Comments to Gertler, Patrinos
and Rubio-Codina (2007)
Emiliana Vegas
The World Bank
October 3, 2007
Mexico has very low levels of
student achievement
Mexico
Brazil
Uruguay
Greece
Portugal
Spain
United States
Polond
Finland
Korea
0% 20% 40% 60% 80% 100%
Below Level 1 Level 1 Level 2 Level 3 Level 4 Level 5 Level 6
Pisa 2003
SBM to raise quality
• SBM is being tried in more and more systems,
including several Latin American countries
• Intuitively appealing – provides the client with
more resources to affect service delivery,
circumventing large, inefficient bureaucracies
• Most impact evaluations have found positive
impacts on student outcomes, though the effects
are not too large (not surprising)
Findings
• GPM find fairly strong, but small, effects of SBM
on failure and repetition, and the effects are
greatest for schools that have received AGE for
more than 1 year
• No observed effects on dropout rates
• Qualitative research strengthens the
interpretations of the findings
• Clear policy implication that resources to parent-
led school organizations can improve school
quality
• But no one policy is a magic bullet
Questions
• Given that most AGE schools are very
small, are there volatility concerns? (See
Chay, McEwan and Urquiola 2005)
– Are some of the observed program effects
possibly due to noise and mean reversion?
– Differences-in-differences can exarcebate
these concerns
• Enrollment decline and dropout. Is there
any relation?