Do Conditional Cash Transfers Function as Safety Nets? Evidence from Malawi.

Sarah Baird (George Washington University)Craig McIntosh (UC San Diego)Berk Özler (World Bank)

Not for citation without explicit permission from the authors. 2

Background and Motivation

The conventional argument for CCTs as safety net programs:

Regular cash transfer payments allow households to smooth shocks, save for future.

Decreasing relative price of schooling may further prevent households from using withdrawal of kids (particularly girls?) from school as a buffering tool.

However: If the incentives provided by a CCT are strong

enough to keep kids in school even in the face of shocks, well and good.

If not, then conditioning the transfers on school attendance will undermine the social protection dimension of CCTs, expose households to the further shock of withdrawal of transfers.

Not for citation without explicit permission from the authors. 3

Background and Motivation

Why may dropout expose girls to additional risks in SSA?

There is a knife-edge relationship between leaving school and marriage & pregnancy in much of SSA.

Teenage girls typically have HIV rates that match men 10 years older, much higher than their male age-mates.

Economic pressure may play a direct role in driving sexual risk debut and other risk factors.

On the face of it, it appears that ‘Conditional Cash Transfer’ programs, giving money in return for regular school attendance, might have an effect on marriage, pregnancy, and HIV in this context as well as simply on schooling.

Not for citation without explicit permission from the authors. 4

Background and Motivation Conditional Cash Transfers (CCTs) are “…

targeted to the poor and made conditional on certain behaviors of recipient households.”

As of 2007, 29 countries around the world had some type of a Conditional Cash Transfer program (CCT) in place, with many others planning or piloting one (World Bank, 2009)

Unconditional Cash Transfer programs (UCT) are also common and have also been shown to change behaviors on which CCTs are typically conditioned.

Not for citation without explicit permission from the authors. 5

Background and Motivation Whether the conditions are necessary has

recently become a bigger debate. Proponents of CCTs point to market failures,

which are addressed by the condition. Evidence (a few slides from now) suggests that the

effects of UCTs are (or would be) small to non-existent.

The Zomba Cash Transfer Program Zomba, Malawi: adolescent girls aged 13-22 who were

never married as of baseline. One arm received CCTs, another arm receives the exact

same transfers, but unconditionally (UCT). Provides an experimental estimate of the impact of the

conditionality itself

Not for citation without explicit permission from the authors. 6

The research questions of SIHR: Do the conditions themselves matter in CCT

programs, or would households have made the same choices anyway? To answer this, two randomized arms, one CCT, other

UCT. Do the behaviors that we care about respond to

transfer size? Transfer amounts were randomized between $5 & $15

per month. Does it matter whether transfers go to

households or girls? Split of transfers randomized between guardian and

beneficiary. Do the programs have a meaning effect on STDs

and HIV? Conduct biomarker testing for HIV, HSV-2, and syphilis.

Are impacts on beneficiaries just ‘diverting’ to non-beneficiaries? Randomize saturation of treatment at village level and

track non-beneficiaries to be able to estimate spillover effects experimentally.

Not for citation without explicit permission from the authors. 7

Zomba Cash Transfer Research DesignMalawi Research Design:

Baseline Dropouts (N=805)

T2a S2 T2b S2Baseline Schoolgirls Within- Uncon- Within- (N=2,741) CCT village ditional village

control CT control

C1

Purecontrol

T2 Unconditional (N=27)

T1CCT

CCT

C2Pure

CCT

S2

Treatment EAs (N=88)

Control EAs (N=88)

controlWithin- village control

Conditional (N=46)

T1

T1 Only (N=15)

T1

Individual transfer randomized at individual level

Household transferrandomized at EA level

EA treatment saturation randomized

Not for citation without explicit permission from the authors. 8

Zomba District & Town

Evidence on the relative effectiveness of CCTs vs. UCTs Evidence points us heavily in favor of

CCTs. de Brauw and Hoddinot (2008); Schady and

Araujo (2008) – using implementation glitches Bourguignon, Ferreira, Leite (2003); Todd and

Wolpin (2006) – using structural models.

Two common themes to these papers: All from Latin America (Brazil, Ecuador, and

Mexico) All with very little ‘income’ effect.Not for citation without explicit permission

from the authors. 9

Not for citation without explicit permission from the authors. 10

Zomba Cash Transfer Program Implementation For CCT recipients, attendance is checked

monthly at each program school using a combination of physical checks and phone calls (with random spot checks in Year 1, i.e. 2008).

For CCT recipients, the payment for the next month is withheld if attendance is below the required threshold. However, the girl remains in the program.

UCT recipients receive their transfers by only showing up.

Not for citation without explicit permission from the authors. 11

Making the initial offers, Fall 2007.

Not for citation without explicit permission from the authors. 12

Program impacts on schooling: Enrollment

(1) (2) (3) (4) (5) (6) (7) (8)

Total Terms

Year 3: 2010

Term1 Term2 Term3 Term1 Term2 Term3 (6 terms)Term 1, Post-

program

Conditional treatment 0.043*** 0.044*** 0.061*** 0.094** 0.132*** 0.113*** 0.535*** 0.058*

(0.015) (0.016) (0.018) (0.041) (0.035) (0.039) (0.129) (0.033)

Unconditional treatment 0.020 0.038** 0.018 0.027 0.059 0.033 0.231* 0.001

(0.015) (0.017) (0.023) (0.038) (0.037) (0.039) (0.136) (0.036)

Mean in the control group 0.906 0.881 0.852 0.764 0.733 0.704 4.793 0.596

Number of observations 2,023 2,023 2,023 852 852 852 852 847

Prob > F(Conditional=Unconditional) 0.173 0.732 0.067 0.076 0.014 0.020 0.011 0.108

Dependent variable: =1 if enrolled in school during the relevant term

Year1: 2008 Year2: 2009

Panel B: Program impacts on teacher-reported school enrollment

Not for citation without explicit permission from the authors. 13

Program impacts on schooling: Test Scores

English test score (standardized)

TIMMS math score

(standardized)

Non-TIMMS math score

(standardized)

Cognitive test score

(standardized)(1) (3) (2) (4)

Conditional treatment 0.140*** 0.120* 0.086 0.174***

(0.054) (0.067) (0.057) (0.048)

Unconditional treatment -0.030 0.006 0.063 0.136

(0.084) (0.098) (0.087) (0.119)

Number of observations 2,057 2,057 2,057 2,057

Prob > F(Conditional=Unconditional) 0.069 0.276 0.797 0.756

Table VI: Program Impacts on Educational AchievementDependent Variable:

Program impacts on marriage and pregnancy

Not for citation without explicit permission from the authors. 14

2008 2009 2008 2009

(1) (2) (3) (4)

Conditional treatment 0.007 -0.012 0.013 0.029

(0.012) (0.024) (0.014) (0.027)

Unconditional treatment -0.026** -0.079*** -0.009 -0.067***

(0.012) (0.022) (0.017) (0.024)

Number of observations 2,087 2,084 2,086 2,087

Mean in control 0.043 0.180 0.089 0.247

Prob > F(Conditional=Unconditional) 0.024 0.025 0.265 0.003

=1 if ever married =1 if ever pregnant

Dependent variable:

Table VII: Program Impacts on Marriage and Pregnancy

How do we reconcile the differential program impacts on schooling vs. marriage & pregnancy? To help think about this issue, we can categorize

the target population into three latent strata:1. UCT compliers: Remain in school under a UCT.2. CCT compliers: Remain in school under a CCT but not

under a UCT. 3. Non-compliers: Drop out of school even under a CCT.

UCT compliers will likely be similar under CCT and UCT.

CCT compliers have a schooling impact under a CCT only.

Non-compliers have an income impact under a UCT only.

Not for citation without explicit permission from the authors. 15

Marriage and Enrollment at Follow-up

Not for citation without explicit permission from the authors. 16

Enrolled Not enrolled Total

(1) (2) (3)

Control 1.7% 49.0% 20.2%

N (row %) 272 (59.7%) 184 (40.3%) 456 (100.0%)

Conditional treatment 0.5% 50.8% 16.2%

N (row %) 174 (69.2%) 78 (30.8%) 252 (100.0%)

Unconditional treatment 0.3% 25.3% 10.2%

N (row %) 82 (60.5%) 54 (39.5%) 136 (100.0%)

Total 1.1% 45.3% 17.4%

N (row %) 529 (62.7%) 315 (37.3%) 844 (100.0%)

Table VIII: Prevalence of being ‘ever married’ by school enrollment status during Term1, 2010

Impacts on partner age:

Not for citation without explicit permission from the authors. 17

0.0

5.1

.15

.2M

ale

HIV

rate

by

Age

, fr

om

DH

S

0.0

5.1

.15

Den

sitie

s of

Part

ne

r A

ge,

Contr

ol v

s T

reatm

ent

15 20 25 30 35Age of male partners

Control TreatmentMale HIV rate (right scale)

Always Active SchoolgirlsAge of Partners, by Treatment Status

Not for citation without explicit permission from the authors. 18

Shock & Program impacts on Expenditures: Food Eaten

at HomeSchool Total

Spent on Beneficiary

Spent by Beneficiary on Others

(1) (3) (9) (10) (11)=1 if Experienced Lower Crop Yield -2.918 1.406 -3.441 0.050 -0.020

(1.683)* (1.397) (4.898) (0.639) (0.031)=1 if Experienced Crop Disease/Pests -0.565 5.573 5.079 0.629 -0.004

(2.121) (2.546)** (4.815) (0.546) (0.062)=1 if Experienced Dead or Stolen livestock -2.770 -3.127 -11.959 0.656 -0.041

(2.048) (2.378) (4.895)** (0.531) (0.039)=1 if Experienced HH Buisness failure (non-ag) 0.401 -4.349 -4.235 0.505 0.037

(2.578) (2.175)** (8.160) (0.613) (0.059)=1 if Experienced Fall in Sale Price of Crops 1.078 7.092 7.997 0.528 0.041

(1.501) (2.157)*** (5.037) (0.547) (0.039)=1 if Experienced Rise in Price of Food 2.589 -1.003 -1.491 -0.154 0.065

(1.717) (1.733) (4.826) (0.378) (0.048)=1 if Experienced Death of HH Head -7.235 3.557 -11.587 1.513 -0.024

(2.187)*** (1.811)* (6.498)* (0.755)** (0.047)=1 if Experienced HH member illness or accident 3.435 -0.956 -2.135 -0.230 -0.027

(2.226) (1.374) (4.164) (0.482) (0.046)Treatment Dummy 5.098 5.870 18.735 1.852 0.122

(3.004)* (1.919)*** (7.193)** (0.771)** (0.057)**

Household-level Monthly Expenditures on:

Girl-level Monthly Expenditures:

Not for citation without explicit permission from the authors. 19

Program doesn’t change incidence of shocks:

02

46

8N

um

ber

of S

hock

s p

er Y

ear

1 2 3

By Treatment Status and Survey RoundNumber of Shocks per year

SG CCT SG UCTSG Control DO CCTDO Control

Not for citation without explicit permission from the authors. 20

Program doesn’t change response to shocks:

(2) (4) (6) (8) (10) (12)

Dropout Treatment 0.008 0.013 0.006 0.029 -0.027 0.009

(0.017) (0.051) (0.012) (0.025) (0.042) (0.011)

Conditional SG Treatment -0.013 0.011 0.011 0.020 0.011 -0.020

(0.015) (0.055) (0.016) (0.033) (0.040) (0.007)***

Unconditional SG Treatment 0.025 0.060 -0.020 0.002 0.005 -0.022

(0.021) (0.056) (0.010)* (0.028) (0.047) (0.008)***

Baseline SG 0.005 -0.028 0.017 0.025 -0.014 0.014

(0.014) (0.025) (0.009)* (0.014)* (0.028) (0.009)

Round 2 dummy -0.050 0.198 -0.029 -0.003 -0.380 -0.073

(0.017)*** (0.035)*** (0.016)* (0.027) (0.047)*** (0.017)***

Round 3 dummy -0.146 -0.372 -0.067 -0.245 -0.653 -0.107

(0.017)*** (0.033)*** (0.014)*** (0.024)*** (0.043)*** (0.017)***

Observations 8,481 8,481 8,481 8,481 8,481 8,481

Dependent Variable: How many times per year the household resorted to each of the following to smooth shocks:

Liquidated Assets or Livestock

Dissaved from Liquid

Savings

Worked More

MigratedNGO

Assistance

Did Nothing in Response

to Shock

Analysis on Response to Shocks; number of responses:

Social Protection from Mom only:

Not for citation without explicit permission from the authors. 21

Program helps insulate girls against shocks, but only in female-headed households!

# shocks experienced -0.016 -0.024 -0.020 -0.025 0.000 -0.019(0.024) (0.036) (0.010)* (0.017) (0.010) (0.016)

Female Headed Household -0.026 -0.112 -0.059 -0.023 -0.048 -0.041(0.053) (0.065)* (0.019)*** (0.026) (0.017)*** (0.025)

Treated with CCT 0.414 0.425 -0.035 -0.058 -0.003 -0.061(0.049)*** (0.068)*** (0.021) (0.029)** (0.023) (0.029)**

Baseline Schoolgirl 1.290 1.303 -0.148 -0.150 -0.222 -0.225(0.086)*** (0.085)*** (0.027)*** (0.026)*** (0.025)*** (0.025)***

Female Headed * # shocks 0.044 -0.026 0.008(0.049) (0.024) (0.024)

Female Headed * Treatment 0.000 -0.004 0.086(0.147) (0.045) (0.041)**

# shocks * Treatment -0.070 0.047 0.087(0.061) (0.025)* (0.026)***

# shocks * Female Headed * Treatment 0.211 -0.053 -0.160 (0.104)** (0.037) (0.038)***

Female Headed Households (CCT Sample versus Control)

Dependent Variable: Terms Enrolled Ever Married Ever Pregnant

Concluding Discussion Cash transfer programs have strong effects on

consumption. Do not appear to change the incidence of or

response to shocks. Social Protection role of UCT appears superior in

the sense that the CCT is cutting household off just at the moment that the girls are most vulnerable.

Social Protection role of the CCT in preventing shocks from adversely effecting girls is found only in female-headed households.

Clear evidence that adolescent girls in Malawi expose themselves to STD & HIV risk because of economic factors; they cease to do this when provided with income support.

Very little sensitivity to transfer amounts, splits. Suggests that a simple inexpensive program with

transfer to households will be just as effective.

Not for citation without explicit permission from the authors. 22

Thank you

Not for citation without explicit permission from the authors. 23

Offer Letters

Conditional Transfers The Zomba Cash Transfer

Program (ZCTP) with funding from the World Bank would like to offer you, _[NAME]_________, a cash transfer to help you and your family with the burdens of school attendance for the 2009 school year. By accepting this offer, in return for going to school you will be given:

You are receiving this money in order to help you return to school or stay in school. In order to receive this money you MUST attend school at least 80% of the days for which your school is in session.

Unconditional transfers The Zomba Cash Transfer

Program (ZCTP), with funding from the World Bank, would like to offer you, __[NAME]________, a cash transfer to help you and your family. By accepting this offer you will be given:

These monthly transfer amounts specified above are given to you as a result of a lottery. You are not required to do anything more to receive this money. You will receive this money for 10 months between February and November, 2009.

Not for citation without explicit permission from the authors. 24

1. The rules of the program were well understood by the girls in the UCT arm:

Not for citation without explicit permission from the authors. 25

2. Girls in the UCT arm knew about the CCT arm:

Not for citation without explicit permission from the authors. 26

Int: Earlier you talked of conditional and unconditional. What did you say about the rules for conditional girls?

Res: They had to attend class all the time…not missing more than 3 days of classes in a week – like I already explained.

Int: How did you say the program managers knew about the missed school days?

Res: They would go to the schools…For example, I have a friend, Jane [not real name], who was learning at Samama. They would go each month to the school to monitor her attendance, and if she was absent for more than three days she would not get her monthly money.

Summary of the conditions under which the UCT experiment took place:1. UCT beneficiaries fully understood their

treatment status and were never worried about not receiving their payments due to school attendance.

2. However, the UCT experiment did not happen in a vacuum. It took place in a district where a CCT program was simultaneously running in neighboring communities. Hence, the UCT experiment took place under a rubric of education.

Not for citation without explicit permission from the authors. 27