S h li d Child L b Schooling and Child Labor: What role ... · [email protected]. Educational policies zCosts of /returns to schooling zSupppp yly - Accessibility zRepetition

S h li d Child L bOslo Workshop on Child Labor and EFA - October 20-21 2008

Schooling and Child Labor: What role for education policies?What role for education policies?

M M dMarco Manacorda Queen Mary University of LondonCEP- London School of Economics & UCW

[email protected]

Educational policiesp

Costs of /returns to schoolingSupply - Accessibility pp y yRepetition policiesQualityQuality MM10

Diapositiva 2

MM10 Marco Manacorda_2, 19/10/2008

Why education policies?y p

School attendance easier to monitor

Caution: work in combination with school / idlenessschool costly (siblings’) y ( g )CAUSALITY crucial to design policies

Repetition in Primary School and Gross E ll t R t i S d S h lEnrollment Rate in Secondary School

6

FinlandoIrelando21.

6

Bahrain

JordanwBelarusBulgaria

Czech Republico

EstoniaHungary oPolando

ArgentinawChilew

JamaicawUruguay w

Austriao

a do

Franceo

Germany o

Irelando

Italy o

1.2

men

t Higher repetition: lower enrolment

Algeria

Kuwait

Lebanon

Saudi ArabiaTunisiawAlbania

Azerbaijan

ChinawIndonesiaw

SamoaThailandwBelize

Boliv iaColombia

Costa Rica

Dominican Republic

Jamaicaw

Mexicoo

NicaraguaPanama

Paraguay w

PeruwTrinidad and Tobago

Indiaw

Botswana

Malawi

.8E

nrol

lm lower enrolment

Morocco

Sudan

Cambodia

Guatemala

Maldiv es

AngolaBurkina Faso

CameroonChad

Comoros

Equatorial Guinea

GuineaGuinea-Bissau

Lesotho

Madagascar

Malawi

.4

Ethiopia

0

0 .1 .2 .3 .4Repetition

Repetition in Primary School and Child L bLabor

.8.

..and higher child

Cameroon

Chad

Ethiopia

4men

t

labor

Albania

Cambodia.4E

nrol

lm

Angola

ArgentinaAzerbaijanBelarus Belize

Bolivia

Chile

ColombiaCosta-RicaDominican Rep

Guatemala

IndonesiaJamaicaLesotho

Malawi

MexicoNicaragua

Panama

PeruSudan

Thailand

T i id dChile Dominican-Rep.Finland FranceGermany

PanamaParaguayTrinidad0

0 .1 .2 .3 .4Repetition

Repetition Rates in Primary School and GDP itGDP per-capita

4Cambodia

.4

AngolaCameroon

ChadComoros

GuineaGuinea-BissauMadagascar

.3io

n

Caution: relationship might

Algeria

M

Tunisiaw

Bhutan

Burkina Faso

Equatorial Guinea

Lesotho

Malawi

.2R

epet

it relationship might be spurious

BahrainKuwait

Lebanon

Morocco

Saudi Arabia

Sudan

AlbaniaBulgaria E t i

IndonesiawThailandw

Argentinaw

Belize

Boliv ia ChilewColombia

Costa Rica

Dominican Republic

Guatemala

JamaicawMexicoo

NicaraguaPanama

Paraguay w

Peruw

Trinidad and TobagoUruguay w

FranceoIndiawMaldiv esBotswana

Ethiopia

.1

Jordanw Belarus

g

Croatia Czech RepublicoEstonia Hungary oPolandoAzerbaijanChinawSamoa

Boliv ia Chilew AustriaoFinlando

Germany oIrelandoItaly o0

0 5 10 15 20 25 30GDP

Grade Repetitionp

Learning benefitsSocio-emotional consequencesSocio emotional consequences Knowledge/discipline Improves quality of match Deterrent against underperformanceDeterrent against underperformance GE + externalities

Does repetition cause drop-out and lower h l tt i t?school attainment?

Natural experiment (Manacorda, 2008)Uruguay: automatic grade failure if more g y gthan 25 days of absenceData follow students for (up to) 5 yearsData follow students for (up to) 5 years

Grade Failure by Number of Missed School D i J i Hi hDays in Junior High

126

.8 Discontinuous rise

25

.6

in repetition

25

.4.2

0

0 5 10 15 20 25 30 35Missed school days

Pupils’ absenteeism and educational tt i t (5 i d )attainment (5 years window)

33.

52

2.5

25

11.

5

Discontinuous fall in attainment

26

0.5

0 5 10 15 20 25 30 35Missed school days

uncensored censored

Pupils’ Absenteeism and stay-on rates (5 i d )years window)

43.

54

53

2522.

5

Discontinuous fall in stay on rate

26

1.5

in stay-on-rate

1

0 5 10 15 20 25 30 35Missed school days

The cost of grade repetitiong p

Grade repetition• After 5 yearsy• Lowers school attainment by half a year

Increases drop out by similar amount (after• Increases drop out by similar amount (after 5 years)

How to remedy repetition?y p

Pre-schoolingEarly Human Capital investmentEarly Human Capital investmentSocialization to schoolEnforcementAdditional benefits (mother’s work)Additional benefits (mother s work)

Does re-school lower repetition?p

Berlinski, Galiani and Manacorda (2008)Compare siblings’ school trajectoriesCompare siblings school trajectoriesUruguay: universal pre-school

Preschool attendance and educational tt i t ( ibli ’ i )attainment (siblings’ comparison)

Diverging 1

g geducational paths between children who attended pre-

.5

school compared to siblings who did not

0

-.5

6 8 10 12 14 16age

Preschool attendance and stay-on rates ( ibli ’ i )(siblings’ comparison)

.3

Diverging stay-on rates between children who

tt d d

.2

attended pre-school compared to siblings who did not

.1 not

0-.1

6 8 10 12 14 16age

Costs and benefits of pre-schoolp

By age 15 +0 79 years of education+0.79 years of education + 27 p.p. in school

Benefits-Costs ratio >2.2

School accessibilitySchool accessibility

Higher distance Reduces attendance Increases specialization Effect on child labour ambiguousEffect on child labour ambiguous

Distance to school and child labour

Kondylis and Manacorda (2008)Compare households in same village p gRural Tanzania: villagization policyControls for distance to other infrastructuresControls for distance to other infrastructures

Distance to school and children’s time use

Naif With Naif estimates

With controls

School 1 670*** 0 365**School -1.670 -0.365

Work 0.425*** -0.270

Work and school -1 101*** -0 519***Work and school 1.101 0.519

Work only 1.526*** 0.249

School only -0.569*** 0.153

Neither school nor work 0.145 0.117

Results by gendery g

males femalesmales females 1. School -0.580*** -0.057 2. Work -0.137 -0.492 3 W k d h l 0 734*** 0 1983. Work and school -0.734*** -0.198 4. Work only 0.597*** -0.294 5. School only 0.154 0.140 6. Neither school nor work

-0.017 0.352

Distance to school and children’s time use

One additional km in distance to school:+ 0.3 p.p. school attendancep pNo significant effect on child labor/idleness

Child l b i l tiChild labor inelasticEffect largely for boysg y yExternal validity? E t i i t i iExtensive vs. intensive margin

What role for education policies?p

I t i ibilitImprovements in accessibilityreduce incentives to drop-outand to engage in economic activity

In poorer countries child labor inelasticIn poorer countries child labor inelasticLabor demand CCT interventions ?

Repetition policies harmfulp pEarly school exposure viable option