Transcript Meeting the Challenge of Secondary Education in East

Meeting the Challenges of Secondary
Education in East Asian and Latin
American Countries
Presentation for Kuala Lumpur Secondary Education
Conference – September 19-21, 2005 (Emanuela di
Gropello)
Objectives of the Study


1) Examine the key challenges in the
delivery of secondary education in EAP
and LAC;
2) Suggest a broad range of innovative
strategies to expand and improve
secondary education, with focus on:


a) Resource mobilization; and
b) Efficiency of service delivery.
Outline of the Report
1.
2.
3.
4.
5.
Chapter 1: Diagnostic of
Secondary Education in LAC and
EAP countries
Chapter 2: Understanding
Demand-side Constraints
Chapter 3: Improving Governance
and Management Structures
Chapter 4: Mobilizing Resources
Chapter 5: Improving the
Efficiency of Service Delivery
Diagnostic of Secondary Education in
EAP and LAC countries – Chapter 1

1.
2.
The chapter:
describes the stock of and flow of the
quantity and quality of education in East
Asia and Latin America countries; and
analyzes demand and supply-side
constraints and opportunities (laying the
ground for further analysis in chapters 2
and 3).
Structure of Secondary Education

-
Secondary education varies considerably across
the countries studied. These differences include:
the age of entrance into secondary education
(lower in LAC);
the duration of compulsory schooling (longer in
LAC and upper-middle income countries);
the length of study (longer in EAP, with an
average of 3.18 years in lower secondary and 3
years in upper secondary).
Secondary Enrollment in EAP and LAC


Latin American and East Asian countries
have rapidly increased primary school
enrollment rates to achieve near universal
coverage. Average primary NER in LAC:
94%; EAP: 92%.
But secondary enrollment levels in many
Latin American and East Asian countries
are below typical levels given national
income.
Secondary Enrollment Gap
180
Gross secondary enrollment rate, 2002 or most recent year
160
Belgium
Australia
Sweden
140
Denmark
Finland
Netherlands
120
New Zealand
Spain
Dominica
Norway
Iceland
France
Slovenia
Barbados
Ireland
Hungary
Japan
Uruguay
Germany
Tonga
Argentina
Italy Switzerland
Estonia
Greece
Uzbekistan
Latvia
Israel
United States
Guyana
Russian Federation
Kyrgyz Republic
Korea, Rep.
Peru
Chile
Macao, China
St. Lucia
Jordan
Bolivia
Mongolia
Jamaica
Thailand
Trinidad and Tobago
Philippines
Fiji
Georgia
Oman
Belize
Hong Kong, China
Mexico
Colombia
Moldova
Vietnam
Panama
Malaysia
Venezuela, RB
China
Saudi Arabia
Paraguay
Nicaragua
Nepal
Ecuador
Dominican Republic
El Salvador
Costa Rica
Indonesia
Brazil
100
Tajikistan
80
60
Yemen, Rep.
40
Malawi
Zambia Benin
Sierra Leone
20
Ethiopia
Burundi
Gabon
India
Bangladesh
Lao PDR
Zimbabwe
Kenya
Luxembourg
Morocco
Ghana
Gambia, TheLesotho
Cameroon
Vanuatu
Cambodia
Guatemala
Equatorial Guinea
Mali
Uganda
Chad
Burkina Faso
Niger
0
6
7
8
9
Log of per capita GDP, 2000 (PPP$)
10
11
12
Lower Enrollment in Secondary
2002 Enrollment in Thailand, by age and sex
Female
Male
19
18
17
16
15
14
13
12
11
10
9
8
7
100 90 80 70 60 50 40 30 20 10
Never enrolled
Never enrolled
Dropout/Finished
Dropout/Finished
0
Tertiary
Tertiary
10 20 30 40 50 60 70 80 90 100
Male
Secondary
Secondary
2003 Enrollment in Colombia, by age and sex
Female
19
18 Primary
Primary
17
16
15
14
13
12
11
10
9
8
7
100 90 80 70 60 50 40 30 20 10
Never Enrolled
Never Enrolled
Dropout / Finished
Dropout / Finished
0
Tertiary
Tertiary
10 20 30 40 50 60 70 80 90 100
Secondary
Secondary
Primary
Primary
9%
Change in GER2
90%
8%
80%
NER2
60%
50%
40%
10%
5%
4%
GER2
20%
6%
3%
NER2
30%
GER2
70%
NER2
7%
GER2
Gross and Net Enrollment Rates
100%
2%
Change in NER 2
1%
0%
0%
Low income
Lower middle income
Upper middle income
Annual Change in enrollment rates
Differences among Income Groups
Internal Efficiency in EAP and LAC:
Over-age is an Issue




Over-age enrollment is a problem in all
countries, but more so in LAC than EAP,
as indicated by the GER/NER gap:
LAC: GER = 80%; NER = 62%;
EAP: GER = 70%; NER= 58%.
Over-age enrollment is related to income
levels, more so in EAP than LAC.
Over-age is Pervasive in LAC
2000 Enrollment in Guatemala by age and sex
Male
Female
19
18
17
16
15
14
13
12
11
10
9
8
7
100 90
80
70
60
50
40
30
20
10
0
10
20
30
40
50
60
70
80
2001 Enrollment in Brazil, by age and sex
90 100
Female
Male
Neverenrolled
Neverenrolled
Dropout/Finished
Dropout/Finished
Tertiary
Tertiary
Secondary
Secondary
Primary
Primary
19
18
17
16
15
14
13
12
11
10
9
8
7
100 90
80
70
60
Never Enrolled
Never Enrolled
50
40
30
20
Dropout / Finished
Dropout / Finished
10
0
Tertiary
Tertiary
10
20
30
40
50
Secondary
Secondary
60
70
80
Primary
Primary
90 100
Over-age Lower at Higher Income
Levels in EAP
1998 Enrollment in Philippines, by age and sex
Male
Female
19
18
17
16
15
14
13
12
11
10
9
8
7
100 90
80
70
60
Never Enrolled
Never Enrolled
50
40
30
Dropout/Finish
Dropout/Finish
20
10
0
Tertiary
Tertiary
10
20
30
40
Secondary
Secondary
50
60
70
80
90 100
2004 Enrollment in Korea, by age and sex
Primary
Primary
Female
Male
19
18
17
16
15
14
13
12
11
10
9
8
7
100 90 80 70 60 50 40 30 20 10
not enrolled
not enrolled
0
tertiary
tertiary
10 20 30 40 50 60 70 80 90 100
secondary
secondary
primary
primary
Internal Efficiency in EAP and LAC:
Low Completion Rates





Survival plots show that the likelihood of
completing secondary varies substantially
across countries, but is generally low,
indicating further inefficiencies:
Cambodia: 10%;
Guatemala: 25%;
Indonesia, the Philippines, Vietnam,
Mexico, Brazil: about 50%;
Thailand and Colombia: about 60%.
0.75
mexico
0.50
0.50
0.00
0.00
0.25
0.25
% completed
0.75
cambodia
0
1
2
3
4
5
6
7
grade
8
9
10
11
12
0
5
10
15
analysis time
sex = Female
hombre = 0
0.50
0.50
0.75
0.75
hombre = 1
Kaplan-Meier survival estimates, by male
1.00
1.00
sex = Male
0.25
brazil
0.00
0.25
vietnam
0.00
% completed
Kaplan-Meier survival estimates, by hombre
1.00
1.00
Low Survival Probabilities
0
1
2
3
4
5
m1c2 = male
6
grade
7
8
9
10
11
12
0
5
10
analysis time
m1c2 = female
male = 0
male = 1
15
PISA 2000 Reading Results
100%
600
80%
500
400
60%
300
200
40%
20%
100
0
Latin America
Japan
Hong
Kong
Korea
Thailand
Indonesia
Mexico
Chile
Argentina
Brazil
0%
Peru
Percentage of Students in Each
Proficiency Level
Quality of Secondary Education: Varies across Regions and
Income Group, Higher Percentage in Lower Proficiency Levels
in Middle and Lower-Middle Income Countries
East Asia
OECD
Avg
Level 5
Level 4
Level 3
Level 2
Level 1
B elo w Level 1
M ean
Quality of Secondary Education: Varies across Regions and
Income Group, Higher Percentage in Lower Proficiency Levels
in Middle and Lower-Middle Income Countries
100%
600
80%
500
Level 5
400
Level 4
60%
300
40%
20%
0%
East Asia
MacaoChina
Korea
Hong
Kong -
Japan
Thailand
Indonesia
Uruguay
Mexico
Latin America
OECD
Avg
Level 6
Level 3
Level 2
200
Level 1
100
Below Level 1
0
Brazil
Percentage of Students in Each
Proficiency Level
PISA 2003 Math Results
M ean
Scores tend to be Lower in TVET
Schools
Inequities in Educational Outcomes



Inequities between urban-rural location and
income quintile exist in all countries, in terms of
enrollment, timely enrollment and completion.
Inequities tend to be higher in LAC countries, with
about three times as many youth completing
secondary in the upper quintile than in the lower
one in Mexico and Guatemala (almost comparable
to Cambodia in EAP).
As enrollment increases, inequities have had a
tendency to decrease in EAP countries, the
evidence is less clear-cut in LAC.
Decreased Inequity in Completion
Rates (Indonesia)
2003
0.75
0.25
0.50
% completed
0.50
0.25
0.00
0.00
% completed
0.75
1.00
1.00
1998
0
1
2
3
4
5
6
grade
quint = poorest
7
8
9
10
quint = richest
11
12
0
1
2
3
4
5
6
grade
quint = poorest
7
8
9
10
quint = richest
11
12
Persistent Inequity in Completion Rates
(Mexico)
1998
2002
1.00
Kaplan-Meier survival estimates, by topquint
0.00
0.00
0.25
0.25
0.50
0.50
0.75
0.75
1.00
Kaplan-Meier survival estimates, by topquint
0
0
5
10
15
5
10
analysis time
analysis time
topquint = 0
topquint = 0
topquint = 1
topquint = 1
15
Main Challenges



Two inter-related objectives:
Expand Gross Secondary
Enrollment, Net Secondary
Enrollment and Completion, in an
equitable way; and
Improve Quality and Relevance of
Secondary Education.
Supply-side Constraints and
Opportunities at the Macro Level




Primary graduates (a constraint in low-income
countries). Primary completion rates < 70%. An
opportunity in the others.
Timely primary completion (a constraint in all
countries).
Quality of primary graduates (a constraint in
several countries).
Demographic trends (favourable in most
countries). Over the next decade, in all but five
countries, total population will increase faster
than the secondary school age population. For
half the countries the secondary school age
population will decline in absolute terms.
Demand-side Constraints and
Opportunities at the Macro Level



Increasing demand for secondary
education from labor markets:
With globalization comes a change
in the type of labor demand,
favoring those workers who have
the skill sets developed through
secondary education.
Innovation and technological
advancement require a labor force
with sufficiently advanced skills.
Demand-side Constraints and Opportunities:
Labor Market Structures
o



A preliminary analysis of household
surveys suggests that:
In most LAC and EAP countries, labor
market composition is increasingly based
on services/sales/trade (more true of
middle-income countries), which require
relevant skills learnt in secondary.
Share of wage employment is also
increasing and associated with more than
9 years of schooling in all countries.
High rates of return (see Chapter 2).
Demand-side and Supply-side
Constraints at the Micro Level


Low transitions between primary
and secondary and sub-cycles of
secondary in several LAC and EAP
countries can suggest supply-side
(access and quality) constraints.
Transitions are more marked in
rural than urban areas.
In all countries, between 20% and
50% of the youth are out of school
at 15, suggesting demand-related
issues, aggravated by over-age.
Understanding Demand-side
Constraints - Chapter 2
This chapter analyzes household demand
for secondary education, with focus on the
two main determinants of household
demand:
a.
b.
expected benefits (labor market returns
to education); and
costs (with a focus on opportunity
costs).
Demand for Education and Private
Rates of Return: Substantial Returns



Private rates of return are higher than 8% in all
EAP and LAC countries, slightly higher in EAP.
Rates of return increase with the education level
in all selected countries, with higher returns for
technical education and very substantial in
tertiary.
Incentives to enroll vary according to the
secondary modality. Rates of return are higher in
TVET, but high returns in tertiary make general
secondary attractive in middle-income countries.
Importance of secondary-tertiary transition.
Increasing Returns by Level of
Education
30
25
20
15
10
5
0
Chile
Argentina
Brazil
primary
Mexico
Guatemala Singapore
general secondary
Thailand
China
technical secondary
Indonesia
tertiary
Mongolia
Demand for Education and Private Rates of
Return: Heterogeneity in Rates of Return



Rates of return decrease with earning
quintiles in lower-income or lower-middle
income EAP countries, but increase across
earning quintiles in upper-middle-income
EAP countries and most LAC countries
(providing little incentive to enroll for lower
income people).
The differences per development stage can be
explained by different labor market structure.
Increasing returns in LAC could be explained by
lower quality of the education received by the lower
quintile. Investing in quality education for the poor
will be key to expand secondary enrollment of the
poor in the region.
Rates of Return Vary per Income Quintile
Chart 4a: Return to One More Year of Education by
Quantile and Gender: Wage Earners 25-65 years,
Indonesia 2003
Chart 3a: Return to One More Year of Education by
Quantile and Gender: Wage Earners 25-65 years, Vietnam
2001-02
MALES
MALES
FEMALES
FEMALES
20
12
Return (%)
Return (%)
10
8
6
4
15
10
5
2
0
0
Q10
Q25
Q50
Q75
Q10
Q90
Q25
Chart 13a: Return to One More Year of Education by
Quantile and Gender: Wage Earners 25-65 years,
Venezuela 2002
MALES
Q50
Q75
Q90
Quantile
Quantile
Chart 16a: Return to One More Year of Education by
Quantile and Gender: Wage Earners 25-65 years, Brazil
2001
FEMALES
MALES
FEMALES
14
20
10
Return (%)
Return (%)
12
8
6
4
2
0
15
10
5
0
Q10
Q25
Q50
Quantile
Q75
Q90
Q10
Q25
Q50
Quantile
Q75
Q90
Demand for Education and School to
Work Transition

Opportunity costs can be a strong
disincentive to continue and/or do well in
school. Using detailed household survey
data, the study will explore:
a.
b.
c.
the extent, intensity, nature, remuneration of
work in teenagers;
the interaction between schooling and work
(age of transition to work, school and work,
work only, school only, neither);
a characterization of working teenagers and
their families.
Demand for Education and Private
Costs of Schooling


Initial evidence shows that fees and
indirect costs of schooling can be
prohibitive for low-income households
across the two regions.
Regional and development patterns will be
identified and innovative interventions
discussed. The LAC case has shown that
demand-side financing has good potential
to address opportunity and private costs
(e.g. Oportunidades in Mexico, Bolsa Escola
in Brazil or PACES in Colombia).
Improving Governance and
Management Structures - Chapter 3
The chapter analyzes:
1.
2.
3.
Structure of Secondary Education
Allocation of Functions and
Decisions across Governments
and Institutions
Impact of Governance Structures
Structure of Secondary Education and
Implicit Governance Issues
Cycles (one vs. lower and upper, basic vs. lower
and upper)
Public and/or Private Schools
General and Vocational Education
1.
2.
3.
a.
b.
c.
review the status and literature—developing
countries & LAC/EAP;
analyze practices in light of the literature review
and conceptual framework;
connect to demand side of the youth employment
equation
i.
ii.
iii.
structure of labor markets
match of demand for skills and what education
system supplies
does demand justify significant increased
investment and separate general and TVET
cycles?
General and Vocational Education:
Enrollment Shares



The average enrollment share in
vocational-technical schooling for LAC is
9% (decreasing), while for EAP it is 11%.
Enrollment share relatively homogeneous
in EAP, but large differences within LAC.
In high income countries, 27% of
secondary schools students are enrolled
in vocational education, vs. 2% in low
income countries.
General and Vocational Education:
New Trends



Traditionally, in LAC as in EAP, vocational schools
have been separated out from general secondary
schools. However, there is an increasing focus
towards broader and more integrated curricula.
In LAC, vocational secondary education is
decreasing but there is growth at the postsecondary level and an increase in industry based
training, together with better integration
secondary-tertiary (Colombia, Chile, etc).
EAP countries have put more emphasis on
modernizing existing secondary vocational
systems.
General and Vocational Education: Basic
Labor Market Facts




With the exception of Argentina and Indonesia,
unemployment rates for secondary graduates are
below 20% in all countries (< average rate).
Unemployment rates tend to be lower for
technical secondary graduates.
Between 30 and 70% of secondary graduates are
wage employed. Shares are higher in middleincome countries.
A deeper analysis of demand for skills is needed
as well as a cost analysis of alternative
modalities.
Secondary Graduate Profiles
Vietnam general secondary
wage employed
self-employed
unemployed
Bolivia general secondary
wage employed
unpaid employment
self-employed
unemployed
Vietnam technical secondary
wage employed
self-employed
unemployed
Bolivia technical secondary
wage employed
unpaid employment
self-employed
unemployed
Allocation of Functions and Decisions
Across Governments and Institutions
1.
Relative spectrum of decentralization
a.
b.
c.
2.
3.
Legal Framework & History
Different Models of Decentralization: to
governments and/or schools
Government relations to private schools
Locus of Decision-making: focus on
administration, personnel, & quality
control
Implementation Issues:
a.
b.
de jure vs. de facto
Weak links in accountability chain
Relative Spectrum of Decentralization


In general, education systems in LAC are
organized regionally or locally. In contrast,
in EAP countries there is a tendency
towards centralized administration of
education.
Schools are decentralized responsibilities
in both regions (more so in lower income
LAC countries and middle or upper
income EAP countries).
Locus of Decision-Making
Impact of Governance Structures: Focus on
Decentralization and School Organization
1.
2.
Assessment of Accountability
Relationships
School-Based Mgmt & School
Committees
a.
b.
3.
Legal Framework, powers & member
selection
Analysis of decision-making
Impact of Decentralization on Outcomes
a.
b.
Cross-country setting
Individual case studies
Evidence on School Based Management
from PISA 2000 and 2003 Analysis


School autonomy generally
higher in Honk-Kong and
Thailand than in the selected
sample of LAC countries.
School boards have slightly
larger roles in LAC. Principals
and teachers have slightly
larger roles in EAP.
ve
ra
ll
N bu
o
d
a g
Sc u to e t:
ho no
ol my
Bo
Pr ard
in
Te ci p
ac al
Bu
he
dg
rs
et
al
lo
N c at
o
a ion
Sc u to s :
ho no
ol my
Bo
Pr ard
in
Te ci p
ac al
he
rs
As
se
s
N sm
o
a en
Sc u to t:
ho no
ol my
Bo
Pr ard
in
Te ci p
ac al
he
rs
Te
x
N tb o
o
a ok
Sc u to s :
ho no
ol my
Bo
Pr ard
in
Te ci p
ac al
he
C
rs
ou
rs
e
C
N on
o
au ten
Sc to t:
ho no
ol my
Bo
Pr ard
in
Te ci p
ac al
he
rs
O
Decision-Making Autonomy within Schools:
Varying Role of the Different Actors
180
160
140
120
100
80
60
40
20
0
EAP
LAC
Evidence on School Based Management
from PISA 2000 and 2003 Analysis



Autonomy in teacher management
positively related to achievement in LAC,
but not EAP.
Autonomy in texts and course contents
positively related to achievement in EAP,
but not LAC.
Accountability argument more important in
LAC; but local informational advantage
argument more important in EAP.
Mobilizing Resources – Chapter 4

1.
2.
3.
This chapter:
estimates resources needed to expand
secondary education coverage and
financing gaps given current expenditure
levels, secondary targets and growth
prospects;
analyzes rationale for public versus
private intervention in financing the
expansion of secondary education; and
analyzes alternative options to raise
resources for secondary education.
Lack of Priority on Secondary
Education, more so in LAC than EAP




Within each region countries differ drastically, but
the two regions are facing similar types of
challenges: increasing gross enrollment rates and
expenditure on secondary education.
Expenditure on secondary education is low in terms
of GDP and per student in both regions (more so in
LAC than EAP).
It is low for expenditure per student as a percent of
GDP per capita in LAC, while high in EAP.
These ratios are lower in low income countries.
4.00
3.00
2.00
1.00
5.00
6.00
Low income Lower middle Upper middle
income
income
EAP
All education, 4.85 %
Secondary, 1.45%
All Education, 5.87%
Secondary, 1.91 %
All Education, 6.03%
Secondary, 1.74 %
All Education, 4.38%
Secondary, 1.64 %
All education, 4.81%
-
Secondary, 0.96%
Secondary Education Shares in terms
of GDP are still Low
7.00
LAC
Expenditure per Student in Secondary
is much Lower than OECD average
7,000.00
Primary
6,000.00
Secondary
5,000.00
4,000.00
3,000.00
2,000.00
1,000.00
-
Low
Lower-
Upper-
income
middle
middle income
income
EAP
LAC
OECD
Average
But Secondary Expenditure per Student in
Percent of GDP per capita is High in EAP
$7,000
Per Student Expenditure
30
Per Student expenditure as % of per
capita income
25
$6,000
20
PPP
$5,000
$4,000
15
$3,000
10
$2,000
5
$1,000
$0
-
Low income Lower middle
income
Upper
middle
income
EAP
LAC
OECD
Average
% of GDP per capita
$8,000
Estimated Fiscal Gap for a given GER target and at
current Unit Costs in percent of GDP: Higher Gaps
for Lower Middle Income and EAP countries
GER Target
75%
85%
95%
15% above
current rates
Low income
Lower middle income
Upper middle income
EAP
LAC
Two regions combined
Low income
Lower middle income
Upper middle income
EAP
LAC
Two regions combined
Low income
Lower middle income
Upper middle income
EAP
LAC
Two regions combined
Low income (68%)
Lower middle income (86%)
Upper middle income (106%)
EAP
LAC
Two regions combined
Resources needed as % of GDP
1.07
2.82
1.87
3.28
1.66
2.23
1.53
2.92
2.15
3.60
1.87
2.47
1.71
3.24
3.00
3.92
2.15
2.80
1.30
2.50
2.59
2.79
2.21
2.40
Fiscal gap as % of GDP
0.55
1.04
0.22
1.58
0.24
0.71
0.55
0.98
0.34
1.41
0.34
0.72
0.73
1.24
1.00
1.52
0.52
0.89
0.32
0.52
0.52
0.47
0.51
0.50
Estimated Fiscal Gap for a given GER target and at OECD Unit
Costs in percent of GDP: Higher Gaps for Lower Income
Countries, increasing gaps for LAC countries
GER Target
75%
85%
95%
Increase by
15%
Low income
Lower middle income
Upper middle income
EAP
LAC
Two regions combined
Low income
Lower middle income
Upper middle income
EAP
LAC
Two regions combined
Low income
Lower middle income
Upper middle income
EAP
LAC
Two regions combined
Low income (68%)
Lower middle income (86%)
Upper middle (106%)
EAP
LAC
Two regions combined
Resources needed as % of GDP
2.73
2.46
2.40
2.55
2.45
2.47
3.10
2.77
2.60
2.73
2.74
2.74
3.46
3.09
2.91
3.06
3.07
3.06
2.52
2.96
3.27
2.74
3.16
3.04
Fiscal gap as % of GDP
1.74
1.37
0.73
1.49
1.07
1.18
2.11
1.46
0.83
1.29
1.27
1.28
2.47
1.78
1.13
1.62
1.59
1.60
1.53
1.65
1. 29
1.30
1.54
1.48
Alternative Options to Raise Resources

Alternative options to raise resources
are analyzed, looking at:
Public sources:
1.
a.
b.
c.
d.
e.
taxing capacity (LAC and EAP lag behind
world average on revenue effort even
controlling for income levels)
fiscal decentralization
revenue earmarked to secondary education
(lotteries, targeted taxes)
cross-sectoral sources
re-allocation across education levels
Alternative Options to Raise Resources
Public-private cost sharing schemes:
2.
a.
b.
c.
Private resources:
3.
a.
b.
c.
4.
household and community contributions
private contributions (firms)
public-private cost sharing in private schools
privately funded/managed formal education
(private schools, private companies funding
TVET, etc)
completing financial markets (savings
accounts, scholarships to poor students, etc)
informal education: tutoring
International funding.
Mobilizing Resources: Some Lessons






Countries that achieved significant
expansion of secondary education share the
following characteristics:
Use multiple resource mobilizing schemes
Have moved towards more decentralized
systems
Have incorporated more private funding in
secondary education delivery
Have successfully and innovatively mixed
public and private resources
Kept per pupil costs under control
Improving the Efficiency of Service
Delivery – Chapter 5

1.
2.
3.
Countries could substantially expand and
improve secondary education with less
resource mobilization efforts by being
more efficient. This chapter:
measures cost-effectiveness levels in
LAC and EAP (output and input
efficiency);
explains efficiency results;
reviews interventions which can be
efficiency enhancing.
Measurement of Cost-Effectiveness


Cost-effectivenes is measured through
DEA non-parametric techniques, relating
selected outputs to one or more inputs
(expenditure, or physical/human inputs).
Existing analysis shows no clear
difference between EAP and LAC countries
in input-efficiency, but that EAP countries
tend to be more output efficient in terms
of secondary enrollment and completion.
Input and Output Efficiency Results
(Herrera and Pang, 2004)
Table 2: Efficiency for educational outputs
Gross primary
enrollment
Net primary enrollment
Gross secondary
enrollment
Net secondary
enrollment
Mean years of schooling
First level complete
Second level complete
Youth literacy
Input efficiency
EAP LAC
0.74
0.74
Output efficiency
EAP LAC
0.79
0.82
0.78
0.69
0.77
0.69
0.90
0.70
0.93
0.61
0.71
0.69
0.84
0.66
0.36
0.43
0.37
0.73
0.32
0.36
0.32
0.72
0.79
0.50
0.38
0.99
0.60
0.36
0.24
0.94
DEA/FDH Estimation on PISA 2000:
Input Efficiency

1.
2.
3.
Main Findings:
more heterogeneity across LAC countries
in terms of input-efficiency (low
efficiency in Argentina, high in Chile and
Mexico);
more homogeneous, but on average
slightly lower, in EAP countries;
in Argentina and Indonesisa, results are
also particularly heterogeneous within
countries.
PISA 2000: Input Efficiency Scores
(higher score is associated with lower input efficiency)
3.5
3
2.5
2
1.5
1
0.5
0
ARG
BRA
CHL
MEX
PER
HKG
IDN
JPN
KOR
THA
DEA/FDH Estimation on PISA 2000:
Output Efficiency

1.
2.
3.
Main Findings:
results are more homogeneous in terms of
output-efficiency, with slightly higher
results for EAP;
Chile and Mexico do well also in outputefficiency terms, while Brazil does poorly;
there is generally more variability within
countries in LAC than EAP, and variability
is particularly high in Brazil and Indonesia.
PISA 2000: Output Efficiency Scores
(higher score is associated with higher output efficiency)
0.94
0.92
0.9
0.88
0.86
0.84
0.82
0.8
0.78
0.76
1
ARG
BRA
CHL
MEX
PER
HKG
IDN
JPN
KOR
THA
Gains from Efficiency Improvements

Initial evidence suggests large potential
gains from input-efficiency
improvements. Potential gains from
output-efficiency improvements are
smaller, although substantial in some
countries such as Brazil and Indonesia
(by eliminating inefficiency, outcomes
could be scaled up by about 20% in
both countries).
What Explains Efficiency Results and
How to Enhance Input Productivity?






Impact of different budget shares;
Impact of general vs technical secondary
education;
Impact of management reforms (school-based
management, public/private schools, publicprivate partnerships);
Impact of flexible and innovative delivery models
of secondary education;
Impact of pedagogical and curricular practices;
Impact of repetition and internal efficiency
improvements.
Impact of General vs Technical
Secondary Education


Vocational and technical education is
more expensive than general academic
school in secondary education (by from
1.14 times to 7.20 times, due to lower
pupil-teacher ratios, more equipment,
etc).
Technical and vocational schools appear
generally less input efficient in the PISA
analysis.
Impact of Flexible Delivery Models



Alternative models tend to have lower per
student costs than conventional schools
even while providing access to
populations and areas that are typically
more expensive to serve.
Three main types of alternative secondary
models are: Distance Education; Open
Learning; and Group Study.
International evidence suggests that
these models can work for specific target
populations. Successful cases in LAC and
EAP will be more thoroughly analyzed and
some systematic lessons extracted.
Impact of Management Structures


Initial evidence suggests that
schools with higher decision-making
autonomy in human resources are
more output efficient in LAC
countries.
Private schools tend to be more
efficient, although more so in LAC
countries.
Impact of Pedagogical/Curricular
Practices


Initial evidence suggests that schools
using external assessment practices are
more output efficient in Brazil, Thailand
and Honk Kong.
Improving the alignment between
different parts of the system to ensure
smooth transitions between sub-cycles
and cycles (case study on Malaysia) is
another promising intervention.
Impact of Reduced Repetition



By decreasing repetition in primary and
secondary, over-age decreases, early drop-out
decreases and educational attainment increases.
The NER increases as well.
Reducing repetition can also, potentially, provide
important savings:
initial evidence suggests that repetition, and the
cost of repetition, are higher in LAC than EAP and
in primary than secondary education. In total,
LAC countries could save over US$ 2 billion per
year by eliminating repetition at the secondary
level, savings are less substantial in EAP.
Cost of Repetition: High in LAC
Cost of repetition as percent of recurrent education
expenditure
6.0%
5.5%
5.0%
4.3%
4.0%
3.0%
Primary
2.6%
Secondary
2.0%
1.0%
0.2%
0.0%
EAP
LAC
Impact of Reduced Repetition


Issues: how can repetition be reduced ? What is
an advisable repetition level ? What is the
likelihood that reductions in repetition translate
into actual cost savings ?
A case study is being undertaken on two countries
in EAP and LAC (Cambodia and Chile) which have
successfully decreased repetition in secondary.
The policy combinations to achieve this reduction
are analyzed, as well as the consequences of this
decrease on coverage and quality indicators, and
on costs.
Some Emerging Messages



Several challenges have emerged in
secondary education in both EAP and LAC,
related to low coverage, completion and
quality.
Coverage and quality challenges are interrelated.
Challenges are more acute in lower
income countries.
Some Emerging Messages


Significant demand for secondary
graduates (both general and technical),
although household demand is still
constrained, in particular for lower income
youth in LAC (contributing to explain
higher inequalities in LAC).
Technical-vocational secondary education
can be effective but is more costly. More
analysis is needed on relevance of TVET
and general education to labor market
needs and relative costing.
Some Emerging Messages

There is scope for mobilizing more resources for
secondary education, while also improving the
efficiency of service delivery in both EAP and LAC:



Secondary education in under-funded in both EAP and
LAC (more so in LAC and in lower-income countries).
Fiscal gaps are substantial for lower-income and lowermiddle income countries and for both EAP and LAC;
Scope for input-efficiency improvements in all
countries, although more so in EAP than LAC;
Also scope, although less so, for output-efficiency
improvements, more so in LAC than EAP and in lowermiddle income countries.
Some Emerging Messages

A broad range of options need to be
looked at to mobilize resources and
improve efficiency:


Options to mobilize resources include using
multiple resource mobilization schemes,
decentralized financing, and innovative publicprivate resource mixes.
Possible options to improve output and input
efficiency include school autonomy, flexible
delivery models, use of assessment practices
and a decrease in repetition (more so for LAC).