Education, Productivity and WellBeing

1


Measuring Well-Being and Societal Progress Milan,
JRC/OECD Workshop Series

• Education, Productivity and Well-Being
• Prof. Albert Tuijnman
• EIB Projects Directorate

6
2
PURPOSETo comment on the complex
relationships between initial education, lifelong
learning for adults, individual economic returns
and self-perceived well-beingSTRUCTURE1.
Measurement issues2. Findings of a Swedish
longitudinal study3. Policy implications4.
Education and productivity revisited5. Broad
policy implications
3
Recurrent Education Well-Being

• Ph.D. thesis published in 1989  Recurrent
  Education, Earnings and Well-Being A
  Longitudinal Analysis of a Cohort of Swedish
  Men  Acta Universitatis Stockholmiensis.
  Stockholm Almqvist Wiksell.
• Aims of the study
• 1. Measurement of cumulative learning over the
  life-span
• 2. Measurement of life-time earnings (until age
  55)
• 3. Measurement of self-perceived well-being at
  age 56
• 4. Establishing individual economic returns to
  recurrent
• education at age 56
• 5. Establishing causal network between
  education, earnings
• and elf-perceived quality of life at age 56

4
LATENT INDEPENDENT VARIABLES

• Study used a longitudinal data set spanning
  almost 50 years. A structural equation model was
  developed, with the following latent independent
  variables
• 1. Home Background - Father education Age 10
  (1938)
• - Mother education
• - Father occupation
• - Mother occupation
• - Family income
• 2. IQ Standardised Score - Sub-tests A-D Age 10
• 3. Youth Education - Years of schooling Ages
  15-25
• - Highest obtained level
• 4. IQ Standardised Score - Sub-tests A-D Ages
  20-25
• 5. Occupational Status Treiman Prestige
  Index Ages 26-55

5
MEDIATING AND DEPENDENT VARIABLES

• Mediating Variables
• 6. Adult Education Cumulative retrospective
  measures Ages 26-55
• Four indices
• Independent Variables
• 7. Income Gross declared income from work Ages
  26-55
• Annual measure, supplied by the tax
• authorities
• 8. Self-perceived Index based on 12 quality of
  life Age 56
• Well-being (QoL) items

6
KEY FINDINGS

• 1. Net wage effect of initial education
  controlling for HB and IQ1-2 peaked at around
  ages 38-44 without the measure of cumulative
  adult education in the model and at ages 28-32
  with adult education specified.
• 2. Net wage effect of initial education
  increasingly mediated by cumulative adult
  education and training.
• 3. Highest obtained level of education
  (attainment) much more robust predictor of
  life-time earnings than years of schooling
• 4. No significant effect of initial ecucation on
  QoL at age 56 Weak effects of OCC on QoL
  inconsistent and weak effect of earnings on QoL
  Weak but significant effect of recurrent
  education on QoL, but causation unclear.

7
THEORETICAL AND POLICYRAMIFICATIONS

• Theory
• Empirical support for screening theory
• Empirical support for human capital theory
• Empirical support for recurrent education theory
  (subsequently relabeled as lifelong learning
  theory)
• Policy
• OECD education indicators (INES, IALS, EAG, EPA)
• OECD support for quantitative education expansion
• OECD support for lifelong learning policies
• OECD focus on knowledge and basic skills (PISA)

8
EDUCATION PRODUCTIVITY

• Prof. Alison Wolfs (provocative) Statement
• Productivity effects of education are exaggerated
  
• 1. Microeconomic evidence Private retuns
  (8-12)
• Social returns (6-8)
• 2. Macroeconomic evidence Labour productivity
  (3)
• Growth effects (ideas, innovation)
• One additional year of schooling raises PR/SR by
  6-8
• One additional year of schooling raised total
  factor productivity by almost the same amount
  (6-8)

9

• The evidence on the micro- and macro-economic
  returns to increased years schooling is rather
  consistent both across countries and over time.
  But how valid and reliable is the evidence?
• Are productivity effects of schooling
  overestimated?
• YES, for two main reasons
• VALIDITY - ACLEV ? Years of schooling
• - No controls for HB and IQ
• - Education ? Skills
• - Crucial omitted variable Lifelong learning
  for adults

10

• RELIABILITY - Cross section data sets, most
  often from 1960s to mid-1990s do not reflect
  the recent quantitative
• expansion of education systems
• - Cumulative lifelong learning for adults
  is known to
• mediate the diminishing effects of schooling
  over
• the life-span
• - Some of the schooling effect is direct and
  some is
• indirect (externalities)
• - Education impacts on other variables
  relevant for
• productivity and economic growth
• 1. Education begets education
• 2. Intergenerational effects of education
• 3. Health and life expectancy
• 4. Labour force participation
• 5. Political awareness, civic engagement
• 6. Effect on the productivity of other persons

11
IN SUMMARY

• Qualified support for Professor Wolfs assertion
  that the productivity effects of initial
  schooling are over-estimated
• Support, at least partially, for Screening Theory
• - But also support for Human Capital Theory, but
  only in the perspective of recurrent education
  and lifelong learning.

12
CONCLUSIONS FOR POLICY

• Past returns are not necessarily a good guide to
  future returns
• Expansion of initial schooling (years) cannot
  continue indefinitely, at some point there will
  be diminishing, even negative returns
• Initial formal education Focus should shift
  from quantitative expansion to more quality
  enhancement with a focus on productivity-relevant
  competencies and skills
• LLL will remain important and needs to be given
  strong continued support. But difficult maintain
  high policy interest over time. Thus important to
  occasionally reinvent the concept
• Strengthen the indirect productivity effects of
  initial education through complementary actions
  in innovation, health and civic engagement
  (social capital)

13
http//www.eib.orgmoney doesnt make you happy