Dime a Day: The Possibilities and Limits of Private Schooling in Pakistan: Dime a Day: The Possibilities and Limits of Private Schooling in Pakistan Tahir Andrabi Department of Economics Pomona College (with Asim Ijaz Khwaja and Jishnu Das) Presentation USIP February 2007


Rise of Private Schools: Rise of Private Schools Most significant and interesting phenomenon of the last 15 years Keywords: Poor, Rural, Female Low cost, Low wages Supply of Teachers What about Quality Incentives and Effort


Slide3: Pakistan Demographics Large country,140 million 50% below age 17 Age 5-15 roughly 50 million Just over half of these are in school Roughly 2 million young people enter the labor market every year Baby boomer barely identified. Population graph looks like a Christmas tree The good news our social security crisis is about 65 years away.


Slide4: Scale and Scope Punjab 246, 764 teachers 63,000 public schools Current enrollment 10 million Potential enrollment 20 million California enrollment K-12 6.2 million Rural Sindh 35,000 schools


Education Overview: Education Overview Low educational attainment net-enrollment 51 % Even in South Asian Context 83 % for India 90 % for Sri-Lanka 70 % for Nepal. Relative to its income Forecasted for its level of income: 77% Not a flat plain—Large disparities 20 percentage point difference between boys and girls Highest expenditure decile twice as much as lowest Rural female enrollment ratio (36%)


Dropouts: By the end of Grade III, 50% of children will have left school forever. : Dropouts: By the end of Grade III, 50% of children will have left school forever.


Government Schools: Government Schools Standard Story Bad Quality, Under funded Reality: Evidence on Government School Quality Scanty and Nonsystematic


What about Madrassas: What about Madrassas Standard Story Fall back option for the vast majority Reality Limited enrollment When school alternatives absent, kids dropout of educational system Household behavior complicated Extremism and Causality


Private Schools: Private Schools Standard Story Because of the use of English as a medium of instruction, and high fee structure, these institutions are best suited to serve the requirement of elite population..” Thus, their role in universalization remains only marginal.


Reality:A Silent Revolution: Reality:A Silent Revolution Explosive growth in private schools since 1980: 2500 new private schools between 1994-95; 8000 between 1999-2000 Increasing enrollment in private schools: 22 percent of enrollment in private schools in 2000 Are setting up in rural areas as well: every 2nd village in Punjab will have a private school within 5-7 years


A Market Based Phenomenon: A Market Based Phenomenon Mostly Self Owned Growth in Self Owned Are affordable: Ratio of private school fees to median per capita income is 2.5% in Punjab, 9% in the US


Low Fees (?): Low Fees (?) Low Wages For Teachers


Teacher Profile: Teacher Profile Single, Local, Female, Young


Teacher Characteristics: Teacher Characteristics Less Educated Less Experienced Less Trained


Slide15: Local Supply of Educated Women Cultural constraints on female mobility across villages Wages of local female teachers dramatically lower where supply of moderately educated women is large Private schools take advantage of that A Supply Side Story


A Striking Finding: A Striking Finding Private Schools Locate Where a generation ago government Girls Schools were set up Government Schools create the supply of educated women Public-Private Complementarity


Slide17: Limitations--Horizontal Don’t exist everywhere Larger villages Better Infrastructure Mostly a Punjab and some NWFP phenomenon Also AJK


Limitations--Vertical: Limitations--Vertical Given that cost advantage comes from low cost moderately educated teachers Not likely to extend to secondary schooling


Private School Utilization: Private School Utilization Where they exist, The Poor Do Use Them. More Female Enrollment


What about Quality: What about Quality


Slide21: Education Inputs and Outputs Measuring Inputs Household Survey Child Questionnaire Teacher Questionnaire Head Teacher Questionnaire School Questionnaire Measuring Outputs Test Scores Student Attendance


Slide22: Measuring Educational Inputs Household Survey ~1800 households (16 per mauza) Detailed two hour questionnaire Information on school choice Parents knowledge of schools Parental time inputs Child Questionnaire ~5000 children (up to 10 per school) Basic questions about families education and assets Health information including weight and height Teacher Questionnaire Remuneration and contracts Education and qualifications Teaching and training history Family circumstances Information about students Head Teacher Questionnaire Remuneration and contracts Education and qualifications Information about teachers School Questionnaire School facilities School competition School expenditures


Slide23: Measuring Education Outputs Test Scores Test piloted in 2002 Covers Urdu, English, Mathematics and Civics (Civics is not scored) Given to over 12,000 class 3 children Attendance Child’s attendance recorded by teachers Child’s attendance will be tracked monthly New enrollment / Dropouts / Switches will be tracked monthly


The adjusted gap between public and private schools: The adjusted gap between public and private schools English 15 times the adjusted gap between rich and poor children. Mathematics 10 times the adjusted gap between children with literate and illiterate fathers. Urdu 14 times the adjusted gap between children with literate and illiterate mothers.


Incentives!: Incentives!


Slide29: The Setting up of Private Schools


Mostly Self-Owned: Mostly Self-Owned


Growth in Self-Owned: Growth in Self-Owned


Dime A Day: Dime A Day Less than Rs. 6 per day!!!


Teacher Profile : Teacher Profile Female, Young, Single, Local


Government vs. Private: Government vs. Private


Private School Location: Private School Location


Females, Poor Use Them: Females, Poor Use Them


Females and Private Schools: Females and Private Schools


Slide45: Salary results are presented as “deviations from mean”. So the number 200 on the vertical axis means that the person’s salary is Rs.200 more than the average salary for the sector The figure is based on a non-parametric plot of deviations from mean salary against teacher’s English test scores. -200 -100 0 100 200 300 Deviation from Mean Salary in Rs 0 .2 .4 .6 Student Test Scores Private Schools Public Schools Student Test Scores and Teacher Compensation


Slide46: -400 -200 0 200 400 600 Deviation from Mean Salary in Rs 0 10 20 30 Days Absent per Month Private Schools Public Schools Teacher Absenteeism and Compensation Salary results are presented as “deviations from mean”. So the number 200 on the vertical axis means that the person’s salary is Rs.200 more than the average salary for the sector The figure is based on a non-parametric plot of deviations from mean salary against the number of days absent.


What Next?: What Next? Teacher Governance Disentangle the production function Value added Interventions Dropouts and repeaters