International Evidence on Public-Private Partnerships to Improve Access and Quality in Education

1
International Evidence on Public-Private
Partnerships to Improve Access and Quality in
Education
Harry Anthony PatrinosWorld BankOctober 2007
2
Public-Private Partnerships Defined

• No fixed definition of PPPs
• Definitions differ in terms of scope and
  formality of arrangements
• Various definitions
• risk sharing relationship based upon an agreed
  aspiration between the public and private sectors
  to bring about a desired public policy
  outcome.
• Commission on UK PPPs
• cooperative venture between the public and
  private sectors, built on the expertise of each
  partner, that best meets clearly defined public
  needs through the appropriate allocation of
  resources, risks and rewards.
• Canadian Council for PPPs

3
Common Elements

• Formal arrangement with contractual basis
• Involve public and private sectors
• Outcome focus
• Sharing of risks/rewards between public and
  private sectors
• Recognize complementary role of public and
  private sectors

4
Role of Government in Education

• Rationale for government involvement in
  education
• Externalities
• Capital market imperfections
• Agency concerns
• Equity
• Information asymmetries
• Government has a variety of policy instruments at
  its disposal in order to meet its policy
  objectives
• Ownership/Delivery
• Funding
• Regulation/Information
• PPPs recognise that governments can meet their
  policy objectives using different service
  delivery models not just traditional public
  finance/public delivery model

5
Financing and Provision
6
Benefits of Public-Private Partnerships

• Increase efficiency improved performance
  incentives and increased competitive pressure
• Improved quality of service delivery
• Secure specialized skills that may not be
  available in government agencies
• Overcome public service operating restrictions
  obsolete salary scales, out of date civil service
  work rules, etc
• Permit quicker response to changing demands and
  facilitate adoption of service delivery
  innovations

7
Benefits of PPPs (contd)

• Benefit from economies of scale regardless of the
  size of the government entity
• Allow government agency to focus on functions
  where it has a comparative advantage
• Increase access, especially for groups who have
  been poorly served under traditional forms of
  service delivery
• Increase transparency of government spending by
  making the cost of services more visible

8
Types of PPPs in Education Types of PPPs in Education Types of PPPs in Education Types of PPPs in Education
What government contracts for What government contracts for Definition Contract types
1 Management, professional services (input) Government buys school management services or auxiliary and professional services Management contracts Professional services contract (curriculum design)
2 Operational services (process) Government buys school operation services Operational contracts
3 Education services (output) Government buys student places in private schools (contracts with school to enroll specific students) Contract for education of specific students
4 Facility availability (input) Government buys facility availability Provision of infrastructure services contracts
5 Facility availability and education services (input and output bundle) Government buys facility availability combined with services (operational) Provision of infrastructure contracts with education services contracts
Source World Bank 2005 Source World Bank 2005 Source World Bank 2005 Source World Bank 2005
9
Examples of Education PPPs

• Government contracting with private schools
• Private management of public schools
• Infrastructure PPPs
• Vouchers/subsidies
• Public/private sector affiliation arrangements
• Private sector regulation
• Innovation and research PPPs

10
Examples of PPPs
Type of PPP Examples
Contracting for the Delivery of Education Services Government sponsorship of private school students, Cote dIvoire Education services contracting, Philippines Alternative education, New Zealand Universal post primary education training policy, Uganda Fe y Alegria, South America Spain
Private Management of Public Schools Concession schools, Bogota, Colombia Railways schools, Pakistan Independent schools, Qatar Quality education for all, Punjab, Pakistan CDG Lahore/CARE schools, Pakistan Contract and charter schools, USA Transformed schools, China
Infrastructure PPPs Private Finance Initiative, UK Proyecto Prestacion de Servicios, Mexico New Schools Private Finance Project, Australia PPPs for Educational Infrastructure, Canada Offenbach Cologne Schools Projects, Germany Swinburne University of Technology, Australia National Maritime college, Ireland Montaigne Lyceum, The Hague, Netherlands
11
Examples of PPPs (contd)
Type of PPP Examples
Vouchers/Subsidies PACES, Colombia Targeted individual entitlement, independent school subsidies, New Zealand Private school subsidies, Cote dIvoire School funding, Netherlands, Denmark, Sweden Milwaukee Parental Choice Programme, USA Punjab Education Foundation programs, Pakistan
Private Sector Quality Assurance USA, Oman, Philippines
Public/Private Sector Affiliation Arrangements Ghana, Sri Lanka, Pakistan, China, New Zealand, South Africa
Innovation and Research PPPs Government programs encourage industry-public research institute research. Netherlands Government programs encourage industry-research institute partnerships and promote commercialization of public research, Australia
12
Contracting Models in Latin America
13
Programs

• Private management of public schools
• Management contracts, professional services,
  operational contracts
• Charter schools, concession schools, etc.
• Involves governments or public authorities
  contracting directly with private (for-
  non-profit) providers to manage public school
• Schools remain free to students no fees
• Schools responsible for all aspects of school
  operation
• Used mostly in disadvantaged areas
• Contract for education of specific students
• Vouchers, scholarships, direct funding of private
  ( public) schools

14
Evidence on PPPs

• Limited rigorous evidence on impact
• Ideal evaluation involves random assignment use
  true control group or natural experiment
• Much debate over impact of vouchers in Chile and
  USA, remain controversial
• Many studies on impact of charter schools, though
  few randomized trials

15
USA EMO Managed Schools/Enrollments
16

• National Charter School Research Project 2007

17
Fe y Alegría Schools in South America
18
Concession Schools, Bogota, Colombia

• Private schools contracted to manage poorly
  performing public schools
• 25 schools serving over 26,000 students
  disadvantaged students
• Autonomous
• 15 year contract
• Designed to overcome problems faced by public
  schools inability of schools to hire own staff,
  lack of labour flexibility, bureaucracy
• Schools paid 500 per student per year below
  public school unit cost

19
Evaluation of Proposals
20
Monitoring Evaluation

• Inspect school property administration
• Supervision visits to observe adherence to
  pedagogical norms and standards
• Independent evaluation of finances to see if
  academic objectives met

21
Impact Evaluation

• Propensity score and matching estimation
  technique (Barrera 2006)
• Hypotheses
• Dropout rates are lower in concessions schools
  than in similar, public schools
• No effects (yet) or small ones on test scores
• Nearby schools have lower dropout rates than
  public schools outside the influence of
  concessions

22
Test Scores
Public schools have lower test scores Concession
and public non-concession schools are similar
23
Findings

• Strong evidence of a direct effect of Concession
  Schools on dropout rates and some evidence that
  they had an impact on dropout rates on nearby
  public schools
• Positive impact on students test scores relative
  to those in public schools
• Dropout rates were 1.7 points lower, while
  mathematics and language scores were 1 point and
  2 points higher than students in similar public
  schools

24
Dropout Results Impact

• Matching
• 10 nearest estimators, common support, balance
  groups
• Direct Effect reduction in 1.7 points dropout
  rates
• Indirect Effect reduction in 0.82 points

25
Test Results Impact

• Matching
• 10 nearest estimators, common support, balance
  groups
• Effect over math test scores improvement of 2.4
• Effect over language test scores improvement of
  4

26
National Voucher Program, Chile

• Nationwide voucher program implemented in 1980
• Applies to public private schools secular
  religious
• Monthly payments made to schools on per-student
  basis
• Voucher schools must follow operational
  guidelines (basic facilities, certified teachers,
  class size)
• Vouchers cover most or all tuition at eligible
  schools

27
Enrollments Shares, Public vs Private Schools,
Chile
28
Test Score Effects from Selected Studies on Chile
Voucher Program
Private Subsidized Catholic Subsidized Private Non-subsidized
Bravo et al (1999)
Gallegos (2002)
Carnoy McEwan (2000) -
Mizala Romaguera (1999)
Sapelli (2003)
Vegas (2002)
McEwan (2001)
Mizala Romaguera (2003)
Sapelli and Vial (2002)
Mizala and others (2004)
Source Adapted from Bellei (2006)
29
When Schools Compete, How Do They Compete?

• While private enrollment rate increased by 20
  points, greater impacts in larger, more urban,
  wealthier communities
• Hsieh and Urquiola (J Public Economics 2006) use
  this differential impact to measure effects of
  unrestricted choice on outcomes using panel data
  for 150 municipalities
• They find no evidence that choice improved
  average outcomes (test scores, repetition, years
  of schooling)
• They do find evidence that voucher led to
  increased sorting, as the best public school
  students left for private sector

30
More on Chile

• Finding a rule about arbitrary assignment to
  treatment that mimics randomization is very
  important (Hoxby 2003)
• Researchers need to find control schools that
  were excluded from the reform for some reason
  that is uncorrelated with factors that affect
  their future performance
• Such arbitrary exclusion can sometimes be found
  in policy rules or natural events
• In some school choice reforms, no arbitrary
  exclusions exist
• When Chile introduced school choice, same law
  applied across entire country so variation in
  choice entirely endogenous, and no pre-treatment
  data exists
• Thus, researchers have neither pretreatment
  trends nor arbitrary assignment to treatment, and
  none of studies on Chilean vouchers is
  sufficiently credible to be given much weight

31
Colombia Plan de Ampliación de Cobertura de la
Educación Secundaria (PACES)

• Introduced in Colombia in early 1990s
• Provided 125,000 vouchers from 1992-1997
• Offered vouchers to students entering 6th grade,
  start of secondary school
• Key elements of program
• vouchers available to children from low-income
  families who had attended a public primary school
  accepted at private school
• renewable subject to satisfactory academic
  performance
• value 190 half the cost of private secondary
  school
• school received voucher funds directly from the
  bank
• schools were allowed to charge top-up fees and
• there was minimal regulation of private schools

32
PACES Voucher, Colombia

• Voucher program designed to give students from
  poor families access to secondary schooling
  (Angrist others 2002, 2006)
• Randomized trial students randomly selected
  through a lottery system and given vouchers to
  attend secondary school
• Findings
• Lottery winners were 15-20 more likely to attend
  a private school, 10 more likely to complete 8th
  grade and scored 0.2 standard deviations higher
  on standardized tests
• Program effects larger for girls
• Program cost less than the unit cost in the
  public sector
• Longer-term positive effects lottery winners
  more likely to take college entrance exam
• Increase in (proxy) high school graduation rates
  of 5-7 percentage points, relative to a base rate
  of 25-30

33
PACES Voucher, Colombia
Indicator Impact PACES Voucher Students Were
Years of Schooling 10 percentage points more likely to finish 8th grade
Grade Repetition 5-6 percentage points less likely to repeat a grade than non-voucher students
Test Scores Scored 0.2 standard deviations higher on achievement tests
Finished High School 5-7 percentage points higher than non-voucher students
Take College Entrance Exam 15-20 more likely to take the college entrance exam
College Exam Scores More likely to score 2 points higher than non-voucher
Employment 2.5-3.0 percentage points less likely to be working than non-voucher students
Being Married or Cohabitating 0.6-1.0 percentage points less likely to be married or living with someone than non-voucher students
34
PACES Voucher, Vocational

• Voucher skeptics argue that even if vouchers
  benefit recipients, they do so by improving their
  peer groups at the expense of others
• Therefore they do no benefit society as a whole
• This requires that voucher recipients have more
  desirable peers than they otherwise would have
• Bettinger, Kremer, Saavedra (2007) look at
  applicants for whom winning voucher did not lead
  to attending schools with peers with
• superior observable characteristics
• They focus on those who applied to vocational
  private schools
• Lottery losers were more likely to attend
  academic secondary schools
• Find that lottery winners had better educational
  outcomes, including higher graduation rates
  reading test scores
• Casts doubt on argument that voucher effects
  operate entirely through improving peers
  available to recipients

35
Education Contract Options and Potential Education Contract Options and Potential Education Contract Options and Potential
Type International experience Policy suggestion
Voucher Extensive Expand, but based on rigorous evaluations
Charter schools Significant, inconclusive evaluation findings Evaluate consider
PFIs Contracting for private financing and construction of schools A few countries have experience shown moderate success Consider with caution evaluate
PFIs Contracting for private actors to run schools, as well as finance and build them None (yet) Given potentially high returns, worth considering, but need a plan to involve providers and financiers, and evaluation of pilot
36
Contracting Guiding Principles

• Enabling policy, regulatory
• Split purchaser/provider role
• Capacity of contract agency
• Transparent, competitive selection
• Staged selection process
• Performance measures, incentives, sanctions
• Effective contract monitoring
• Providers maximum flexibility
• Long-term contracts with providers
• Independent evaluation

World Bank 2006