From Protection to Production:

The Productive Impacts of Social Cash Transfers

Paul Winters

Department of Economics, American University

Building Resilience and Assets for Food Security: Evidence and Implications for Feed the Future

presented by USAID and the Assets and Market Access CRSP

September 29-30, 2011, Washington DC

Motivation for studies

Objectives of Social Cash Transfers programs Social protection Reduce consumption poverty Induce investment in child health and education Break the intergenerational transmission of poverty

Motivation

Criticism of SCTs Focus solely on long-term poverty reduction

Weak link to productivity human capital, skills, labor market insertion

Agriculture? Miss opportunities to complement broader development

programs, particularly productive investment Investment in human capital to what end

Intergenerational transmission of welfare receipt? Productive vs. social (unproductive?) investment

Motivation

Productive cash transfers Can they be used for productive purposes as a

complement to broader development agenda? Agricultural technology adoption Agricultural asset accumulation With land formalization projects Promoting microenterprises…

Under what conditions?

Social cash transfers and productive choices

Transfers can influence consumption and production decisions when markets fail or are incomplete

Transfers as a source of liquidity Credit constraints potentially limit productive spending

and investment Transfers can induce spending and investment altering

production and the allocation of resources, including labor

Social cash transfers and productive choices

Transfers as a secure source of income Insurance and credit market imperfections limit the ability

of poor households to smooth consumption Poor households take action to manage risk ex ante and

cope with risk ex post Transfers provide regular income uncorrelated with other

income sources, potentially altering risk management and coping strategies, and therefore production choices

Objective

To look beyond the social protection function of SCT programs and analyze the impact of these programs on productive activities

Mexico’s Oportunidades Todd, J., Winters, P. and Hertz, T., 2010. “Conditional Cash

Transfers and Agricultural Production: Lessons from the Oportunidades Experience in Mexico.” Journal of Development Studies 46(1), 39-67

Malawi Social Cash Transfer Program Covarrubias, K., Davis, B. and Winters, P., 2011. “From Protection

to Production: Productive Impacts of the Malawi Social Cash Transfer Scheme.”

Boone, R., Covarrubias, K., Davis, B. and Winters, P., 2011. “The Impact of Social Cash Transfers on Agricultural Production: The Case of Malawi.”

Mexican Oportunidades program

Initiated in 1997 as PROGRESA in marginal rural areas (506 localities in 7 states)

Targets poor using proxy means test

Transfer primarily go to women

Food/nutrition and education transfers conditioned on health and education behavior

Oportunidades: Data and method

Random assignment of eligible communities to treatment and control (320 treatment 186 control)

9936 households – 6281 treatment, 3655 control Baseline: 1997 census and March 1998 survey Follow-up: Oct 1998, March 1999 and Nov 1999

Questionnaire limited in production questions Own food production, land use, livestock, agricultural

spending Different agricultural seasons

Consumption from own production

Control mean Oct 1998

Control mean

May 1999

HH consumes from own production

0.70 0.04**(0.035)

0.68 0.04**(0.023)

Per capita value of own production

25.11 3.00***(0.008)

23.61 0.97(0.517)

Number of food groups from own production

1.16 0.12***(0.004)

1.13 0.12***(0.011)

Number of foods from own production

1.26 0.14***(0.003)

1.17 0.14***(0.004)

P-values in parenthesis

Cons. own production: Food groupsControl mean Oct 1998

Control mean

May 1999

Cereals0.52 0.02

(0.278)0.50 0.03

(0.124)

Beans0.06 0.02*

(0.080)0.06 0.00

(0.895)

Fruit 0.24 0.05**(0.024)

0.16 0.21(0.282)

Vegetables0.12 0.00

(0.964)0.23 0.37**

(0.031)

Meat 0.07 0.02**(0.012)

0.05 0.01(0.113)

Eggs0.10 0.01

(0.455)0.11 0.01

(0.536)

P-values in parenthesis

Agricultural investment and spendingControl mean Oct 1998

Control mean

May 1999

Use land0.59 0.34**

(0.048)0.55 0.23

(0.203)

Per capita hectares of land use

0.26 0.32*(0.094)

0.24 0.01(0.378)

Own livestock 0.77 0.03*(0.060)

0.72 0.03**(0.037)

Per capita livestock ownership

0.14 0.02*(0.099)

0.14 0.03***(0.007)

Agricultural spending 0.43 0.05**(0.013)

0.34 0.02(0.321)

Per capita agricultural spending

210 21.1(0.478)

181 30.1(0.403)

P-values in parenthesis

Effects by land holdings

Consumption from home production Landless & 0-3 hectares: similar to general effects >3 hectares: no effects

Cons. from own productions by food groups Landless & 0-3 hectares: similar to general effects >3 hectares: effects on fruit in Oct

Agricultural investment and spending Landless: Only land use effects and livestock on Oct 0-3 hectares: Land area effects, strong livestock effects

(Oct and May) and agricultural spending in Oct >3 hectares: no effects

PROCAMPO and Oportunidades

PROCAMPO Decoupled cash transfer provided to compensate (1993-

1994) staple producers given expected impact of NAFTA 34% receive PROCAMPO—22% receive both Highly agriculturally oriented

Impact of Oportunidades on PROCAMPO producers Impact on all households not just PROCAMPO But significantly larger impact of Oportunidades for

PROCAMPO recipients on own food consumption (cereals), land use, livestock ownership, agricultural spending

Malawi Social Cash Transfers

Initiated in 2006 with intention to reach poorest 10% of population Expanding with hope of reaching 300,000 households

by 2015

Targets ultra poor, labor constrained households—geographic combined with community targeting within Village Development Groups (VDGs)

Unconditional cash transfer but “encouraged” to invest in children (soft conditions)

Malawi SCT: Data and method

Evaluation 2007-2008

Random assignment of eight VDGs into treatment and control in Mchinji district

751 households – 386 treatment, 365 control Baseline: March 2007 Follow-up: Sept 2007 and April 2008

Questionnaire limited in production questions Agricultural assets, some time use, own production

Agricultural assets (preliminary)

Total Labor>0 Labor=0Male head

Female head

Hoes0.13***(0.000)

0.10**(0.044)

0.16***(0.007)

0.02(0.755)

0.19***(0.000)

Sickles0.29***(0.000)

0.23***(0.000)

0.34***(0.000)

0.19***(0.000)

0.34***(0.000)

Goats 0.48***(0.000)

0.58***(0.000)

0.41***(0.000)

0.43***(0.000)

0.52***(0.000)

Chicken 0.59***(0.000)

0.70***(0.000)

0.50***(0.000)

0.58***(0.000)

0.60***(0.000)

P-values in parenthesis

Time use (preliminary)

Sept 07 April 08

Own farm-0.01

(0.400)0.12***(0.000)

Casual labor-0.33***(0.000)

-0.44***(0.000)

Casual labor days -2.32***(0.000)

-3.00***(0.000)

Household work 0.14***(0.000)

0.14***(0.000)

P-values in parenthesis

Source of food: Sept / Apr (preliminary) Own

Production Purchase Gift

Cereals + / + + / + - / -

Tubers + / + + / + - / -

Pulses + / + + / + 0 / 0

Vegetables + / + + / + - / -

Animal products + / + + / + 0 / 0

Fruits + / + + / + 0 / 0

Regular maize flour + / 0 + / + - / -

Fine maize flour + / + + / + - / -

Conclusions

SCTs, while linked to long-term human capital investment, have an impact on productive decisions Crop/food diversification, agricultural spending, time

use, land use, investment

Impacts are evident among land poor (Oportunidades) and poor/labor constrained (Malawi) Impact possible linked to credit and insurance market

imperfection

Conclusions

Impacts appear greater for those with agricultural potential (PROCAMPO) Targeting agricultural households may induce greater

productive impacts

Suggests possibility of using productive cash transfers Under what conditions?

Need for further analysis FAO/UNICEF funded by DFID

From Protection to Production

New CTs in SSA accompanied by rigorous impact evaluation

Malawi SCT Mchinji pilot, 2007-2009 Expansion, 2011-2013

Kenya CT-OVC Pilot 2007-2011 Expansion, 2011-2013

Mozambique PSA Expansion, 2008-2009

Zambia Kalombo pilot, 2005 Monze pilot, 2007-2010 Expansion and child grant,

2010-2013South Africa CSG Retrospective and

expansion, 2010-2013

Ethiopia • PNSP, 2006-2010

• Regional minimum social protection package, 2011-2013

Ghana LEAP• Pilot, 2010-2012

Lesotho CGP• Pilot, 2011-2012

Uganda, begins in 2011

Zimbabwe, begins in 2011

Tanzania, TASAF

FAO/UNICEF funded by DFID From Protection to Production

Countries and programs

• Formally:1) Lesotho Child Grant Program (baseline 2011, follow up

2012)

2) Ethiopia Tigray SP package (baseline 2011, follow up 2012)

3) Malawi SCT expansion (baseline 2011, follow up 2012)

4) Kenya CT-OVC (2nd follow up 2011)

5) Ghana LEAP (follow up 2012)

6) Zimbabwe SCT (baseline 2012, follow up 2013)

• Informally: 1) Zambia SCT (baseline 2010, follow up 2012)

2) South Africa CSG (retrospective, 2010)

Strengthen data collection and analysis

• Design, pilot and supervise implementation of additional modules in household surveys– Analyze household economic decision making on

productive activities and labor allocation; climate change adaptation; risk coping; time use and social networks using baseline and follow up data---disaggregated by gender

• Economic “linkages” questions throughout household questionnaire, as well as business enterprise survey– Simulate local economy impacts using village SAM/CGE

models

• Integrate qualitative/quantitative design and methods

• Lead a research network on impact evaluation of CT programs in SSA