Child Sponsorship and Educational Aspirations

Evidence from Rural Mexico, Summary

Daniel Prudencio

M Csapek SA, Tecnológico de Monterrey
Published in Education Economics, 2025

The Puzzle: Why Doesn’t Education Pay Off Here?

The gap: Returns to education are positive, yet educational attainment stays very low

  • Adults in Oaxaca and Chiapas average only 7.5 and 7.2 years of schooling, barely above primary
  • Both regions among the poorest in Mexico

Standard explanation: external constraints, poverty, distance to schools, poor health

But also: Growing evidence that internal constraints matter too

Research question:

Can an international child sponsorship program shift the level of education children aspire to reach?

Compassion International (CI): Third-largest child sponsorship program worldwide

≈ 2.2 million children across 29 countries

In Mexico: 33, 360 children in >185 centers

What Does CI Do? And How Do We Study It?

Three program pillars:

  1. Material support, school supplies, uniforms, food, healthcare
  2. After-school program, 5–6 hrs/week of academic tutoring and socio-emotional development (self-esteem, self-efficacy, life skills)
  3. Letter exchanges with international sponsors, broadens career horizons and life possibilities

Average 9.3 years of participation

De facto: school attendance required for continued sponsorship

The study:

  • 2017 survey, Oaxaca and Chiapas, Mexico
  • 8 rural communities: 4 with CI, 4 matched controls
  • 403 children aged 12–15 (163 sponsored, 240 non-sponsored)
  • Key outcome: Does the child aspire to a higher education degree?

How Do We Measure CI’s Causal Impact?

The challenge: CI deliberately targets the poorest children, sponsored and non-sponsored kids differ in ways we cannot fully measure

  • Raw comparison of aspirations = not causal

The solution: A binary Roy model (Aakvik et al. 2005)

  • Instrument (\(Z\)): Age when CI arrived in the village, younger children more likely selected, but age-at-arrival should not affect current aspirations

Identifies ATE (average child), ATT (sponsored), MTE (by disadvantage)

Z Age at CI arrival S CI Sponsorship Y Higher-Ed Aspiration U Unobservables ↑ prob. effect? ✗

The Roy model corrects for \(U\) by exploiting variation in \(Z\)

✗ = exclusion restriction: \(Z\) does not directly affect aspirations

Main Result: CI Raises Aspirations

Effect on sponsored children (ATT):

  • Estimated +17 to +20 percentage point increase in the probability of aspiring to higher education
  • Direction consistent with prior CI studies on adult outcomes (Wydick et al. 2013)

Statistical precision:

  • Not statistically significant, small sample (\(N=163\) sponsored) combined with a complex estimation model amplify uncertainty
  • Results are suggestive, not definitive, but the direction is clear and meaningful

Back-of-envelope:

If higher aspirations translate into behavior, a 20 pp increase implies

≈ 8 more months of schooling

Consistent with Wydick et al. (2013): 1.0–1.5 additional years of education for adult CI participants

Even an imprecise result is informative: the program appears to be working in the right direction

Two Additional Insights

Insight 1: CI’s targeting is efficient

The Marginal Treatment Effect (MTE) shows that children most likely to be selected into CI are also those who benefit the most from it

No efficiency-equity trade-off: Aid programs often face a dilemma, targeting the most vulnerable may not reach those with the highest gains. CI avoids this dilemma: the most disadvantaged children extract the greatest benefit

Insight 2: A surprising gender gap

  • Girls aspire more to higher education than boys (boys are ≈ 12 pp less likely to aspire, among non-sponsored)
  • Yet girls expect ≈ 50% lower income than boys at the same education level

The gender gap in aspirations is not explained by income beliefs, structural barriers and identity formation dominate at ages 12–15

Information campaigns about female earnings alone are unlikely to close the gap

Conclusions and What It Means

What we found:

  1. CI has a positive, economically meaningful effect on aspirations for higher education (ATT ≈ 17–20 pp), suggestive, not conclusive
  2. CI’s targeting is efficient: the most vulnerable children benefit the most, no equity-efficiency trade-off
  3. Girls aspire more than boys despite expecting much lower earnings, the gap is not explained by beliefs about future income

What it means:

Holistic programs that combine material support with socio-emotional development are well-positioned to shift aspirations, not just economic constraints

  • Informal attendance requirements may be a valuable design feature, more time in school strengthens educational identity
  • Intervening early (ages 9–12) may be more effective than waiting for aspirations to be fully formed
  • Gender-specific barriers require targeted interventions beyond information provision

References

Aakvik, Arild, James J. Heckman, and Edward J. Vytlacil. 2005. “Treatment Effects for Discrete Outcomes When Responses to Treatment Vary: An Application to Norwegian Vocational Rehabilitation Programs.” Journal of Econometrics 125 (1–2): 15–51.
Cunha, Flavio, and James J. Heckman. 2009. “The Economics and Psychology of Inequality and Human Development.” Journal of the European Economic Association 7 (2–3): 320–64.
Dalton, Patricio S., Sayantan Ghosal, and Anandi Mani. 2016. “Poverty and Aspirations Failure: A Theoretical Framework.” Economic Journal 126 (590): 165–88.
Heckman, James J., and Tim Kautz. 2012. “Hard Evidence on Soft Skills.” Labour Economics 19 (4): 451–64.
Wydick, Bruce, Paul Glewwe, and Laine Rutledge. 2013. “Does International Child Sponsorship Work? A Six-Country Study of Impacts on Adult Life Outcomes.” Journal of Political Economy 121 (2): 393–436.