Returns to Education The Gender Perspective

1
Returns to EducationThe Gender Perspective
Harry Anthony Patrinos World Bank
Global Symposium Education A Critical Path to
Gender Equality and Empowerment, October 2-3,
2007 World Bank, Washington DC
2
Human Capital Not a New Idea

• The most valuable of all capital is that invested
  in human beings
• Alfred Marshall
• 1890

3
Returns to Schooling

• Returns to investment in education based on human
  capital theory
• Estimated since 1950s
• Several reviews of empirical results in attempts
  to establish patterns
• More estimates from wide variety of countries,
  over time, new econometric techniques
• Reaffirm importance of human capital theory
• Education as investment

4
(No Transcript)
5
ln Wi a ßSi ?1Xi ?2Xi2 ei
6
Returns to Year of Schooling ()
Source Psacharopoulos and Patrinos 2004
7
Higher Returns in Developing Countries
8
Highest Returns to Primary
27
19
17
10
Primary
Secondary
Higher
9
(No Transcript)
10
Rising Higher Education Returns
points
Primary
Higher
11
Returns to Additional Year of Schooling
12
Private Returns

• Undisputable
• Universal, global
• Explaining behavior
• Analyzing distribution effects
• Example Student loans
• But not sufficient for funding policies

13
Social Benefits

• Beyond private benefits of increased
  productivity
• Private non-market effects
• Externalities spillovers innovation community

14
Social Returns to Education

• Effect of others education on wages
• (Acemoglu Angrist 1999, Moretti 2002)
• Average schooling at the state level is highly
  correlated with average wage

15
Non-Market/External Benefits
Wolfe and Zuvekas (1997)
16
Intergenerational Benefits

• Parents with more schooling spend time with
  children effectively
• Parents with more schooling better at assessing
  returns to schooling
• Parents with more schooling add more to household
• Parents with more schooling serve as role models

17
Returns to Schooling by Gender
18
Benefits of Girls Education

• Year of schooling for girls reduces infant
  mortality 5-10 (Schultz 1993)
• Children of mothers with 5 years primary 40 more
  likely to live beyond age 5 (Summers 1994)
• Double proportion of women with secondary reduces
  fertility rate 5.3 to 3.9 per woman (Subbarao
  Raney 1995)
• Providing girls with extra year of education
  wages by 10-20 (Psacharopoulos Patrinos 2002)
• More productive farming due to increase female
  schooling, 43 of decline in malnutrition (Smith
  Haddad 1999)
• Womens education has more impact than mens on
  childrens schooling (Filmer 2000)
• Brazil womens resources have 20X impact on
  child health compared with mens (Thomas 1990)
• Young rural Ugandans with secondary 3X less
  likely to be HIV (De Walque 2004)
• India women with formal schooling more likely to
  resist violence (Sen 1999)
• Bangladesh educated women 3X likely to
  participate in political meetings (UNESCO 2000)
• (Herz and Sperling 2004 Watson 2005)

19
Schooling Gap
Source Barro Lee
20
Gender Differences in Returns to Schooling

• 95 estimates of male female schooling returns
• Coefficients from 49 countries
• 63 are greater for females (66)
• 3 equal
• 23 greater for males

21
Returns to Schooling by GenderSelected Countries
22
Returns to Schooling by Level
23
Gender Differences

• Rate of return to schooling for men is highest
  for primary education
• decreases for secondary education
• increases for university education
• For women, returns lowest for primary education
• highest for secondary education
• Decreases for university education

24
(No Transcript)
25
A Major Policy Concern

• Families may decide to send sons rather than
  daughters to school
• Jeopardizes MDGs, at least 2
• Lower returns to females at primary level is
  puzzling opposite situation in transition
  countries

26
Explanations

• Detrimental impact of discrimination and other
  factors
• cause women to accept wage offers that undervalue
  their characteristics
• Better educated woman more able willing to
  overcome sex handicaps compete with men in
  labor market
• Male-female differences in quality of schooling
• (Dougherty 2005)

27
(No Transcript)
28
Chile, IALS

• Working women score higher on standardized test
• 219 gt 209
• Women have more schooling 10.5 gt 9.3
• But earn significantly less

29
Chile, IALS

• Working women score higher on standardized test
• 219 gt 209
• Women have more schooling 10.5 gt 9.3
• But earn significantly less
• Receive lower returns to Year of Schooling

30
Chile, IALS

• Working women score higher on standardized test
• 219 gt 209
• Women have more schooling 10.5 gt 9.3
• But earn significantly less
• Receive lower returns to Year of Schooling
• And receive lower returns to Cognitive Ability

31
Explanations

• Women may choose to work in sectors where
  education is relatively highly valued
• Lower return to primary may be evidence of
  discrimination
• such that women need to achieve more schooling in
  order to earn sufficient wages
• or option value of schooling since women are
  likely to perform better, stay in school longer,
  experience higher returns for each year of
  schooling, then the most able go beyond primary
  schooling, thus depressing the returns at that
  level
• (Dougherty 2005)

32
Gender Differences

• Possible that differential returns due to
  different job opportunities available to men and
  women at each schooling level
• In Thailand, manufacturing jobs that pay more
  require secondary have grown more rapidly for
  women (Hawley 2004 Phananiramai 1996 Kurian
  1999)
• Jobs that require only primary (eg,
  construction), proportion of men may be higher
  because of tradition or labor intensity, so men
  have higher returns to primary

33
Conclusions

• Overall, returns high justify public private
  investment
• Differential returns signal problems in labor
  market
• and need for further research
• including program evaluations
• Lower returns to primary for women in developing
  countries a major policy concern
• Jeopardizes MDGs, at least 2, limits overall
  benefits of womens schooling
• Quality necessary, but not sufficient
• Enrollment incentives may be needed