The average education budget in developing countries amounts to just $416.87 per child per year to cover everything from buildings to teacher salaries to books and administration. In the United States that budget is $12,296 per child per year. Meanwhile in Afghanistan the budget amounts to only $100 per year. As the majority of the world’s children live in developing countries, achieving more with less isn’t just desirable: it’s essential. Yet as I read about many innovative education projects that claim to improve education in such settings, I see very few that provide any information about cost let alone return on investment.
Research impact and then consider cost? WRONG!
If you’re designing an education project, you need to consider the budget constraints and marginal cost per student from the start. If a project demonstrates impact but requires expensive broadband satellite Internet connectivity, power, and computer labs then even if it demonstrates impact it’s A: more expensive than other ways to achieve the desired outcomes and B: it’s a waste of research money. Before a project is even considered the per student per year marginal cost should be calculated.
Innovation in education more often than not isn’t free; but when an “innovation” has no viable path to deliver a return on investment, it needs to be shelved before it’s even started. Upfront fixed costs like curriculum design are less of a problem than systematic per student costs like giving out devices (often not a good idea), infrastructure and training needs. These issues are far better caught at the design phase (e.g. by swapping giving out devices in favor of using people’s own devices – even if they are less sophisticated).
Those donors who fund innovation in education who want to see those innovations scale need to have a space in their application template for marginal cost per student. The sustainability section of a proposal needs to have some hard maths as well as a narrative.
Keep costs low by design
As mobile technology becomes more advanced, ubiquitous, and (very importantly) cheaper, it’s not surprising that many (but by no means all) education innovation projects involve some technology. When looking to improve reading fluency in Kenya, the PRIMR study found it was at least 6.5 times more cost effective to give tablets to teachers than it was to give e-readers to pupils. It was even more cost effective to give tablets to teacher tutors than to teachers themselves.
At Ustad Mobile, we develop software that runs on the devices that people already own: we still support the non-smartphone (feature phone) models that are owned by the majority of people in low income countries. It requires a small upfront investment in software development but costs a lot less than giving out and maintaining higher end devices. It also keeps our marginal cost per learner at less than $1 USD per year.
Whether or not the intervention involves tech, it’s important to keep things as simple as possible to keep training costs as low as possible.
Consider behind the scenes innovation
Many of the edtech photos you see have a picture of a learner (often in an austere setting) in front of a mobile device of some kind. Yet some of the best return on investment can be found behind the scenes. Without accurate data on student numbers, how can you distribute the right number of books to schools? Without a tracking system, how many children will be left without the books that they need? Without feedback systems, how will parents and teachers report and resolve issues? It might not make for as great a photograph as “frontline” innovations, but it can often deliver incredible bang for the buck.
Innovate to fight waste and fraud
Teacher absenteeism in primary schools in India costs $1.5 billion per year with a 23.6% absenteeism rate in rural schools. In Balochistan province in Pakistan it was found that there were “no records” of 15,000 teachers, 900 schools and there were 300,000 fake student registrations. Such waste and fraud disproportionately affects those in low income and more rural areas. To improve equity in education, waste and fraud must be tackled. At Ustad Mobile, we even started integrating attendance monitoring with photos (and soon fingerprint scanning) into our app that originally was focused on the delivery of learning content.
Open source it
The digital development principles endorsed by almost all the major donor agencies rightly advise that software, educational content and so forth should be made available under open licenses. Such licenses allow for anyone to remix their own versions of content and software. It doesn’t mean all suggested changes have to be accepted: it only means people must be free to create their own adaptations. This drastically reduces the barriers to innovation allowing others to freely build on what went before rather than having to reinvent the wheel.
It’s not a particularly useful question to ask “can this innovation improve learning?” on it’s own. Even if it can, it’s ultimately useless unless it can improve learning AND do so better and more cost effectively than what came before. Ultimately every dollar in education comes from somewhere: spending money on an innovation that has no path to provide a viable return on investment for the context that it’s intended for isn’t just wasteful: it’s destructively depriving both conventional programs such as teacher training that work and scalable innovations of the funds they need to improve education outcomes.
This article was written by Mr. Michael Dawson, CEO of the 2014-15 WISE Accelerator project Ustad Mobile.