Barbara M. Fraumeni

1
A Human Capital Approach to Measuring Government
Education Output
Barbara M. Fraumeni Muskie School of Public
Service, University of Southern Maine the
National Bureau of Economic Research,
USA Measurement of Non-market Output and
Health OECD/ONS/Government of Norway Workshop,
Session 5 London, England, UK October 4, 2006
2
Muskie School of Public Service
Ph.D. Program in Public Policy

• BOTTOM LINE
• A human capital approach could significantly
  increase the measured ROG of government education
  output
• It looks beyond a body count to the impact of
  education on lifetime income

3
Muskie School of Public Service
Ph.D. Program in Public Policy

• J-F-C Project Goal
• As a first step, draws upon Jorgenson-Fraumeni
  (1992a, 1992b) methodology
• The current Jorgenson-Fraumeni-Christian (JFC)
  project aims to
• Develop estimates from 1960-2004
• Experiment with a smaller set of data
• Modify and refine the J-F methodology for various
  purposes

4
Muskie School of Public Service
Ph.D. Program in Public Policy

• Context
• The Atkinson Report Consideration of
• an adjustment to reflect the value of education
  for future earnings
• Simplistic view Looking at the real rate of
  growth in wages
• J-F-C view Directly measure the impact of
  education on lifetime earnings

5
Muskie School of Public Service
Ph.D. Program in Public Policy

• Output Targets
• All formal education (J-F)
• Government-provided (public) formal education
  only
• Output from school inputs

6
Muskie School of Public Service
Ph.D. Program in Public Policy

• Variations of J-F
• Literature on relative earnings of publicly vs.
  privately educated individuals?
• Output from school inputs only requires adjusting
  for
• Childrens time
• Parents time helping
• Compositional effects such as non-native speakers

7
Muskie School of Public Service
Ph.D. Program in Public Policy

• Additional Refinements
• Graduate school issues
• Dealing with investment in education for
  elementary school enrollees with a more
  aggregated data set
• Rose-colored glasses?
• Time spent in school
• Years to complete a level (college)
• Hours spent in school (or doing homework)

8
Muskie School of Public Service
Ph.D. Program in Public Policy

• Methodology
• Construction of expected future lifetime income
  (market plus nonmarket) for all individuals in
  the US
• Determined from the cohort alive in a particular
  year
• Investment in education (education output) is
  the difference between expected lifetime income
  of an individual enrolled in school and someone
  of same sex and age who has one year less of
  education

9
Muskie School of Public Service
Ph.D. Program in Public Policy

• Stages of Life
• Lifetime incomes are calculated by a backwards
  recursion, starting with age 75
• Stage 5 Age 75, retirement
• Stage 4 Age 35-74, work only
• Stage 3 Age 16-34, work and school
• Stage 2 Age 5-15, school only
• Stage 1 Age 0-4, no work or school

10
Muskie School of Public Service
Ph.D. Program in Public Policy

• So What?

Growth Rates Enrollments Investment in Education (constant)
1960-1986 (J-F) .6 2.7
1986-2004 (J-F-C) 1.2 ?
11
Muskie School of Public Service
Ph.D. Program in Public Policy
1960-1986 vs. 1986-2004 College The Critical
Factor
Growth Rates Enrollment Growth Rates Wage Gap College to Other Levels
1960-1986 J-F 4.6 (.6) Decreases often, but also increases
1986-2004 J-F-C 1.8 (1.2) Increases substantially
12
Muskie School of Public Service
Ph.D. Program in Public Policy

• 1960-1986 vs. 1986-2004
• College The Critical Factor
• 1960-1986 Enrollments are the driver
• 1986-2004 The wage gap as reflected in lifetime
  income will be the driver
• Actual Estimates J-F-C Project