0:02 let me take you to an urban slum in the
0:05 North Indian city of
0:08 chandigar here water buffalos and
0:11 chickens are looking for their grain of
0:15 Life Pools of stagnant water act as
0:17 breeding ground for
0:21 mosquitoes I cross one such open
0:24 sewage and enter the residential area of the
0:26 the
0:28 slum step into a small room
0:31 room
0:36 this room houses 25 kids between the
0:41 ages of 3 and 6 years and is staffed by
0:43 a worker called the angan Bari
0:47 worker this daycare center is called an angan
0:49 angan
0:52 Vari even though there are alphabet
0:55 charts on walls there is no
0:59 lighting the temperature is over 100° fah
1:00 fah
1:02 but there are no ceiling
1:05 fans I take a picture of these kids and
1:08 show it to them many of them haven't
1:15 before I weigh all these kids who are
1:17 present at the
1:19 center and
1:28 malnourished this angari in chandigar is
1:31 not unique
1:35 all around the world 3 million children
1:37 die every year because they are
1:40 malnourished that's about four times the
1:41 size of
1:45 Boston half of all child mortality
1:49 happens directly because of
1:50 undernutrition even though there has
1:53 been steep declines in
1:56 undernutrition from 1980 to
2:00 2015 the global uh malnutrition rate for
2:05 instance has dropped from 40% to about
2:08 23% much remains to be
2:17 outlier on the xaxis here you see
2:20 malnutrition rates and on the y- AIS is
2:23 the per capita income each bubble here
2:25 represents a
2:28 country and the size of the bubble is
2:31 the population
2:34 as you can see India has higher
2:37 malnutrition than Ethiopia Democratic
2:39 Republic of Congo
2:41 Liberia and even though the per capita
2:44 income is the same as Ghana the
2:49 malnutrition is three times as
2:52 high what is the reason that despite
2:54 being one of the fastest growing
2:57 economies in the last 20
3:03 years India's malnutrition has
3:07 been declining only very
3:10 slowly has the income not trickled down
3:12 to the poor that doesn't seem to be the
3:16 case the poor are increasing in their
3:20 incomes so what could be the reason
3:24 behind High malnutrition in India and
3:30 undernutrition India's home to over a
3:33 third of the world's stunted children
3:36 and 1.27 million children approximately
3:39 die every year in India because of
3:48 worker why it is that almost half the
3:51 class is
3:54 malnourished she blames the mother she
3:57 says that the mothers lack the knowledge
3:58 on what to feed the
4:01 child they they don't feed the right
4:04 amount of calories and
4:06 protein now an angeni worker
4:10 herself can affect the health of a child
4:12 through two main channels she's supposed
4:16 to provide a midday meal to the kids one
4:18 the porridge Bowl as you can
4:21 see um so she's supposed to distribute
4:25 this amongst her kids in the class and
4:27 the other channel through which she can
4:30 change the health of the child
4:33 is by talking to mothers and guiding
4:38 them on nutrition and health
4:42 behavior so the worker blames the
4:47 mother I go further into the
4:52 slums and interview a mother and ask her
4:58 malnourished the mother
5:00 mother
5:06 state angan Vari is part
5:09 of uh the integrated child development
5:11 services which was set up by the
5:14 government of India in
5:18 1975 it's the world's largest Child Care
5:22 Program it encompasses over 1.3 million
5:25 daycare centers and each Center is
5:28 staffed by only one worker so there are
5:31 1.3 million workers in India trying to
5:38 children these workers are paid very
5:41 meager wages their wages are about $50 a
5:45 month which is about a tenth of what a
5:49 government assistant would earn in in uh
5:55 region so she blames a lack of worker
5:59 effort the worker doesn't teach me what
6:02 I should should cook the worker doesn't
6:04 show me that my child is undernourished
6:06 the worker doesn't show me the growth
6:08 chart that she's supposed to show and
6:15 absent so here was a problem the supply
6:18 side was blaming the demand side and the
6:27 side what could be done so we designed
6:30 an experiment in the city of chandigar in
6:36 2010 and uh and there were three
6:38 components to this experiment the first
6:42 component was a recipe book this was
6:44 done to address the lack of maternal
6:49 knowledge so to speak in the 20056
6:51 demographic and health survey for
6:54 instance nine out of 10 mothers don't
6:55 increase the fluid intake in their
6:58 children when their CH child has diarrhea
7:00 diarrhea
7:01 this is against the recommendation of
7:02 the World Health
7:05 Organization so we designed a recipe
7:06 book in collaboration with the local
7:08 food and nutrition
7:12 board that contained 10 simple easy
7:16 toake nutritious delicious calorific
7:20 recipes all of them contain local
7:23 ingredients and we said that what if we
7:25 just give this to the mothers this would
7:28 solve the maternal knowledge problem
7:30 hopefully and that could improve the
7:36 kids now remember that the mother had
7:39 blamed the state so we had to also think
7:42 about ways to motivate the worker and
7:44 when we had spoken with the workers
7:47 workers were obviously they they were
7:49 not satisfied with their wages they
7:51 wanted their wages to go up for the
7:53 amount of work that they were putting in
7:56 so we thought of an incentive scheme
7:57 wherein the worker would get a
8:00 performance-based bonus
8:03 and the performance would be based on
8:05 the decline in malnutrition that would
8:08 happen in her Center so she would get $3
8:11 for every child that improved a grade of
8:13 malnutrition so if a child went from
8:15 severely malnourished to moderately
8:16 malnourished that's one grade
8:19 Improvement or from moderate to mild
8:21 that's another gr Improvement so the
8:23 worker would get $3 for nutritional
8:25 grade improvements and then we'd also
8:27 subtract from the total bonus if there
8:29 were kids who declined in their nutritional
8:31 nutritional
8:34 status so for instance if three kids
8:37 went up and one kid went down then the
8:48 $6 4,11 children were weighed twice
8:50 during the course of this experiment in
8:53 2010 in the city of chandigar that I
8:54 just showed
8:57 you there were three treatment arms the
8:59 first treatment was the recipe
9:03 group treatment where um about 900
9:06 mothers got a recipe
9:08 book in the incentive treatment which
9:10 was the second treatment and the bonus
9:12 treatment for the
9:16 worker there were, 1,61
9:19 children and 38
9:22 workers in the combined treatment again
9:24 they were close to a thousand children
9:26 and the combined treatment the mothers
9:28 got the rescue books and the workers got
9:30 the incentives
9:33 the intuition behind the combined
9:41 tango
9:45 perhaps you need both the mother and the
9:48 worker to reduce undernutrition and not
9:50 one or the
9:54 other in economics this is called
9:56 complimentarity so to test for
9:58 complimentarity we had to test whether
10:00 the effect of of the combined treatment
10:04 was greater than the sum of the effects
10:06 of the recipe treatment and the
10:08 incentive treatment and we can
10:14 that the fourth group was the control
10:16 group where we just follow the kids over
10:23 grow and what we
10:27 found was very insightful because the
10:29 recipe treatment had no signific ific
10:34 effect the incentive scheme alone had no
10:36 effect and these were precisely
10:38 estimated zero
10:41 effects but the combined treatment
10:43 actually led to a significant drop in
10:45 malnutrition which was about 4
10:48 percentage points over a
10:51 quarter once we discontinued the incentive
10:53 incentive
10:55 scheme and weighed these kids a year
10:59 later they had the the gains that that
11:01 they had accumulated initially those
11:05 gains had persisted over a
11:08 year then we delved into why this
11:10 happened what were the mechanisms
11:14 driving this increase in weight for age
11:22 malnutrition a team of enumerators had
11:24 not only weighed these kids but they had
11:27 gone to the mothers and interviewed them
11:29 both before the treat treat M and after
11:32 the treatment at end
11:35 line what we found was that workers had
11:37 started making more home visits when
11:44 incentivized and when they went to the
11:46 homes they gave specific nutritional
11:49 advice to the mothers when the mothers
11:50 had recipe
11:53 books so the communication was much more
11:55 effective when the mothers had the
11:57 recipe books at home and the workers
11:59 when the workers were incentive
12:01 vised so this
12:04 communication Effectiveness actually
12:07 drove the complimentarity in Improvement
12:08 an improvement of
12:11 Weights in turn we also asked mothers
12:13 about the diet that they were giving to their
12:14 their
12:18 kids and we found that mothers improved
12:25 children that experiment was carried out
12:28 in 2010 since then I've carried out a
12:30 similar EXP experiment in Kolkata among
12:32 209 daycare
12:41 20145 um in 165 daycare
12:43 centers in both of these experiments we
12:45 have been able to
12:48 corroborate a decline in malnutrition of
12:50 roughly the same
12:53 magnitude every
12:55 quarter where workers got
12:58 performance-based pay and mothers got information
12:59 information
13:01 so this has been replicated over and
13:05 over again what we've also found is that
13:07 just increasing the wage without
13:11 conditioning it on the weight gain of
13:13 the kids is not as effective so just
13:16 increasing the wages as a fixed lump sum
13:21 amount is not as effective we found that
13:23 making workers compete against one
13:25 another is not as effective as just
13:33 you might think that after giving it
13:35 once if you give it if you implement
13:37 this incentive scheme again it would
13:39 have no effect but a renewal of
13:41 incentives also leads to a decline in
13:44 malnutrition of about 4 percentage
13:47 points finally withdrawal of incentives
13:49 did not crowd out intrinsic motivation
13:52 of workers it did not reverse the gains
13:53 that the that the children had
13:57 accumulated when the incentives were in
13:59 place and the main main channel that was
14:01 driving this was Mother worker
14:03 communication and we in a recent uh
14:04 paper I also
14:09 find that worker attendance went up so
14:12 um enumerators went in randomly and saw
14:13 uh whether the workers were present in
14:15 their classroom or not and they they
14:17 were more likely to be present when they were
14:19 were
14:22 incentivized now different contexts will
14:23 involve different
14:27 strategies what we can do is try and experiment
14:29 experiment
14:31 and sometimes experiments are going to
14:33 be successful and when they are
14:35 successful in the realm of Early
14:38 Childhood nutrition the benefit cost
14:40 ratios can be really really
14:43 high so think about what happens when a
14:46 child is no longer undernourished not
14:49 only is that child less likely to die
14:51 because of diseases such as diarrhea and
14:54 pneumonia but that child is also more
14:55 likely to come to
14:58 school the child is more likely to have
15:00 high higher grades in school to get more
15:02 schooling to have higher labor
15:05 productivity and eventually higher wages
15:07 economists can back out the lifetime
15:09 increase in wages and say that's the
15:12 benefit of not being undernourished the
15:13 benefit cost ratio estimate of the
15:16 combined treatment was close to 16 which
15:18 means that if you spend a dollar today
15:20 on implementing the combined treatment
15:22 over a long enough period you're
15:25 actually going to get $16 back so it's
15:28 in the interests of governments and NOS
15:30 and communities to actually
15:33 Implement policies that have high
15:40 ratios finally Improvement in child
15:43 health is just the first step towards
15:46 kids realizing their potential and
15:49 leading fulfilling lives thank you
