File2-page144 - Duflo (2001) 7.1 Human Capital Duflo...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Duflo (2001) 7.1 Human Capital Duflo (2001) Schooling and Labor Market Consequences of School Construction in Indonesia 1973-1978: Indonesian government constructed over 61000 primary schools Use this policy experiment to look at impact on education and wages Identification strategy: (i) only some cohorts affected (ii) intensity of program in inverse proportion to level of enrollment – education of men should be higher who were young enough when the program was launched should he higher than education of older men in all regions – difference should be larger in the regions that received more schools Difference in difference estimator – controls for systematic variation of education both across regions and across cohorts – allows Duflo to isolate program effects Only combination of the two variables is treated as exogenous Specifically – exogenous variables (and the instruments in the wage equation) are the interactions between dummy variables indicating the age of the individual in 1974 and the intensity of the program between 1973 and 1978 – See Table 3 Clear effect on education – effect of one school built per 1000 children is to increase eduction for children aged 2 to 6 in 1974 by 0.12 years in the whole sample and by 0.19 years in the sample of wage earners Control experiment – comparing 12-17 and 18-24 cohorts – no impact Clear effect on wages – increase in completed years of education has translated into an increase in wages of up to 3.8% for each additional school built Estimates of exogenous returns to education using this exogenous variaDevelopment Economics, LSE Summer School 2007 141 ...
View Full Document

Ask a homework question - tutors are online