SCHOOL CHOICE, QUALITY & VALUE ADDED
LEAPS research uncovers how school quality is measured and what shapes parent and student choices in Pakistan’s evolving education system. We study the effectiveness of public and private schools, what value-added estimates reveal about learning gains, and how school fees, distance, amenities, and vouchers influence parental decisions. Our research also examines patterns in religious school enrollment to provide a complete picture of where and why families send their children to school. To know more, explore the studies below.
Heterogeneity in School Value Added
Authors: Tahir Andrabi, Natalie Bau, Jishnu Das and Asim I. Khwaja
Citation: Andrabi, Tahir and Natalie Bau and Jishnu Das and Asim I. Khwaja. 2025. “Heterogeneity in School Value Added and the Private Premium”, American Economic Review 115(1): 147–182.
Questions and Findings
Do private schools consistently outperform public schools? What explains differences in student learning gains (“value added”) across schools, and how do parents respond to information about school quality?
School Quality is Highly Uneven: Value added varies widely, even within villages; the top school outperforms the bottom by an average of 0.89 SD in student test scores per year.
Not All Private Schools Are Superior: Private schools average 0.08 SD higher value added than public, but the top 30% of public schools outperform the bottom 30% of private schools.
Substantial Student Learning Growth: Students gained an average of 0.40 SD per year—matching or exceeding learning gains in many low-income countries.
Parental Choices Reflect School Quality: Parents actively favor higher-value-added schools, enrolling their children and paying more for demonstrated learning gains; a one SD increase in value added raises a school's market share by 7 percentage points.
Price Responds to Quality: Private schools with one SD higher value added charge, on average, 18% higher fees than those with lower value added.
Why This Matters
This study reveals that the “private premium” in education is not simply due to all private schools being better, but is largely driven by substantial quality differences within both sectors. Many public schools perform as well as or better than private ones, and parents actively seek out higher-quality options when given information and choice.
These findings suggest that expanding informed school choice, increasing the availability of transparent quality data, and focusing on raising quality standards in all schools—not just promoting privatization—are crucial steps for policymakers and education leaders seeking to improve learning. The study’s robust measurement of value added also provides actionable benchmarks for future research and accountability efforts in low-resource settings.
The Value of Private Schools
Authors: Pedro Manuel Carneiro, Jishnu Das, Hugo Reis
Citation: Carneiro, Pedro Manuel and Jishnu Das and Hugo Reis. 2016. “The Value of Private Schools: Evidence from Pakistan”, IZA Discussion Paper No. 9960.
Questions and Findings
What factors (such as school fees, distance, test scores, and amenities) most influence parental choices between public and private primary schools in poor villages in Pakistan? How would introducing school vouchers affect school enrollment patterns and household welfare in these communities?
School Fees and Distance Matter: Families are willing to pay 75–115% of the average private school fee to reduce school distance by 500 meters, and a 10% fee cut increases enrollment by 2% for boys and 5% for girls; making schools free (from $13 to $0) raises enrollment by 7.5 percentage points for girls and 4.2 points for boys.
Peer and School Characteristics Influence Choice: Parents consider peer quality and amenities when choosing schools, and access to low-fee private schools provides substantial welfare value equal to 25–100% of monthly per capita income for users.
Vouchers Raise Enrollment and Welfare: Both simulations and a real voucher program show that subsidizing private school fees increases enrollment and raises welfare even for families whose children stay in public schools.
Equity Gains for the Disadvantaged: Welfare benefits from vouchers and low-fee private schools are largest for the poorest and least educated families, helping to reduce educational inequality in rural Pakistan.
Why This Matters
These results point to important priorities for education policy in low-income countries. Families value factors like distance, school environment, and peer quality—not just test scores—so interventions should reflect diverse parental preferences. Since parents respond to changes in school fees, targeted vouchers or subsidies could meaningfully boost access to quality education, especially for disadvantaged families.
A key policy implication is that expanding support for affordable private schools can promote both educational access and equity. Altogether, the findings guide policymakers toward effective regulation and resource allocation for both private and public schools.
Do Value-Added Estimates Add Value?
Authors: Tahir Andrabi, Jishnu Das, Asim Ijaz Khwaja, Tristan Zajonc
Citation: Andrabi, Tahir and Jishnu Das and Asim Ijaz Khwaja and Tristan Zajonc. 2011. “Do Value-Added Estimates Add Value? Accounting for Learning Dynamics”, American Economic Journal: Applied Economics 3(3): 29–54.
Questions and Findings
Does the value-added model (VAM)—which estimates the contribution of schools to student learning based on year-to-year test score growth—accurately measure school effectiveness in primary schools in Pakistan, especially in settings where learning persistence is low?
Most test score gains don’t last: After one year, only 20–50% of students’ test score improvements remain; 50–80% of the gains from one year are gone by the next. This means annual test score gains largely fade out over time.
School performance rankings are unreliable from year to year: Because learning gains fade so much, 43% of schools that are ranked as "top" or "bottom" performers in one year move out of those categories the next year when using standard value-added models. This instability makes it hard to identify consistently good or bad schools.
Private schools have a real, but sometimes overstated, learning advantage: When properly accounting for fading, private schools produce about 0.24–0.34 standard deviations more learning per year compared to government schools. However, if the model ignores the fade-out problem, this advantage is artificially inflated by about 45%.
Incorrect models lead to mislabeling many schools: If learning fade-out is ignored, value-added models can misclassify up to 26% of schools—meaning that many truly effective schools might be labeled as underperforming, and some less effective schools might appear to be doing well.
Why This Matters
Traditional value-added models, which aim to assess school quality based on student progress, can be misleading in contexts where learning gains are not sustained over time. For education systems—especially in low-income countries—this can distort accountability, resource allocation, and policy incentives, affecting school funding, teacher evaluations, and parent choices.
To improve educational decision making, policymakers need to factor in learning persistence when evaluating schools or interventions, and develop more accurate methods that reflect the actual dynamics of student achievement in their contexts.
Religious School Enrollment in Pakistan
Authors: Tahir Andrabi, Jishnu Das, Asim Ijaz Khwaja, Tristan Zajonc
Citation: Andrabi, Tahir, Jishnu Das, Asim I. Khwaja, and Tristan Zajonc. 2006. “Religious School Enrollment in Pakistan: A Look at the Data.” Comparative Education Review, 50 (3): 446-477.
Questions and Findings
What is the actual share of children in Pakistan enrolled in religious (madrasa) schools, and are concerns about rapid growth in madrasa enrollment supported by the data?
Madrasa Enrollment Is Very Low: Madrasa enrollment constitutes less than 1% of all enrolled children in Pakistan nationally, and even in rural Punjab—the province with the highest rates—enrollment is only about 2%.
No Evidence of Rapid Growth: Survey data across multiple years show that madrasa enrollment rates have remained almost unchanged since the mid-1990s; for example, national madrasa participation in 1997 and in 2001 is effectively the same.
Poverty Is Not the Main Driver: The vast majority of out-of-school or poor children do not enroll in madrasas. Even among the poorest households, less than 1-2% of children attend madrasas; enrollment in public or private schools overwhelmingly dominates.
Lack of Alternatives Matters Only Rarely: Madrasas are most likely to be attended when no other schools are present in a settlement, but even in those areas, overall madrasa participation remains modest.
Why This Matters
Understanding the true scale of madrasa enrollment is essential for informed education policy and public debate in Pakistan. By dispelling myths of widespread or rapidly growing religious school attendance, this study helps redirect attention and resources toward the real challenges faced by the education system—such as improving the quality of and access to public and private secular schools.
Effective policy must be grounded in facts, not sensationalized narratives, and these findings encourage a shift toward evidence-based reforms that address the needs of the vast majority of Pakistani children.