Power and Progress

Between 1840 and 1900, real wages for unskilled British laborers rose by a staggering 120%. A century later, from 1980 to the present, real wages for American men without a high school diploma fell by over 20%. Both periods saw explosive technological innovation, from the steam engine and telegraph to the microchip and internet. Yet the outcomes for working people were polar opposites. One era translated technological marvels into widespread prosperity, lifting millions from poverty. The other concentrated the gains at the very top, leaving a significant portion of the population behind. This divergence presents a fundamental puzzle: if technology is the engine of progress, why doesn't it always benefit everyone?

This exact question has been the central focus of Daron Acemoglu and Simon Johnson's careers. As two of the world’s most influential economists—Acemoglu at MIT and Johnson, a former chief economist of the International Monetary Fund—they spent decades analyzing centuries of data, from medieval Europe to Silicon Valley. They noticed that the dominant narrative of technological inevitability didn't match the historical facts. They wrote Power and Progress to dismantle the myth that progress automatically creates a rising tide for all, arguing instead that the direction of technology and the distribution of its benefits are choices. The book is their definitive answer, showing how societal power structures determine whether innovation leads to shared prosperity or deepening inequality.

Module 1: The Illusion of Automatic Progress

The central myth the authors dismantle is the "productivity bandwagon." This is the classic economic idea that as technology makes companies more productive, everyone benefits. Higher productivity is supposed to lead to higher wages for all. But history shows this is rarely automatic.

A thousand years ago, medieval Europe saw major technological advances. Better plows, windmills, and new crop rotations increased agricultural output. But did the peasants who made up 90% of the population get richer? No. Their living standards stagnated or even declined. The surplus wealth from new technology was captured entirely by a small elite of lords and clergy. They used it to build massive cathedrals and expand their land holdings. The church, for example, came to own one-third of all agricultural land in England. Technology made the pie bigger, but only for those who already held power.

This pattern isn't just ancient history. Fast forward to the early British Industrial Revolution. New machines in textile factories created staggering wealth for their owners. But for nearly a century, workers saw no increase in their real income. They faced brutal fourteen-hour workdays in dangerous, polluted environments. Technological progress often serves to deskill and discipline labor. The Luddites, who famously smashed machines, weren't just anti-technology. They were protesting a system where automation was used to enrich a few at the direct expense of the many.

So what's the lesson here? The authors argue that technology is not a neutral force. Its development and deployment are shaped by power. When a small elite controls the vision for progress, they naturally direct it toward their own interests. This often means automation that replaces workers, surveillance that controls them, and systems that extract wealth without sharing it. Without a countervailing force, this is the default outcome.

Module 2: The Two Paths of Innovation

Now, let's explore a critical distinction the authors make. All technology is not the same. It can follow two very different paths. The first path is automation. This is technology designed to replace human labor. The second path involves creating new tasks. This is technology that augments human capabilities and creates new roles for people. The direction we choose has massive consequences for inequality.

For a clear example, let's compare two moments in manufacturing. In the early 20th century, the rise of the American automotive industry was a powerful engine of shared prosperity. Yes, there was automation on the assembly line. But the explosion of mass production also created a vast number of new tasks. These included roles for designers, engineers, machinists, clerical workers, and maintenance crews. This created strong demand for labor across many skill levels, pushing wages up.

But flip the coin. In recent decades, the introduction of industrial robots into manufacturing has followed a different script. These robots often automate tasks like welding or painting without creating enough new roles for the workers they displace. This is what the authors call "so-so automation." It might offer a small productivity bump for the company, but it leads directly to layoffs and regional economic decline. Think of self-checkout kiosks at the grocery store. They eliminate cashier jobs but provide minimal real benefit. The work is just shifted to the customer.

And here's the thing. The current direction of AI is heavily biased toward this "so-so automation" and surveillance. Instead of being a tool for human empowerment, it’s often used for control. AI-powered systems monitor warehouse workers' every move. They track delivery drivers' routes. They enable "flexible" scheduling that destroys income stability for service workers. This use of technology simply shifts rents from labor to capital. It’s a tool for extracting more effort for less pay. This isn't progress. It's a modern-day panopticon.