Exploring aspects of private, public and private-public par

Exploring aspects of private, public and private-public partnership (PPP) schooling in Pakistan.

Ravish Amjad and Gordon MacLeodAnnual Status of Education Report (ASER) Pakistan

Background

Sub-study from ASER, Pakistan (2011) (Thank you ASER!)Large, nationwide household survey accompanied by smaller school survey.Linked individual child data with school data.Individual data included success (or not) on tests of Urdu, arithmetic and English.Samples of:26,054 Government children3997 Private children, and 159 PPP children

The Regression Analyses

Regression Analysis 1Checks for the ‘school type effect’ alone on assessment outcomesRegression Analysis 2Checks for the ‘school type effect’ on assessment outcomes after controlling for other factors, other than the type of school.

Do private and PPP school students outperformgovernment school students?

Regression 1: Yes (substantial and significant).(Private and PPP perform at same level as each other).Regression 2: After controlling for other variables (Child, household, school) a. private school children outperform govt. children b. PPP school children underperform govt. children. Conclusions:Private school children’s superiority seems due to school type.PPP school children’s superiority is due to factors other than school type.

How much do private school students pay in fees?

Fee levels per month (in Pakistan rupees)Number paying% payingCumulative%Rs. 0 – 1997361818Rs. 200 – 39921475472Rs. 400 – 5999492496Rs. 600 and above1654100

Do children from lowest cost private schools outperform government school children?

Regression 1: Yes (significant and substantial).Regression 2: After controlling for other factors, a. Yes for Arithmetic and English b. No for Urdu.Conclusion: lowest-fee private school children’s superiority in arithmetic and English seems associated with school type.

Do children from high cost private schools outperform lower-cost private school children?

Regression 1: Yes (significant and substantial) for those paying 400 or more rupees per month (top quartile of fees)Regression 2: After controlling for other variables, No. In fact highest levels of fees (top 5%) seems associated with lesser performance.Conclusion: superior performance of higher fee children seems due to factors other than school type.

Summary of outcomes

FindingApparently dueto School Type?Private > governmentYesPPP > governmentNoLowest-fee private > governmentYesHigher-fee private > lower-fee private.No

If not school type…then what?

PPP/government difference attributable to private tuition. The tuition effect alone accounts for more than 90% of the total effect after controlling for all factors!Higher fee/lower fee difference attributable to combination of Child factors (age, gender, preschooling, private tuition).

For further investigation…

Determinants of taking private tuition? Tuition is taken by 11% of government school children; 52% of PPP school children and some 66% of private school childrenDoes higher cost tuition produce better outcomes? Average amount paid for tuition rises alongside levels of private school fees.Do private schools benchmark against local govt. schools? Private schools offer facilities just a notch superior to attack demand, after which they do not have incentive to improve furtherNeed for systematic evaluation of PPP schools’ learning outcomes. Family savings on fees are being applied to private supplementary tuition.

The big issue: Quality

Private schools seem to be the winners!But:—40% of Grade 5 private school students cannot read a simple story in Urdu (end-of-Grade 2 task);—41% of Grade 5 private school students cannot divide a three-digit number by a one-digit number (end-of-Grade 2 task);—50% of Grade 5 private school students cannot read simple sentences in English (end-of-Grade 2 task).Quality should be a concern for all and not just government schools!

Thank

youravish.ita@gmail.comgordonmacl@gmail.com