Effects of school inputs on academic performance

1
Effects of school inputs on academic performance

• Hanushek (1986, 1996)
• Card and Krueger (1996)

2
The effects of school inputs on students
academic outcome

• Students outcome variables
• Test scores (Hanushek)
• Post- graduation earnings (Card and Krueger)
• School input e.g.
• computer use
• lower class size
• Teacher pay
• Controversies
• no strong relation between school expenditures
  and student performance
• The Colman report (1966), Hanushek (1986)
• Strong positive results
• Card and Krueger (1996)

3
Basic framework

• Education production function
• Control for family inputs
• Control school inputs for the school attended
• Hanushek see presentation on Class Size
• Earnings determination
• Typically didnt control for family inputs
• Typically didnt control for the resources
  relevant to the student (rather, they use
  aggregate school data e.g. state-level)
• Card and Krueger

4
Sources of conflicting results

• Omitted variables
• Student ability, family background, state-level
  political variables, school-specific effect

5

• Omitted variables lead to a spurious relation
  between school inputs and student performance
  that can go either directions
• Positive spurious relation Kids from wealthier
  families often attend schools with better
  teachers and smaller class size. If we cannot
  fully control for family background, we will see
  a spurious positive association between school
  resources and measured outcomes, even if school
  inputs have no effect per se.

6

• Negative spurious relation students with weaker
  background may be assigned to remedial classes
  with higher resources per student

7

• Card and Krueger use aggregate measures for
  school inputs in their earnings determination
  equations
• This generates omitting variables problems
  (omitting school-specific effects relating to
  students academic or labor market performance)
• See Hanusheks (1996) comments on the use of
  aggregate measures of school inputs