The Automation Paradox: When Technology Is Designed to Keep You Cheap
The Question Everyone Stopped Asking
For more than four decades, American workers have watched robots, software, and algorithms reshape their workplaces while their paychecks stayed stubbornly flat. Economists have long treated this as a puzzle: why haven't productivity gains from technology translated into wage growth? But MIT researchers Daron Acemoglu and Pascual Restrepo have reframed the question entirely, according to their study published in the Quarterly Journal of Economics. What if it's not a puzzle at all? What if firms are deploying automation specifically because it suppresses wages, not in spite of that effect?
The research, titled "Automation and Rent Dissipation: Implications for Wages, Inequality, and Productivity," exposes a mechanism that's been hiding in plain sight (Quarterly Journal of Economics). Companies aren't just automating to boost output. They're automating to strip workers of bargaining power, ensuring that any productivity gains flow upward to shareholders rather than outward to paychecks. The technical term is "rent dissipation," but the human translation is simpler: technology designed to make you replaceable, not productive.
This isn't speculation about corporate motives. It's a documented pattern in how firms choose which tasks to automate and when, per the MIT researchers' findings. The strategy works because it targets a specific vulnerability: your ability to ask for more money when you become more valuable to your employer.
How the Wage Suppression Machine Actually Works
The mechanism operates through precision, not wholesale replacement. Firms don't need to eliminate your job entirely to eliminate your leverage. They just need to automate enough of your tasks that you become easier to replace if you demand higher wages. A warehouse worker still picks items from shelves, but now a algorithm determines the route, the pace, and the priority, making the human component interchangeable. A call center employee still handles complex customer issues, but chatbots have absorbed the routine questions that used to build expertise and seniority.
The research by Acemoglu, who serves as Faculty Director at MIT's Shaping the Future of Work initiative, and Restrepo, an Associate Professor of Economics at Yale University and Research Affiliate at the Stone Center on Inequality, documents how this pattern has played out across industries (fact bank cache). The original version of their work appeared in May 2024 under the title "Labor Market Consequences of Technological Change" before being updated and sharpened in November 2025 to emphasize the rent dissipation framework (fact bank cache).
What makes this particularly insidious is that it creates a productivity paradox by design. The technology does make operations more efficient in aggregate. But because workers can't capture any of those gains through higher wages, they have no incentive to embrace the changes or invest in complementary skills. The result: firms get cheaper labor but not the productivity explosion that genuine innovation should deliver. Everyone loses except shareholders in the short term, and eventually even shareholders lose when demand collapses because workers can't afford to buy what they produce.
The Pattern Across the Economy
This wage suppression strategy explains patterns that have baffled observers for years. Consider healthcare, where employment has grown even as other sectors automated. Those jobs persist not because healthcare resists technology, but because direct care work is harder to automate in ways that suppress wages without eliminating the position entirely. You can't make a nurse 10% more replaceable; you either need the nurse or you don't. The strategy reveals itself through its absence.
Or look at financial markets, where firms have spent billions chasing microsecond advantages in trading speed. When you've already squeezed labor costs to the floor through automation-driven wage suppression, capital seeks returns wherever it can find them, even in advantages measured in millionths of a second. It's the same extractive logic, just applied to a different bottleneck.
The geographic concentration of this phenomenon compounds its effects. Automation-driven wage suppression doesn't distribute evenly across regions. It concentrates in specific communities, creating pockets where entire generations watch their earning potential flatline regardless of how hard they work or how much the local economy grows. The technology becomes a mechanism for engineering inequality into the economic geography itself.
What This Reveals About the System
The Acemoglu-Restrepo research forces an uncomfortable reckoning with how we think about technological progress. We've been treating automation as innovation, subject to light-touch regulation designed to encourage advancement. But if firms are deploying technology primarily to control wages rather than boost productivity, we're regulating the wrong thing. This isn't innovation policy; it's labor policy disguised as technological inevitability.
The implications extend beyond individual firms or sectors. If wage suppression through automation is a deliberate strategy rather than an unfortunate side effect, then the productivity gains we've been promised aren't missing. They're being captured, just not by workers. The technology works exactly as designed. The inequality isn't a bug; it's the feature that makes the investment worthwhile for firms.
This reframing also explains why conventional policy responses have failed. Tax credits for automation investment, retraining programs for displaced workers, and education initiatives all assume the problem is workers lacking skills for a changing economy. But if the actual mechanism is firms using technology to eliminate worker bargaining power, then no amount of retraining will restore wage growth. You can't skill your way out of a system designed to ensure skills don't translate into leverage.
The Uncomfortable Truth
What makes the MIT researchers' findings so significant is that they name the thing we've been dancing around for decades. Stagnant wages amid technological advancement isn't a mystery to be solved through better education or more innovation. It's the logical outcome of firms choosing automation strategies that prioritize wage suppression over productivity gains. The system is working exactly as intended, just not for workers.
The question now is whether policymakers will continue treating automation as neutral technological progress or start scrutinizing it as a labor practice with distributional consequences. Because once you see the mechanism, you can't unsee it. Every new automation announcement becomes a question: is this technology making workers more productive, or just more replaceable? And if it's the latter, why are we subsidizing it?