Pupils in 2017 at the Martha Tubman Elementary School, run by Bridge (now NewGlobe), in Nimba County, Liberia

In 2016, Liberia embarked on a radical educational experiment. At a stroke, the government outsourced 93 primary schools, covering 8.6 per cent of state pupils, to eight private organisations.

Those school operators — ranging from Bridge, a Silicon Valley-style for-profit technology company, to BRAC, a Bangladeshi charity — agreed to submit themselves to a three-year randomised controlled trial (RCT). As well as assessing the performance of individual operators, the intention was to draw conclusions about educational best practice. The results came out in 2019. Two years later, assessors, educators and providers are still arguing about what they mean.

The broad-brush findings were that, in aggregate, the outsourced schools brought modest educational gains, though in some cases at substantially higher cost — raising questions about long-term sustainability. There were, the study found, also possible negative effects on the rest of the educational system, which had to absorb teachers and students jettisoned from outsourced schools.

Children at the 93 schools scored 0.18 standard deviations higher in mathematics and literacy, an improvement of roughly four words per minute in reading fluency for pupils starting in the first grade. The programme also reduced the amount of corporal punishment used.

However, dropout rates increased. Nor did outsourcing have any impact overall on levels of alleged sexual abuse of children, a persistent problem in Liberia. That topic became the focus of scandal when More Than Me Academies, a US charity and one of the eight operators, was found to have covered up the rape of students by co-founder Macintosh Johnson before the launch of the programme.

Teaching English with a tablet at Martha Tubman school
Teaching English with a tablet at Martha Tubman school

“I wouldn’t say these results were a failure, but they were disappointing,” says Justin Sandefur, a senior fellow at the Center for Global Development, which led the assessment. “And instead of growing, what learning gains there were plateaued,” he adds, referring to findings that the effects of outsourcing, both on educational attainment and teacher absenteeism, faded over time.

Some of the operators disagreed with both the findings and the critical tone of the assessment, which, they said, significantly underestimated the difficulties of working in Liberia. Paul Skidmore, founder of Rising Academies, which performed among the best overall, says allowances had to be made for working in a country that was emerging from a civil war and where some teachers were functionally illiterate.

Two young students looking at the chalkboard with math problems
Maths at a school run by Omega Schools in Kenlay, near the Ivory Coast border

Aggregate results disguised significant divergence. Three of the operators — BRAC, Omega Schools and Stella Maris — had no statistical effect on learning outcomes, with Stella Maris never actually taking control of its four assigned schools (although operators disputed the RCT findings, saying it underestimated their impact on learning). If those three are excluded, the other five operators improved maths and literacy scores by a statistically significant 0.4 standard deviations.

Operators say they did even better than that. In particular, Bridge (now better known as NewGlobe) has disputed the methodology, which it argues underestimates the impact of students actually attending its schools.

Under the RCT methodology, all children on the school roster the year before the study began, including those who dropped out or went on to other schools, were counted as though they were taught at the outsourced schools. Sandefur says this “intention to treat” analysis was standard academic practice because it ensured operators could not manipulate the data by, say, weeding out underperformers.

Students standing around a flag pole during break time
Break time at the state Robert Stanley Caulfield school near Monrovia

Bridge has contested the methodology, saying it improperly assigned “intent”. It was, argues Steve Cantrell, chief impact officer at NewGlobe, the equivalent of assessing Harvard not by the performance of its graduates but rather by everyone who had applied to the university, including those who had been rejected. It is an apparently arcane dispute, but it changes the results markedly. NewGlobe says that, by including only pupils that Bridge intended to teach, its 23 schools improved outcomes at year three by 0.62 standard deviations.

NewGlobe points to an endorsement by Liberia’s current vice-president, Jewel Taylor, the ex-wife of convicted war criminal and former president Charles Taylor. At an event in August in the country’s capital, Monrovia, she commended the “tremendous efforts” of Bridge Liberia, including what she said was its encouragement of young mothers to return to school — now a national policy.

George Werner, the education minister between 2015 and 2018, was the driving force behind the experiment and a big believer in the use of technology and standardisation to improve teaching. He considers the programme — which has since survived the transition to a new government and has been significantly expanded — to be a success.

A teacher marking work using a notepad and a gadget
A teacher marking work at Bridge’s Upper Careysburg school, near Monrovia

“A gain is a gain in our context,” he says. Many of the outsourced schools were in districts with poor roads where teacher absenteeism was high. “Every step in the right direction is the right thing to do,” he says, adding that “the best operators produced remarkable results”.

“Part of what the RCT didn’t quite capture is the longer term prospects of having operators on the ground, strengthening the ecosystem,” says Rising Academies’ Skidmore. His company now runs 95 schools, up from five monitored during the RCT. Skidmore also points to his company’s swift response to Covid-19, when, within weeks of schools closing last year, it had devised a series of open-access radio programmes, called Rising on Air, that provided home-study options at five different grades.

Whatever the rancour around the Liberian experiment, the idea of outsourcing education to third-party operators has taken hold in the west African country and elsewhere. Werner dislikes the term outsourcing, preferring to describe the arrangement as a “partnership”.

African students sitting in a desk
Classes at the Kenlay school

NewGlobe, in particular, has gone from running schools to offering services to school authorities. In Nigeria, it has partnerships with the states of Edo and Lagos to support more than 2,000 public primary schools in a large scaling-up of its model. Added to schools in Kenya and Uganda, the company says it is helping to teach nearly 1m pupils.

Governments increasingly have come to the conclusion that “the only way to achieve transformational change at speed and scale is with non-state actors”, says Omowale David-Ashiru, managing director of NewGlobe Nigeria. Though the concept of outsourcing education remains controversial, it is gaining momentum.

This article has been amended to correct a typographical error.

Get alerts on Education when a new story is published

Copyright The Financial Times Limited 2022. All rights reserved.
Reuse this content (opens in new window) CommentsJump to comments section

Follow the topics in this article