Schooling and Labor Market Consequences of School Construction

1
Schooling and Labor Market Consequences of School
Construction

• Paper by Esther Duflo
• Presentation by Radha Iyengar

2
Motivation

• Estimated returns to education are thought to be
  larger in developing countries than in
  industrialized countries.
• Family and community background are important
  determinants of both schooling and labor market
  outcomes

3
Questions to be answered

• whether investments in infrastructure can cause
  an increase in educational attainment
• whether an increase in educational attainment
  causes an increase in earnings

4
The Natural Experiment

• Look at the effect building schools has on
  education and earnings in Indonesia
• Use a major school launched by the Indonesian
  government launched (Sekolah Dasar INPRES)
  program
• Between 19731974 and 19781979, more than 61,000
  primary schools were constructed an average of
  two schools per 1,000 children aged 5 to 14 in
  1971.
• Enrollment rates among children aged 7 to 12
  increased from 69 percent in 1973 to 83 percent
  by 1978.

5
Identifying Assumption

• Treatment Intensity of school construction in a
  geographic area (high vs. low)
• Treatment Group Compare younger cohort (2-6) who
  were in school the whole time
• Control Group Compare older cohort (12-17) who
  were done with school by time of constructions
• Worry that this is all mean reversion so have a
  second falsification check comparing the older
  cohort (12-17) to and even older cohort (18-24)

6
Wald Test is the Ratio of these two (Change in
Earnings/ Change in Schooling)
7
(No Transcript)
8
Bottom Line

• A school construction program takes a very long
  time to generate positive returns (because the
  costs are incurred early on, whereas the
  benefits are spread over a generation).
• The returns generated are large. The internal
  rates of return range from 8.8 to 12 percent,
  well above the average interest rate on
  government debt in Indonesia during the period.

9
Major Issues 1

• Causal interpretation requires
• No omitted time-varying and region-specific
  effects correlated with the program.
• No changes in trends for different cohorts over
  time
• Some support from Control Experiment but
• Control Experiment wont be valid if both
  region-specific and cohort specific trends
• Unclear what the true counterfactual trend is

10
Cohort-Specific Effects

• Large changes in growth also varied by regionbut
  its not clear what the overlap is here
• Given this, the Returns to Education are not
  stable, so that the investment decision is not
  the same
• May not be that school construction helped
• May be that expected future returns changed
  differentially by cohorts by region and so
  investment increased
• This wouldnt happen in older cohorts because no
  large changes in growth

11
Major Issue 2

• The benefits are, to a large extent, driven by
  the rapid growth of Indonesias GDP from 1973 to
  1997
• As Duflo Notes
• If the growth rate had been very low from 1973
  until today, the net present value of the program
  would actually have been slightly negative,
  according to all specifications but one.

12
External Validity

• Given Indonesias expected increase in
  growthinvestment in school is a good idea BUT
  how much growth is needed is sensitive to
  specification
• May not be appropriate to extend to other
  populations