On Steven Lukes's three layers of power
Power isn't a topic people bring up easily. It can feel too political, too uncomfortable, like something that belongs to governments and institutions and not to ordinary life. So we file it away: over there, not here.
Recently I read Steven Lukes's Power: A Radical View, and it gave language to something I'd been sensing for years. Power isn't over there. Power is like air. We breathe it. We live inside it. It's everywhere, including the places we assume are free of it.
Lukes's argument is simple and unsettling: power operates in layers. First, in who wins. Second, in what gets discussed. Third, in what people come to want.
The first layer is the one everyone recognizes. There is a conflict, a decision is made, someone gets their way.
Lukes's point is that if this is all you look for, you will systematically underestimate power. The most effective forms rarely appear at the decision-making table. They've already set the table. They chose the chairs, wrote the menu, and decided who gets invited. By the time a visible conflict erupts, the most important work has already been done.
Power also operates by shaping what counts as a legitimate topic, ensuring certain questions never reach the agenda. Lukes calls this the power of non-decision: the ability to keep issues from becoming contestable in the first place.
I think about this whenever immigration comes up in American politics. The debate is reliably framed as "legal versus illegal." But the prior question, why legal immigration can take decades, rarely becomes a political issue at all. I know people, educated and employed, who have waited fifteen or twenty years for a green card. For some countries, the backlog stretches beyond a human lifetime. Whatever the intention behind the system, the effect is structural. Delay becomes normal, and the question of fairness is treated as background noise rather than a problem that demands a solution. It's not on the ballot. It's not framed as an injustice. It is, in practice, off the agenda.
This second layer greatly expanded the study of power. But Lukes thought it still didn't go far enough. The thinkers who developed it still relied on conflict. They widened the lens from open conflict to hidden conflict, but they assumed excluded groups at least had unexpressed grievances.
What if the grievance itself never forms?
This is the layer Lukes calls radical, and the reason the book matters.
The first two layers assume conflict, open or suppressed, is necessary for power to exist. If people consent, we assume power is absent. Lukes argues the opposite: the most effective use of power is to prevent conflict from arising at all. It shapes perception and preference so that demands never take shape.
One thing worth pausing on is that power over, the form of power that operates between people and shapes who dominates and who defers, does not require bad intentions. Often it arrives wrapped in care.
I once sat in on a dharma talk at my Zen center. Afterward, a mother asked the teacher for advice. Her daughter refused to practice piano, and it had become a constant battle. The mother was exhausted. She genuinely believed piano would benefit her daughter. The daughter genuinely wanted to stop. What I noticed was the framing. The mother never asked whether her daughter should play piano. She asked how to make her daughter play piano. The child's own judgment about her life wasn't part of the equation. This wasn't cruelty. It was love. And it was power.
I grew up hearing a version of this at a national scale. When people in China justified the Great Firewall, the system that blocks access to most foreign websites, the language often sounded parental: we aren't ready; outside influence would be destabilizing; people need protection. Whether or not every actor believed it, the structure is familiar—someone decides what others can be exposed to, and that decision is narrated as care.
This is part of what makes power over hard to see. We look for it in coercion, in cruelty, in visible domination. Much of it works through people who believe they know what is best for others, and whose position allows that belief to harden into the reality everyone else has to live in.
Lukes pushes further. At this depth, power works by making people accept their place in an order that may not serve them. It shapes what they want, what they believe they deserve, what they consider possible. In its strongest form, it can lead people to act against their own interests without ever experiencing that as a violation, because the sense of violation never fully emerges.
His example comes from Matthew Crenson's study of air pollution in Gary, Indiana. U.S. Steel was the city's largest employer. The factory polluted the air for decades. Residents got sick. Yet no meaningful movement formed. No sustained organizing, no pressure campaign, no city-council confrontation. Economic dependence didn't just constrain people materially. It shaped the town's mental landscape. "Air pollution as a problem worth fighting" struggled to crystallize. The company didn't need to suppress dissent. Dissent didn't reliably arise.
I felt a quieter version of this during a video call with my mother during the pandemic. She was in China. She told me not to go outside because things were chaotic. I said we weren't allowed to go outside. She replied, almost casually: "It's better when the government is strict about these things." It landed as common sense, obvious, reasonable, even comforting. That is the heart of this layer. When preferences have been shaped deeply enough, conflict is extinguished before it is born.
Of course, "real interests" is contested. James Scott's idea of the "hidden transcript" complicates Lukes's picture. Dominated people may comply outwardly while privately sustaining narratives of resistance. Apparent consent can mask a sharper awareness than observers can access. The third layer is messier than Lukes sometimes implies. Still, his lens forces a question that is otherwise too easy to skip: when we see consensus, is it genuine, or has it been manufactured?
There may be a reason people don't talk about power much anymore. Power hasn't weakened. It has gotten better at disappearing.
First-layer power is crude. Someone wins, someone loses, everyone can see it. Visible power provokes resistance. It can be protested, challenged, voted out. So power evolves. It migrates into the second layer: agenda-setting, rule-writing, quietly defining what counts as "reasonable," what counts as "serious," what counts as "unrealistic." Even that can be exposed. You can point to a rule and ask who wrote it.
The third layer is where power becomes nearly frictionless. It shapes what you think rules should be. It shapes what you experience as choice. It shapes what feels embarrassing to say out loud.
It can look like surveillance normalized as convenience, carried voluntarily in our pockets. It can look like algorithmic curation. A system decides what you see, and slowly trains your sense of what matters. In education and work, it can look like the system has defined what "qualified" means.
The quieter power gets, the deeper it reaches. And the deeper it reaches, the less it feels like power at all. It feels like reality.
That feeling, the way things are, is where the third layer lives. One way to test it is to ask two questions, gently but repeatedly. What is missing from the agenda? And which options have become unthinkable, so unthinkable they don't even feel like options?