SUSAN W. PARKER, MARYLAND
TOM VOGL, TEXAS
T R A N S F E R P R O J E C T W O R K S H O P
A P R I L 2 - 4 2 0 1 9
A R U S H A , T A N Z A N I A
Do Conditional Cash Transfers Improve
Economic Outcomes in the Next
Generation? Evidence from Mexico
Motivation
Conditional cash transfer (CCT) programs have become a central strategy for
poverty alleviation in over 80 countries over past two decades.
Began in Latin America, spread to Africa and Asia, now in US and Europe.
Innovation: condition transfers to the poor on investment in human capital.
Dual aims: reduce current poverty and poverty of the next generation.
Does CCT exposure in childhood improve economic outcomes in adulthood?
Unclear: returns to human capital investment in poor areas may be low.
I School quality low, labor markets mostly agricultural.
Short-term studies: positive schooling effects, mixed evidence on test scores.
Benefits to children necessary for justifying conditionality.
This paper estimates long-term effects in a nationwide program.
Outcomes: schooling, labor market performance, living standards, migration.
Setup
The Progresa program
Methods: Data and Design
Research design:
Implementation phases: 1997-1999 (Zedillo versus 2001-2005 (Fox)
Exposure classifications: age 7-11 vs. 15-19 in 1997
Data:
10% sample of the 2010 Mexican Population Census
Merged to administrative government data:
Yearly Progresa enrollment at municipality (county) level
“High” and “very highly” marginalized municipalities where program began.
Pre Program Education and Labor Market Outcomes
Men Women
(1) (2) (4) (5)
A. Working
Mean (SD) 0.81 (0.40) 0.26 (0.44)
B. Working for wage
Mean (SD) 0.48 (0.50) 0.17 (0.36)
C. Working in agriculture
Mean (SD) 0.39 (0.49) 0.03 (0.17)
D. Health insurance from job
Mean (SD) 0.14 (0.35) 0.13 (0.34)
E. Weekly labor hours
Mean (SD) 36.18 (24.24) 10.30 (20.54)
F. Years of completed schooling
Mean (SD) 7.9 (3.96) 7.65 (4.07)
Table 7: Evaluation of Program Benefits and Costs
No ann. earnings growth 2% ann. earnings growth
DWL = 0.2 DWL = 0.6 DWL = 0.2 DWL = 0.6
(1) (2) (3) (4)
Lower bound on benefits, 2010 pesos 47,037 47,037 81,915 81,915
Costs, 2010 pesos 7,456 16,985 7,456 16,985
Lower bound on net benefits as % of 2010 GDP p.c. 38% 29% 71% 62%
Lower bound on B/C Ratio 6.31 2.77 10.99 4.82
Note: Costs include administrative and private costs = 0.113 per peso of transfers (Coady, 2000), plus the
deadweight loss of taxation to finance direct costs and transfers. The discount rate is assumed to be 0.05.
Conclusions and some lessons learned
Lasting program impacts on education, labor market outcomes,
household economy, mobility, especially for women.
Pre program context of low LFP and low women’s status.
Migrating out of rural villages may be key to long term benefits.
Results encouraging for long-term trajectories of children from
households receiving CCTs.
Note: next generation effects only visible about 15 years
post program
→Implies need for longer term evaluations.
On estimating long run impacts
Combining Census data with administrative data on beneficiaries
(geographic or individual level) can work as general estimation
strategy.
May provide more representative impacts for large
scale programs than small scale RCTs.
Census should include birth place or early residence and
some outcome variables e.g. income.
Complementary strategy to long term follow up of RCT
sample-Progresa case-some limitations due to short
experimental period.

Do Conditional Cash Transfers Improve Economic Outcomes in the Next Generation? Evidence from Mexico

  • 1.
    SUSAN W. PARKER,MARYLAND TOM VOGL, TEXAS T R A N S F E R P R O J E C T W O R K S H O P A P R I L 2 - 4 2 0 1 9 A R U S H A , T A N Z A N I A Do Conditional Cash Transfers Improve Economic Outcomes in the Next Generation? Evidence from Mexico
  • 2.
    Motivation Conditional cash transfer(CCT) programs have become a central strategy for poverty alleviation in over 80 countries over past two decades. Began in Latin America, spread to Africa and Asia, now in US and Europe. Innovation: condition transfers to the poor on investment in human capital. Dual aims: reduce current poverty and poverty of the next generation. Does CCT exposure in childhood improve economic outcomes in adulthood? Unclear: returns to human capital investment in poor areas may be low. I School quality low, labor markets mostly agricultural. Short-term studies: positive schooling effects, mixed evidence on test scores. Benefits to children necessary for justifying conditionality. This paper estimates long-term effects in a nationwide program. Outcomes: schooling, labor market performance, living standards, migration.
  • 3.
  • 5.
  • 8.
    Methods: Data andDesign Research design: Implementation phases: 1997-1999 (Zedillo versus 2001-2005 (Fox) Exposure classifications: age 7-11 vs. 15-19 in 1997 Data: 10% sample of the 2010 Mexican Population Census Merged to administrative government data: Yearly Progresa enrollment at municipality (county) level “High” and “very highly” marginalized municipalities where program began.
  • 9.
    Pre Program Educationand Labor Market Outcomes Men Women (1) (2) (4) (5) A. Working Mean (SD) 0.81 (0.40) 0.26 (0.44) B. Working for wage Mean (SD) 0.48 (0.50) 0.17 (0.36) C. Working in agriculture Mean (SD) 0.39 (0.49) 0.03 (0.17) D. Health insurance from job Mean (SD) 0.14 (0.35) 0.13 (0.34) E. Weekly labor hours Mean (SD) 36.18 (24.24) 10.30 (20.54) F. Years of completed schooling Mean (SD) 7.9 (3.96) 7.65 (4.07)
  • 11.
    Table 7: Evaluationof Program Benefits and Costs No ann. earnings growth 2% ann. earnings growth DWL = 0.2 DWL = 0.6 DWL = 0.2 DWL = 0.6 (1) (2) (3) (4) Lower bound on benefits, 2010 pesos 47,037 47,037 81,915 81,915 Costs, 2010 pesos 7,456 16,985 7,456 16,985 Lower bound on net benefits as % of 2010 GDP p.c. 38% 29% 71% 62% Lower bound on B/C Ratio 6.31 2.77 10.99 4.82 Note: Costs include administrative and private costs = 0.113 per peso of transfers (Coady, 2000), plus the deadweight loss of taxation to finance direct costs and transfers. The discount rate is assumed to be 0.05.
  • 12.
    Conclusions and somelessons learned Lasting program impacts on education, labor market outcomes, household economy, mobility, especially for women. Pre program context of low LFP and low women’s status. Migrating out of rural villages may be key to long term benefits. Results encouraging for long-term trajectories of children from households receiving CCTs. Note: next generation effects only visible about 15 years post program →Implies need for longer term evaluations.
  • 13.
    On estimating longrun impacts Combining Census data with administrative data on beneficiaries (geographic or individual level) can work as general estimation strategy. May provide more representative impacts for large scale programs than small scale RCTs. Census should include birth place or early residence and some outcome variables e.g. income. Complementary strategy to long term follow up of RCT sample-Progresa case-some limitations due to short experimental period.