The Impact of Outsourcing School Management in Liberia
Public-private partnerships to provide education in low-income countries are common, yet controversial. In Liberia, researchers worked with IPA, the Ministry of Education, and a set of eight private operators to conduct a randomized evaluation that measured the impact of 93 partnership schools—free public schools with management outsourced to private operators. After three years, partnership schools raised test scores by 0.21 standard deviations in math and 0.16 standard deviations in English (equivalent to 4 words per minute additional reading fluency for the cohort that started in first grade). Looking beyond learning gains, the program reduced corporal punishment (by 4.6 percentage points from a base of 51%), but increased dropout (by 3.3 percentage points from a base of 15%) and failed to reduce sexual abuse. Results vary by provider: some produced uniformly positive results, while others present stark trade-offs between learning gains and other outcomes.
Despite progress in primary school access around the world, learning levels remain low. An important question in low- and middle-income countries is whether the private sector can improve learning outcomes when public school systems fail. Public-private partnerships in education are common around the world. However, their role is controversial. Proponents of such partnerships argue that private operators may be able to operate schools more effectively, delivering the same or better results for students at lower cost than resource-limited governments. Opponents argue that monitoring challenges, limited outside options for citizens in remote areas, and a narrow focus on test scores at the expense of other aspects of students’ well-being such as access and child safety may create harm. The rigorous evidence on this question is mixed, with little experimental or quasi-experimental evidence from low- and middle-income countries.
Context of the Evaluation
In 2016, the Liberian government outsourced the management of 93 randomly selected public schools, comprising 8.6 percent of public school students, to eight different private operators in a program called the Liberian Education Advancement Partnership (LEAP). (The program was later expanded to 101 additional schools.) The government’s primary, stated motivation for the outsourcing program was the low level of learning outcomes in public schools. According to Demographic and Health Survey data, only 25% of adult Liberian women who had completed primary school were literate, one of the lowest rates in the world.
The model is similar to charter schools in the United States or academies in the United Kingdom. However, in contrast to many public-private partnerships in education, the teachers in the Liberian public schools that were outsourced to private providers were to remain civil servants and were still paid directly by the Liberian government. Each school was managed by one of eight private contractors, including three for-profit companies and five charities, which took responsibility for everything from teacher in-service training to fixing leaks in the roof. Providers were paid on a per-pupil basis (with some exceptions) and forbidden from charging fees or screening students based on ability.
Details of the Intervention
Researchers worked with IPA, the Liberian Ministry of Education, and the group of eight private operators to evaluate the impacts of outsourcing management of public schools with a randomized evaluation after one and three years.
A total of 185 schools took part in the evaluation. The sample of schools included was not representative of the country as a whole; rather, schools were required to meet certain minimum infrastructure requirements, and private providers were further allowed to filter the list of potential schools to be included in the evaluation based on proximity to roads and availability of cell service.
Of these 185 schools, 93 were assigned to one of the eight private operators. The other 92 schools continued to be operated by the government and served as a comparison group.
Of the eight private organizations, three are for-profit companies: Bridge International Academies (which was allocated 23 schools), Omega Schools (allocated 19 schools), and Rising Academies (allocated 5 schools). The other five non-profit providers include BRAC (allocated 20 schools), Street Child (allocated 12 schools), More than Me (allocated 6 schools), Youth Movement for Collective Action (allocated 4 schools), and Stella Maris (allocated 4 schools). Youth Movement for Collective Action and Stella Maris are Liberian organizations, while the other six organizations are international.
After one and three years, the research team measured the impacts of LEAP on school access, learning, and financial sustainability. (Child safety, measured through corporal punishment and sexual abuse, was also added to the survey that took place after three years.) Pupils who were in a school included in the evaluation before the study began were surveyed, regardless of where they went. Researchers also surveyed teachers and collected school-level information about facilities, the teacher roster, the availability of materials such as textbooks, and expenditures.
Results and Policy Lessons
Access: Outsourcing led to increased enrollment in privately-operated schools, but had a negative effect on educational attainment for the students originally in them. Negative effects on pupil-level enrollment emerged in the first year due to mass expulsions by one private operator, Bridge International Academies. After three years, there were still negative impacts on the probability of children still being in school or transitioning to secondary school. (The phenomenon, still concentrated in Bridge schools, appears unrelated to the expulsions in the first year and is driven in part by higher rates of drop-out attributed to pregnancy.)
Learning: After one year, LEAP improved learning outcomes by roughly 0.18 standard deviations in both English and mathematics. After three years, the positive impacts remained statistically significant but were essentially unchanged since the first year. The gains, modest in absolute terms, varied starkly by operator: three of eight produced zero impact, while the remaining five produced somewhat larger, very similar results that drove the overall positive impact.
Child Safety: Overall, corporate punishment declined, but was not eradicated. The percentage of students reporting being hit by their teachers at least occasionally declined 4.6 percentage points from a level of 51 percent in the comparison group.
Despite an influx of new resources and external oversight in partnership schools, sexual abuse did not decline. Survey results from 2019 show 3.6 percent of students reported sexual intercourse with a teacher, and 7.5 percent reported some form of sexual contact with a teacher. While few of the operator-specific results are significant due to limited statistical power for this outcome, there were a few operators who effectively reduced reported sexual abuse to zero, while some others did not.
Sustainability: Costs were high, but fell considerably during the course of the study. At the launch of the program, the Liberian government spent roughly USD $50 per pupil in public primary schools and aimed for private operators to spend an additional $50 per pupil. However, in the first year, per-pupil expenditure by private operators was roughly $300 on average. By year 3, the overall average cost (beyond the government’s $50 per pupil) was $119.
Detailed results after one year are available in an American Economic Review article here.
Detailed results after three years are available in this working paper.
This project was made possible with support from the UBS Optimus Foundation, the Economic and Social Research Council (ESRC), the Aestus Foundation, and the Center for Global Development.
 The program was originally known as Partnership Schools for Liberia (PSL).
 Liberia Institute of Statistics and Geo-Information Services. (2014). Liberia demographic and health survey 2013. Liberia Institute of Statistics and Geo-Information Services.