Lant Pritchett on 5 Ways We Can Learn From Biology to Shift Education Systems
An intro to “The Rebirth of Education” — the most important book on why ed reforms in the Global South fail and what we can do about it.
Want EdWell’s articles straight to your inbox — with jobs & events? Sign up for our newsletter!
Did you know that most of our efforts to improve schools have failed?
That even though we know a lot about what works to teach kids in the Global South — and spend billions — our schools are still pretty much staying the same, year after year?
I didn’t. Until I read The Rebirth of Education: Schooling Ain’t Learning. This one book by the economist Lant Pritchett showed me just how much all our hard work and innovations are maintaining the status quo.
I meet so many people trying to lead or fund change in education systems across Latin America, Africa, or South Asia. I recommend they read this book, hoping they can learn from all those mistakes that others made. I even co-founded an organization, Metis, because of it. Pritchett gives us a powerful vision for a better way.
But almost no one actually reads it. It’s a dense 240 pages of charts and statistics — not what you want to dig into after a long day at work.
So I’m going to make it easy. Here is a cheat sheet with the five concepts you need to know: spiders, mimicry, monoculture, evolution, and starfish. Take 13 minutes and learn from Pritchett’s advice.
To explore Pritchett’s full argument, read this brief by the Center for Global Development, download a free chapter, or buy the book ($11).
1. Spiders
We define success as INPUTS, because they are easier to track. We should shift our focus to learning OUTCOMES.
The Push for Access
Typically, we define successful education reforms as ones that increase inputs — changes that can be paid for or easily counted. These include:
- Infrastructure like classrooms, desks, or toilets
- Learning materials like textbooks, chalkboards, or paper
- # of teachers or class size
- # of hours teachers are trained
- Amount for teachers’ salaries or the Ministry of Education’s budget
- # of students enrolled in school
The last item is most important here, because since 1948, most countries in the Global South focused on increasing access. Free primary education was our main goal, and this had results: by 2015, 91% of children worldwide were in school. This was a huge task for many countries; think about how difficult it was for a country emerging from colonization or civil war, to build schools and train teachers for their entire population.
Our Learning Crisis
However, Pritchett argues that even though many more students are in school, MOST ARE NOT LEARNING ACTUAL SKILLS. (In caps because of how surprising and important this message is!) He cites plenty of evidence for this including:
- Tests that compare countries: TIMSS shows that learning for science/math skills “is slow (or moving backward)” (92) — by grade 8 students from 1999–2007, in countries like Indonesia, Malaysia, and Iran.
- Tests that compare change within a country: ASER in India showed that in 2008, “only 30 percent of children in grade four could read a simple passage” (28). Uwezo found that in Kenya from 2012–2015, even though there was nearly universal primary enrollment, there was no significant increase in learning outcomes across basic literacy and numeracy.
As he writes, “what does not get measured does not get done” (19). We measured inputs but not learning. Thus, insight 1: increased inputs do not necessarily cause improved learning.
Centralized Systems
But there are two reasons why we focus too much on measurable inputs. The first is simple: measuring quality is much more difficult.
The second is more complex and it has to do with how education systems are like spiders. Sounds strange, right? But get this: although they create a massive web, spiders can only process limited information through their one tiny brain. As Pritchett writes, “all information created by the vibrations of the web must be processed, decisions made, and actions taken by one spider brain at the center of the web” (5).
Education systems are the same way. They are top-down bureaucracies where a small group of experts at the top (typically a central HQ at the national or state/provincial level), control what happens in the rest of the system. Power is highly centralized, which is great for tracking logistics, but becomes tricky if local principals need the power to change things at their school that they know can help students.
Pritchett goes off on a fascinating tangent about why education systems became this way (because schools were a tool for postcolonial governments to indoctrinate their young citizens with a sense of nationalism, so that they wouldn’t revolt or protest the ruling party). For more, read the book :)
2. Mimicry
We COPY what seems to work for other countries. But we should decide what works best for our context based on EVIDENCE.
Tested Solutions
Okay, so now that we know our systems are broken, what should we do? If we are wasting so much of our money, as Pritchett argues, what should we spend our money on?
What has been proven to work. Interventions that have been evaluated by researchers and proven to improve learning outcomes.
Pritchett says that the greatest shifts in learning outcomes come from “organizational and systemic changes that change the scope of actions, incentives, and accountability” of people within the system (122). In non-professor terms, this means things that change how teachers operate, or that create pressure on a school to improve instead of delivering the same mediocre lessons year after year. Evidence supports interventions like:
- Community-hired or locally-contracted teachers
- Low-cost tutors to supplement teachers
- Cameras to capture student attendance
- Regulation that permits low-cost private schools to operate, so that parents have options and there is pressure on schools to improve
- Radio-based instruction
But there are two key reasons why systems do not adopt these kinds of innovations. This is where I’m going to throw some fancy science concepts at you.
Isomorphic Mimicry
Because of this principle, governments often do what other governments did because they assume it will work. But it does not always work in their local context. Education systems copy what seems to impact learning, rather than what is actually effective. They follow what Pritchett calls “the educational fads of rich countries” (90). Imagine — Limit class sizes to 20 children. Distribute laptops to all students. Re-paint classroom walls so they look nicer. Etc. Pritchett says this is a problem because we’re copying systems that are at a fundamentally different stage of development — aiming for incremental change rather than huge gains:
“No living education expert from the West has had experience implementing programs…that led to the type of massive learning improvements that developing countries aim to achieve today. Instead…[they] inherited systems with high learning performance. Therefore, their professional experience is in operating or improving functional, high-performing systems, not in building them” (116).
Camouflage
This one is more sinister. If leaders adopt logistical changes like improving infrastructure, they can hide or distract from the fact that they are not changing the more important parts of their system (the ones that will actually change whether students are learning). As Pritchett writes, “The real purpose of reform efforts is to create certain appearances to legitimate failing and flailing systems, without making demands or threatening existing political interests” (135). As I said before, measuring learning quality is hard. Changing a system is HARD. You’ve got powerful teachers unions, corruption, patronage-based appointments, all kinds of entrenched interests that make it much easier to just sit still, so that you don’t make anyone angry that you’re challenging their power.
3. Monoculture
We impose ONE model across a system. We need to allow leaders to test DIVERSE solutions.
One-Size-Fits-All
To have quality control, what most public systems do is require that all schools look the same. They mandate that principals use the same curriculum, teacher training, and schedule. This is easy to enforce, because you can send around inspectors and all they have to do is go through the same checklist as they observe each school. It works well for a spider system, because they report back to a centralized Minister who ensures that every school is following the rules.
But to understand why this is a problem, we turn to agriculture. Pritchett teaches that sameness in a school system is similar to monoculture in fields or forests. This is when a farmer plants a single species and eliminates all others from their land.
The Problem With Monoculture
Although this may look pleasing to the eye (see those tidy rows in the photo above), it is actually pretty bad for whatever you’re growing. In the short-term, say a 1-year growing season, you’ll be fine. But after multiple times of planting and picking your crop, your crop will become lower quality. It will be more susceptible to pests, more likely to be damaged in storms or floods — and less healthy due to fewer nutrients in its soil.
You see, there is a reason that nature created forests in all their messy glory. A diversity of species in a system makes every species stronger in the long-run.
The Value of Diversity
When we apply this lens to education systems, we see that a system needs a diversity of schooling approaches to be healthy, too. Basically, teachers need freedom to test and discover new innovations that have the potential to improve learning. If they don’t have freedom, and are required to conform to a one-size-fits-all model, they will not improve how they teach, and students will not learn more.
As Pritchett argues, systems that are centralized like spiders and managed like a monoculture field, “can block great ideas from scaling” (202). With these concepts, Pritchett builds on the excellent book Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed. James Scott shows that when a government tries to impose control over society through a central plan, they end up making things worse. They lack knowledge about what is actually needed locally and how conditions are changing (see: the planned city of Brasília, Soviet collectivized farms, etc).
Our goal here, as Pritchett says, is to transform our systems from the “uniform implementation of a top-down program…into systems that free, foster, and scale innovations and dynamism” (194).
4. Evolution
We should focus more on holding schools ACCOUNTABLE to outcomes, and less on restricting the methods they use to get there.
Pressure From Predators
To make the next layer of Pritchett’s argument easier to grasp, I’ll use a term most of us are familiar with from our science classes: evolution. Remember that this is the process by which a species evolves — over generations, they shed or gain certain characteristics based on what works best for their species to survive.
To teach you about this, I’ll use another animal metaphor (what fun!). Pritchett explains:
“Suppose we wanted to increase the average speed of things that swim in a given ecosystem. One might set about to genetically engineer the perfect swimmer. Alternatively, one might just get more sharks in the water…Those that can’t swim fast get eaten and those that don’t get eaten reproduce…millions of choices of individuals in the system: lots of swimmers doing different things… just enough sharks in the water to increase the pressure” (10).
To break this down, what he’s saying is that an education system is the ocean. Schools are the swimming fish. And we need something to play the role of a shark. That leads us to:
Weeding Out Low-Performing Schools
Governments can be the shark. Imagine: government leaders choose a core set of skills that are most important for their students to learn. They design assessments to measure progress on those skills. Students across the system take them regularly. As a result, everyone knows to what extent each school is enabling students to learn.
There are incentives for schools to improve — if they do well, they get more funding/ classrooms/ teachers to grow so that they can accept more students. But schools are also under pressure and are held accountable if they fail. If they are clearly not teaching their students, they are closed. Their space is given to another school that is doing a better job.
Keep in mind this is very much an economist’s way of looking at a system — it’s what they call creative destruction. In this world, there are many different types of schools all iterating and evolving, ineffective models are weeded out, and the system as a whole improves faster than it would have if all schools were forced to follow one prescribed method.
The Next Frontier: Measuring Quality
Now this sounds great in theory, but when systems try to implement it, it gets complicated. Pritchett does not give us a clear answer on what the accountability structure should look like; he says that he offers a vision and it is up to local leaders to adapt it to their specific context. But in my opinion, the biggest challenge for all of us in implementing his recommendations is figuring out how to measure quality and hold schools accountable. There are two thorny issues we need to face:
- Assessing students has too often looked like high-stakes standardized tests, and we’ve seen the problems and limits of these. In a future where robots take over many tasks, the skills that are increasingly valuable are those that robots cannot take over. Skills for interacting with humans: managing a team, starting a business, traits like empathy, grit, etc. Right now our tests are adequate at measuring basic skills like literacy, numeracy, and information that can be memorized. How do we build better ways to measure what matters at scale?
- When an accountability framework like this was implemented in my own country, the United States, it didn’t account for the fact that schools deemed ‘failures’ were often working with students who had the lowest levels of skills in the first place. We increasingly know that teachers are only one factor that shapes whether a student learns. The equal, if not more important, factors are how many words they heard as a toddler, whether they have enough to eat, have access to healthcare for things like glasses, are getting enough sleep, or are experiencing violence, homelessness, or psychological trauma at home. How do we support schools to overcome these challenges to enable learning, instead of punishing principals/teachers who choose to work with the most marginalized students?
I believe Pritchett’s argument about accountability is still important — but there is a long road ahead as we learn how to implement it.
5. Starfish
There are six traits we can use so that our system will ACCELERATE schools to improve over time.
Decentralized Control
Here’s where we finally find a better way than being like a spider. Being like a starfish. As Pritchett explains, starfish have no brain. They are a decentralized organism with a nervous system throughout their arms; the arms each make decisions about where to go, and that adds up to the whole starfish moving. If one arm is cut off, they can regenerate another arm. Cool, right?
We need education systems to look less like spiders, and more like starfish.
The Playbook
Now for my favorite part, where we finally get to the most important page in the book! Pritchett gives us a vision for how to create a starfish education system. It has six traits for where we want to go (center column) and for where we are right now (right column). This is the whole book summarized in one chart:
Translated Into Action Steps
In practice, this could mean things like:
- Open: New schools are allowed to open and serve students, such as charter schools (publicly financed with more autonomy than traditional public/government schools), private schools, or community-controlled schools started by parents. Parents have options between multiple public/government schools to choose from and have the power to decide where to send their children.
- Locally operated: Schools are allowed to choose their teachers and fire them if they are incompetent.
- Performance pressured: Standardized assessments measure skills students have learned, so that progress is visible and it is easy to see whether a school is improving over time. Schools’ results from tests (without student names) are available to be viewed by local, province/state, and national governments, along with journalists, researchers, and parents.
- Technically supported: Schools have autonomy to identify their training needs and choose from a range of options for training their teachers.
- Professionally networked: Teachers use networks with other teachers to observe classrooms in other schools, learn from other teachers, and share practices that strengthen learning.
And with that, I hope you go try to create some of these traits for your system. Good luck!
Kat Pattillo writes about education reform and innovation across the Global South. She is currently studying how social movements accelerate systems change in education, through an MPhil in Politics at Oxford. Kat previously consulted as a researcher and facilitator, taught at African Leadership Academy in South Africa, and co-founded Metis in Kenya. For more of her writing, follow her on LinkedIn or sign up for the EdWell newsletter.
