There’s a story that circulates in Addis the way all modern myths do, fast, emotional, slightly exaggerated, and always just believable enough to stick.
A 19-year-old drops out after Grade 10, picks up video editing on a cracked phone, starts posting reels from somewhere around Bole, and within a year is making more money than a fresh economics graduate who spent four years at Addis Ababa University and now sells airtime near Mexico Square. One has a degree. The other has distribution. One has transcripts. The other has attention.
And in the age of TikTok, AI tools, and overnight virality, the conclusion feels obvious: education is losing relevance.
Except it isn’t. The data refuses to cooperate with the story.
What is actually happening is more uncomfortable. Education is not becoming less important in the digital economy. It is becoming more important, more valuable, and more decisive than at any point in modern economic history. The only thing that has changed is that bad education systems now expose themselves faster.
The myth is loud. The evidence is louder.
Across global labour markets, economists have been measuring the returns to schooling for decades. The most comprehensive global synthesis, drawing from over a thousand estimates across more than a hundred countries—shows a consistent pattern: each additional year of schooling increases earnings by about 9 percent on average. That figure is not collapsing in the digital age. It has stayed remarkably stable through the transition from industrial economies to software-driven ones.
But the geography matters. In Sub-Saharan Africa, the return rises sharply to around 13.5 percent per year of schooling. That makes education one of the highest-return “investments” available to individuals in the region. For women, the returns are often even higher, with measurable gains in wages, health outcomes, and long-term household resilience.
Even university education, the most criticized layer in today’s “dropout discourse” has not lost relevance. In fact, returns to tertiary education have increased over time, rising from roughly 13 percent in the 1980s to over 17 percent in the early 2000s. The more complex the economy becomes, the more valuable structured learning becomes.
This is where theory catches up with reality. Endogenous Growth Theory argues that economic growth is not just driven by machines, roads, or capital accumulation. It is driven by knowledge itself. Ideas generate more ideas. And ideas do not depreciate when copied, they scale. A single well-trained mind can produce output that replicates across millions of users at near-zero marginal cost in a digital economy.
So education does two things at once. It raises individual productivity and it expands the frontier of what an economy can produce. One affects wages. The other affects national growth. The digital economy did not weaken this mechanism. It amplified it.
If anything, technology has made the labour market more selective, not less.
The fear that AI replaces skilled labour has become a popular narrative, but the empirical direction is more precise. What is happening is skill-biased technological change on steroids. Technology does not remove the need for human capability. It concentrates value around it.
Recent global labour data tracking close to a billion job postings shows that workers with AI-related skills now earn wage premiums of around 56 percent compared to their peers. That premium was roughly half that only a year earlier. In other words, the value of skill is not declining, it is accelerating.
At the same time, jobs most exposed to AI are not shrinking. They are expanding in complexity and productivity. Exposure to AI has been associated with faster job growth in those roles, rising demand for AI-specific skills, and productivity gains that significantly outpace less exposed sectors. In many industries, revenue per employee has accelerated multiple times over.
The pattern is not replacement. It is intensification.
The bar is rising faster than the floor is falling.
And that leads directly to the creator economy, which is often presented as the ultimate proof that formal education no longer matters. A teenager with a smartphone and a ring light can, in theory, outperform a graduate with a degree.
The theory is partially true. The reality is brutally uneven.
The global creator economy is now worth nearly $200 billion and could double within a few years. Over 200 million people identify as creators. But the income distribution tells a very different story than the social media fantasy.
Roughly half of creators earn less than $15,000 annually. Only a small fraction, around 4 percent, cross the $100,000 threshold. Most people who enter the space never build stable income streams. Many take months just to earn their first dollar. A small elite captures the majority of revenue.
What looks like democratized opportunity is structurally closer to a winner-take-most attention market.
And the winners are not unskilled. They are just differently skilled.
Successful creators understand audience psychology, content strategy, pricing, analytics, negotiation, branding, storytelling, and increasingly AI-assisted production systems. They are not outside the knowledge economy. They are inside it, just wearing a different interface.
The platform gives distribution. It does not give judgment. It does not give discipline. It does not give economic reasoning or narrative structure or financial literacy. Those are still learned capabilities. Whether in classrooms or through self-directed learning, the pattern does not change: sustained success requires accumulated human capital.
The dropout did not escape education. He found another version of it.
The illusion becomes even more dangerous in contexts where education itself is visibly failing.
Ethiopia is one of the clearest examples of this paradox. The country has expanded access to schooling dramatically over the past two decades. Primary enrolment is now close to 89 percent, and secondary enrolment has surged into the millions. On paper, the system has expanded. In practice, learning outcomes have collapsed.
International assessments estimate that around 90 percent of ten-year-olds cannot read and understand a simple text. This is not an education system producing low performance. It is a system producing near-zero foundational literacy at scale.
The consequences appear most visibly in the national Grade 12 examinations.
A decade ago, roughly half of students could pass the threshold for university entry. Today, that figure has fallen to single digits, with recent years showing pass rates between 3 and 8 percent. In some subjects, performance has collapsed from above 80 percent success to around 5 percent. In certain years, entire schools report zero students passing.
The inequality inside the system is just as revealing. Students in elite or international schools pass at rates above 80 percent. Urban private schools sit around the middle. Government schools, which educate the overwhelming majority of students, struggle in the low single digits.
What this produces is not just an education crisis. It produces a credibility crisis.
When degrees stop translating into skills, and skills stop translating into jobs, the public does not reject the labour market, it rejects education itself. That is where the dropout narrative becomes emotionally persuasive. It is not built on theory. It is built on lived disappointment.
But the distinction matters.
The failure is not that education has lost value. The failure is that meaningful education is not being delivered at scale. Underfunded classrooms, large student-to-teacher ratios, weak training systems, outdated curricula, and low investment relative to global benchmarks all converge into one outcome: schooling without learning.
The World Bank’s Human Capital Index captures this in a single number. A child born in Ethiopia today is expected to reach only about 38 percent of her productive potential under current health and education conditions. That means more than half of potential lifetime productivity is lost before adulthood even begins.
That is not a labour market failure. It is a foundation failure.
And it explains the economic ceiling the country keeps running into. Infrastructure projects, industrial parks, and mega-dams can shift supply constraints, but they cannot substitute for human capability. Without skilled labour, the returns to physical capital eventually flatten.
This is the central prediction of endogenous growth thinking: capital without knowledge hits diminishing returns quickly. Knowledge does not.
So the real tension in Ethiopia is not between education and the digital economy. It is between weak education systems and the demands of a digital economy that is becoming more skill-intensive every year.
AI does not reduce this pressure. It increases it. The creator economy does not remove the need for learning. It hides it behind new labels. The labour market does not reward credentials. It rewards capability.
And capability is still built the same way it has always been built: through sustained learning, structured thinking, repetition, exposure, and adaptation.
The modern economy has not made education optional. It has made it unstable, fast-moving, and unforgiving of low-quality learning.
Which leads to the final inversion of the original myth.
The dropout story feels like a sign that education is irrelevant. In reality, it is a sign that informal learning is trying to replace formal systems in environments where those systems are failing. The successful dropout is not evidence against education. He is evidence of its necessity in a different form.
And the struggling graduate is not evidence that education is useless. He is evidence that credentials without competence are becoming economically invisible.
The digital age did not weaken the value of education.
It exposed the cost of getting it wrong.
