Relative Development and the Intelligence Divide: Human Capital, Technology Diffusion, and AI

For decades, economists have watched the gap between the world's richest and poorest nations, asking a simple but stubborn question: why do some countries stay poor while others pull ahead? The answer has often seemed to lie in what a country owns. Nations with more factories, better roads, and more schools generally produce more wealth. But a deeper mystery remains. Two countries can have similar amounts of machinery and similar numbers of educated workers, yet one produces far more than the other. This difference is often called productivity, a measure of how well a society turns its resources into useful things. For a long time, the hope was that technology would act as a great equalizer. If a poor country could simply buy the same machines or software as a rich one, it should be able to catch up. Yet history shows that while countries have become more alike in their schools and factories, they have not become nearly as alike in how much they produce. The tools have spread, but the ability to use them effectively has not.

This puzzle is now at the center of a new debate about artificial intelligence. As powerful new computer systems become available, many wonder if they will finally help poorer nations close the gap. A new study by Patrick A. Imam and Jonathan R. W. Temple, researchers at the International Monetary Fund and an independent economist, suggests the answer is not as simple as access. The researchers argue that having the technology is not the same as having the capacity to use it. They set out to understand why some countries manage to move from low productivity to high productivity while others remain stuck, and whether artificial intelligence will help or hinder that journey. Their work combines a detailed look at historical data with a fresh way of thinking about how nations develop, moving beyond simple averages to track the actual movement of countries over time.

The researchers examined over a thousand five-year periods for nearly a hundred countries, tracking how their positions in the global economy changed. They looked at three main things: the amount of physical capital a country had, the level of education of its workforce, and its overall productivity. They found a striking difference in how these factors moved. Countries were able to increase their capital and improve their schooling relatively quickly. A nation could build a new factory or send more children to school within a few years. However, moving from low productivity to high productivity was much harder and much slower. Even when countries added more machines and educated more workers, they often failed to improve how efficiently they used them. The gap in productivity proved to be far more stubborn than the gap in resources.

To understand why, the authors looked closely at the role of human capital, or the skills and knowledge of the workforce. Standard economic accounts suggest that education explains only a small part of why some countries are richer than others. If you simply equalized the number of years of schooling across the world, the gap in income would shrink, but a large distance would remain. The researchers found that this static view misses a crucial dynamic. While education might not explain the current level of wealth perfectly, it is a powerful predictor of whether a country will move up or stay down. They discovered that countries with higher levels of measured human capital were significantly more likely to escape the lowest productivity states.

The study identified a specific threshold in human capital that seemed to separate two different worlds. For countries below this level, the average time to leave the lowest productivity state was about sixty-two years. For countries above it, that time dropped to just twenty-five years. This difference is not just a number; it represents generations of development. A country stuck in the low-productivity trap might wait a lifetime to see its economy transform, while a country with slightly more capable workers could achieve the same shift in a single generation. The researchers tested this finding in many ways, using different measures of skill and different ways of grouping the data, and the pattern held true. Even when they looked at cognitive skills rather than just years of schooling, the countries with higher skills moved faster.

This leads to a sobering conclusion about artificial intelligence. The technology itself is not a magic wand that will automatically lift all boats. If artificial intelligence mainly helps those who are already skilled, it could widen the divide. A country with a strong base of educated workers, reliable data, and capable firms will be able to use these new tools to become even more productive. A country lacking those foundations might get access to the same tools but fail to integrate them into its economy. The researchers argue that for artificial intelligence to help poorer nations catch up, it must do more than just lower the cost of knowledge. It must lower the cost of learning, adapting, and implementing that knowledge in places where the supporting systems are weak.

The study suggests that the real barrier is not a lack of technology, but a lack of the complementary capabilities needed to make that technology work. It is not enough to have a tractor; you need the fuel, the repair shops, and the farmers who know how to use it. Similarly, having artificial intelligence is not enough if a country lacks the management, infrastructure, and institutional strength to turn it into real productivity gains. The researchers found that the ability to convert cheaper intelligence into actual economic movement is the key. If artificial intelligence can help weaker economies learn faster and adapt better, it could shorten the decades-long wait for development. But if it only rewards those who are already strong, it may simply push the frontier further away, leaving the gap between the rich and the poor unchanged.

The findings offer a clear path forward for policymakers. Instead of focusing solely on adopting the latest technology, nations need to invest in the broader ecosystem that allows technology to thrive. This includes improving the quality of education, strengthening management practices, building reliable infrastructure, and ensuring that institutions can support innovation. The study does not promise that this will be easy or quick, but it clarifies the problem. The divide is not just about who has the tools, but who has the capacity to use them. For artificial intelligence to be a force for equality, it must be paired with a genuine effort to build the human and institutional foundations that turn potential into progress. Without that, the promise of a connected world may remain just that—a promise, while the reality of the intelligence divide continues to grow.