any impact that we find on eventual household size or, uh, which is, or household configuration, so no effects there, um, yeah. Let's take a few more. I saw a hand up there before. Um, OK, there's 1234. Um Hello, uh, I'm Priya. So I'm a nonprofit and we actually work on the micro entrepreneurship in India. So, uh, from the last 10 years we're working with the women and the youth. So we understood that when the first year when you start the business, everyone, when we do the as a cash, you give a grant. Sometimes they just come to get the cash grant, that's why they attend the training. And after the attending the training, many times it happened that just, you know, uh, that incentive towards the getting the cash because of that they continue with the, you know, training part. OK, I want to do the business, but, you know, at the next, at the last step, we understood that many times this woman actually don't want to be an entrepreneur, they just want to be, have that cash grant, and because of that they get involved. So that is one thing which we face the challenge. Second challenge was that, you know, many times it happened that. The women and the youth, when we talk about the both ways, youth start a business, we help them to start, and once they get a good job. They switch to the job. Once they're laid off from the job, they can come to the, you know, the livelihood opportunity. So that is one of the challenges with the youth. So where we have to work on the model, what are the models are good for the youth because women and the youth, we can't keep it on the together. So I think there needs to be a different model. So, yeah, what is your point of view on that? Thank you. Hi, um, thank you so much for the presentation. My name is Sonia, and I am an undergrad student from the University of Notre Dame. So my question is more about like the landscape about, uh, financial incentives like right now. I was wondering like when we were talking about like financial incentives, are we talking more about like grant or we're talking more about like franchisement or are. Are we moving like more towards like microfinance, like micro loans, something like that. And also when we're talking about like offering those incentives to people, are we, like, how do we make sure, like how do we determine like how much are we offering people and also like are we offering to like individual like household or like who, like who in the household. So thank you. Over there and then behind him. Hello and uh thank you for your presentation. Uh, my name is Frederic Aubury. Uh, I'm working at the World Bank. Uh, I have uh, a couple of questions for uh Isaac and BT about apprenticeships. Uh, the, the first is about the, your opinion on other. The advice you would give to a policymaker in your uh experiment you are targeting the intervention is targeting the masters to indirectly, uh, target the apprentices, and I was wondering if you have any opinion about policies that target the masters or policies that would target apprentices directly. Uh, the reason why I'm, I'm asking that is because I'm, I'm wondering to, to what extent the fact that the master is retaining apprentices. Is telling us about they're willing to retain this productive employee, or does that mean that apprentices do not have outside option, uh, outside option outside of the workshop? They, they are lacking of opportunity to use their skills outside of the workshop. So I was just wondering whether it might make sense to target also apprentices in some way to help them in their transition outside of the workshop. Thank you. And last one there. Thank you for your presentation. My name is Aida Martinez. I'm an intern in the World Bank and I'm a student in the Master of Public International Development Policy at Duke University. And well this is more common for Owen. I was thinking that as well that uh maybe the lack of willingness of being an entrepreneur uh it was the reason to look those uh fates and the effects of both franchise and the grants and the second thing it was that, uh, women. Uh tend to like dusting more a larger share of their incomes to uh household expenses that could be another reason and when you mention that uh the well like the grand beneficiaries uh were not married, uh, I was thinking that maybe is not the case for all of the participants of this uh like RTT but um. Economic violence is not only perpetrated by uh husbands but also for parents, for sisters, for uh siblings in general so I was thinking about that like maybe uh looking at that uh could help to understand a little bit more why those effects change. Thank you, thank you very much, Isaac. Why don't we start with you because there was a specific question to you. Yes, um, so thank you so much for the question. Um, so, um, in terms of, um, you know, how do you target masters or train or apprentices and sort of how you think about the sort of the, I guess the outside options, what I would say is, um. There's a very complex sort of uh tradition I would say of you know determining when a uh against the traditional system right so you you enter the apprenticeship and basically the you know the apprent the master trainer basically has a lot of power over when you compete right. Um, and so, and this is, I think, part of the, I think the contractual sort of issues and the monitoring issues, um, you basically end up, and again at the end of the day, basically they graduate, so to speak, and they will move away from the firm, and there's a question of like, OK, when are they going to graduate, and the, and the master is the one who actually determines that, right, so what you see is basically a lot of training that just gets dragged out for a really, really long time. And I think that this is evidence of a part of this monopsonistic sort of um power that they have over them to sort of control their lives because basically. When you sort of say I did an apprentice, I, you need to have like like a reference letter from them, right, so they can't leave until they get that reference letter and there are these sort of um trade associations that are very powerful that sort of, you know, have these customs about so so there's just a lot of I think um um I think norms and sort of societal sort of structures here that I think, um, really I think um. Put apprentices at a disadvantage where they end up sort of staying in these really low paid apprenticeships for a long time and the master trainer is able to basically extract a lot of rents, um, so yeah I think there are, I think ways to really speed up training to targeting the um uh trainer as we did. Uh, and I think, yeah, there are also probably complementary ways we can sort of help the apprentice move faster, but I think one of the things we need to do is sort of systemic where we basically need to sort of put some more regulation and oversight to make sure we're not getting people in apprentices for 5 years, right? So I think this is, I think, a role for government to really come in and sort of, I think, regulate the system a little bit more and sort of put standards and, um, and sort of I think really just. Upskill and upgrade this apprenticeship system a little bit more. Thanks. Let's move to the rest of you to pick up on the questions around. What if people really don't wanna be an entrepreneur, um, and, and then some of the design elements that were brought up, uh, why don't you start Owen and then we'll go. OK, uh, yes, I'm not sure that people think that entrepreneurship is the thing that they have always wanted to do, uh, but in an environment where there's really low labor demand from the wage sector, uh, it's helpful to have at least as a coping strategy if not if not something more. So I think that's an accurate description of a lot of the people here. I should say we didn't advertise cash grants. We said if you're interested in starting a business and then we said, uh, good and bad news. The business training is full, but I do have something good, and people didn't know that that was coming, so I don't think we faced particularly adverse selection of the type that you described, but, uh, completely fair point, um. With respect to the point about uh the the roots of economic violence may not exclusively originate in the husband and things like that uh. Completely agree and one of the things we did that we were frustrated by is that we looked for the predictable heterogeneity. We looked for baseline characteristics, who the woman was residing with, uh, how much education she had, a host of background characteristics, and we used a variety of machine learning methods to ask, does any of those background characteristics predict whether this is gonna work out for her in the long run or not, and we were unable to find a reliable predictor along those lines. Uh, so I think it's still open to figure out what to do there, and I forgot my best anecdote in answer to Kehinde's question. We actually didn't come here originally to study, uh, hairdressing and, uh, food carts. The original micro franchising model that the IRC had done in West Africa and was really happy about was a refrigerated goods micro franchise called Ice Ice Baby. So I think there's at least one more sector out there, Andy. Um, very quickly, yeah, very briefly, I think the, the Priya, your, your anecdote about, um, conditionality, I think is an interesting one. I think speaks to, uh, an important challenge. So one thing is, you know, do people wanna become an entrepreneur, and then the other thing is kind of once you're on a path toward that, you know, should you force them or, you know, economically incentivize them in strong ways to go through particular, you know, trainings targeted at particular things, um, I think that model is predicated on the idea that there's some market. Market failure out there that people don't know that this training is really going to be great for them, don't realize the value of it. That may be true in some circumstances. I think the evidence we're seeing in a variety of these studies is that those benefits are modest and that has therefore a double tax, right? You're both spending money from the organization's perspective on things that aren't, you know, that could be cash that does good things for somebody else and you're taking up people's time that they could be putting into the enterprise itself or other things that they might enjoy in their life. So I think that's worth thinking really carefully about. Right, so there was a question about how the financial amounts are set, how much to offer. I think I was looking at the range of amounts. It's somewhere between $200 to $300 tends to be the amount, and I think it's based on what could be used to start up a micro enterprise in the setting, um. It's very I completely agree with the other panelists that this is really a stopgap measure in many cases when I was making the title of my slide when I wrote the choice of entrepreneur, I chuckled to myself that no really these are not entrepreneurs, these are, you know, some kind of sustain like sustenance subsistence, um, livelihood activities and you know I when I always think of this big problems of youth unemployment I. The questions are much bigger than I feel like these are all micro solutions to a macro problem. There are no jobs. That's why we have all of these programs that are that are at play and you know we're talking about the Bangladesh story early and I feel like if all of us could get that. In our countries like people that can employ en mass young people and um really find ways to to address these you know this pyramid that all of us are looking at and and very worried about so um that's a very good place to end this discussion. Thanks to all of you and thanks very much to the panel. A round of applause please. OK. I Oh, OK, they don't have to. All right. Should we? Uh, Uh, I was told that would be an issue, so I'm fine. I'll just sit here. OK, all right, um, I wanna ask how many of you, uh, participated in the discussions yesterday as well? OK, quite a, quite a few, all right, um. I wanna thank uh the Center for Global Development in particular for giving me this opportunity not just to moderate this wonderful panel but also to give uh closing remarks. And I was really wondering yesterday after hearing the discussions, what I can bring to this audience by way of closing remarks at the end of the two full days of amazing presentations, big ideas, fruitful discussions, and I thought I would lean on my comparative advantage. And you might ask what is that? And I would respond by saying I am retired, so I have been away from these conversations for almost 10 years and what that does for me is that it gives me some perspective, um, it also allows me to plead ignorance because I will say some things that I don't know a great deal about. So what I want to do is I want to talk about what I heard during these two days, uh, and especially today and reflect on how different. What I heard was from what I knew, what the discussions were 10 years ago, where I was in the midst of these conversations. So there was a lot of discussions on challenges and what I heard on that side is that uh many of the issues facing the global community in the developing world today are not entirely new, but they are definitely qualitatively different. Yesterday's presentations focused on these big issues. Digital technology that was definitely around 10 years ago, but it is much more ubiquitous today and AI in particular promises to transform our lives, our societies, and our economies in ways that could not be imagined 10 years ago. Second, the climate agenda again, we have had warnings about the climate crisis for many years unheeded, but I have a sense that the urgency of the climate agenda now has become a mainstream topic, and there are constant calls for action on raising trillions of dollars and changing the trajectory of economic production. Third is that all these, this is happening, these two major drivers of change, uh, they're happening at a time when we have historically low growth rates globally. And in a geopolitical context where many OECD countries have turned inward. And both those things severely curtail the opportunities for low-income countries to grow through trade. That has been the traditional path as demonstrated by East Asian countries to growth into middle-income and higher-income status, and also to take advantage of migration because the same inward-looking policies on the part of wealthy countries are also closing off those opportunities. On top of this is the accumulating debt burden for many developing countries that we heard about yesterday. Which ends up reducing the fiscal space for investment and infrastructure and the social sectors. So that's for the challenges. What did I hear on the side of solutions, and I'll focus on today. Um, and today was meant to focus on, as I said, on the economic inclusion. Of especially women and youth. A lot of the conversations that we had today, the discussions at the panels are topics that I was very familiar with in many ways the same, including creating jobs through beauty salons and cosmetology. Um, And many of the discussions focused on reducing the constraints to the labor market participation of women through various means, uh, including changes in societal norms and, um, through safety nets, cash transfer programs, uh, credits. And outside encouraging labor, greater labor force participation on the part of women by reducing these constraints, we had the discussion most notably in my panel today on how to encourage our programs to foster more entrepreneurship. Entrepreneurship, as we all ended up saying may be too big of a board for what's going on. It's really, uh, self-employment and household businesses that we're talking about and both of these types of interventions, removal of barriers as well as working on the small micro enterprises really take labor demand as a given. And in the constant in the context of a given labor demand, what can we do to encourage the economic inclusion of women and youth and these are the conversations that we had today, some incredible creative experimentation. That ultimately. I feel may not move the needle. On the discussion around women, I think there are some big ideas there. If we're able to change societal norms, if we're able to remove some of the barriers as in providing childcare services, for example, to women that we heard that would encourage them, enable them to participate more, I can see the potential for scale and economy-wide impacts. I have a difficult time doing the same for the types of interventions that we talked about in this panel. They are important, as we said, because for the moment at least these are the only things that are available to poor people that don't have access to many opportunities and certainly not wage labor. So, Um, two main takeaways for me. I think we need to think more boldly and more creatively about addressing the challenges of the day which are the ones that I talked about in terms of the. Uh, technological improvements and the climate agenda we have learned tremendous amount through a portfolio of impact evaluations we have learned how to think about value for money, which is, um, what, uh, uh, Rachel has been emphasizing. And these are all very important learnings, but I think the challenges of today require us to get outside our comfort zones and think more boldly. I think there's significant room for enhancing our knowledge about what economic inclusion will look like in a world that's dealing with the climate change and the promise and threats of digital technology. How can AI technology and innovation more generally be harnessed to improve labor productivity, especially for low-skilled workers? Can we think about experiments that demonstrate how AI can complement rather than supplant current workers in the education sector, in agriculture, in the health sector, in social work. In the area of the green economy and. I really don't know what that means, and I found out that many people don't know what that really means, but it is a thing. What scope is there for low income countries to be part of the supply chain for some products? If not solar panels and batteries, which I was disabused of yesterday, something else perhaps, I don't know exactly what. What about um investments in adaptation to reduce the costs of climate induced disasters in low income countries, same countries that are paying for the costs of a problem that they really had no part in producing. Experiments to produce heat resistant crops, I think Rachel mentioned this yesterday, adapt infrastructure to become more resilient to climate change. There was also an interesting panel yesterday on the revival that industrial policy seems to be experiencing. And this exploring the scope for more activist government policies to spur growth. What I did not hear in that panel in that discussion was any attention to distributional concerns. So if innovation, the green economy, and some new form of industrial policy are to produce gains for the poorest segments of society, for youth and for women. I think that people that worry about these concerns have to raise these issues and be part of those conversations. So my second point is that I believe this agenda, distributionally sensitive innovation, climate transition, and new industrial policies need to form a big part of future research. For people who worry about poverty reduction, jobs, and human capital. I think there's a significant need for generating evidence on what works and learning through experimentation. And for this to happen, I really think there needs to be a much closer collaboration between macro and microeconomists. In other words, more coherence, the word of the conference between day one and day two of this conference. I was expecting to see more of a connection between the issues raised on day one. And what we talked about today, I mean, Rachel touched upon why it is that we're talking about human capital today, which is an intributor contribute uh important contributor to growth, but I think a more explicit link I would have found more helpful. I was very happy to hear Michael Kramer's thoughts yesterday on how to encourage innovation that delivers social benefits. More government and private sector collaboration through appropriate incentives potentially mediated through new institutional arrangements. There is clearly room for more creative thinking along those lines, and I trust that you all will rise to the challenge because I'm gonna go off into my retirement again. Thank you very much for listening to me. I.

showAllTranscripts

no

duration

PT1H41M5S

scene7File

worldbank/ABCDE2024-Day2-Session3-1

scene7Domain

scene7FileAvs

worldbank/ABCDE2024-Day2-Session3-1-AVS

title

ABCDE2024 Day2 Session3

description

ABCDE2024 Day2 Session3

showTimestampAndTranscript

no