Why Reform Fails

The reform trap

Every few years the same cycle runs. Something terrible happens. Public outcry follows. Politicians promise reform, commissions form, reports get written, a few policies get tweaked, and everyone declares victory. Then, a few years on, the same terrible thing happens again, sometimes worse, and we act surprised.

That's not failure. It's the system working as intended. Reform, in the ordinary political sense, was never built to fix broken systems. It was built to keep them running.

What gets called reform

In practice "reform" usually means taking a fundamentally unjust system and making it slightly less obviously horrible. Not the root causes, not the distribution of power, not the structures that produce the harm.

Just the roughest edges, sanded down. Oversight committees with no power to enforce. Trainings everyone ignores. Guidelines that are optional once the meeting ends. New language in official documents while the underlying machinery runs unchanged. Reform as a pressure valve: let out enough steam to keep the thing from blowing, then back to business.

The ratchet effect

Watch which changes survive. Progressive reforms get implemented and then quietly rolled back or starved. Reactionary changes stick. They become the new baseline.

Expand surveillance powers during a crisis and the powers outlast the crisis. Grow enforcement budgets and they never shrink back. Build new prisons and someone will find people to fill them. The ratchet turns one way only. Reach, budget, authority, all of it expands, and reform that works inside the framework treats each expansion as the given it has to start from.

"We need better training for ICE agents" concedes that ICE will keep operating at scale. "We need more humane detention centers" concedes mass detention. "We need clearer enforcement priorities" concedes that enforcement at this level is fixed. The premise of the reform has already lost the argument it claims to be having.

The expertise trap

Reformist politics loves expertise. Policy wonks, think tanks, white papers laying out the seventeen-point plan to improve outcomes by adjusting budget lines and tuning regulatory frameworks.

Expertise does matter; you have to understand a system to change it. The trap is what the expertise quietly does to the question. It moves the fight from "should this system exist" to "how do we optimize this system," turning a normative question into a technical one. We stop asking whether to have mass incarceration and start debating the most effective recidivism programs. We stop asking whether immigration enforcement at this scale serves any legitimate purpose and start ranking which enforcement priorities are most defensible.

The experts are usually smart and well-meaning. But their expertise lives inside a frame that takes the system's legitimacy as settled. They tune the machine. They don't ask whether it should be running.

Electoral politics and structural limits

The reformist pitch runs roughly: elect better people, pass better laws, and the systems improve over time. It's a seductive story and an ahistorical one.

The largest expansions of harmful systems often arrive under nominally liberal administrations, because those administrations want to look "tough" on whatever the right has weaponized, because they inherit the machinery and find it easier to run than to scrap, and because the economic and political incentives that built the system in the first place never moved.

You can't vote your way out of structural violence when the structure itself decides which reforms are even on the menu. The system selects for candidates who won't seriously challenge it; campaign finance, media access, and party gatekeeping screen out anyone who might. That doesn't make electoral politics worthless. It makes treating elections as the main engine of systemic change a category error. The master's tools, to borrow Lorde's phrase, won't take down the master's house.

The incrementalism myth

"We can't change everything at once, so we make incremental progress." Reasonable, pragmatic, realistic. It also rests on two assumptions that usually don't hold: that small changes accumulate into big ones, and that the system will permit reforms that actually threaten its core functions.

Systems are very good at swallowing incremental reform without changing. They adapt, take on the vocabulary of reform while keeping the substance of oppression, and implement exactly enough to defuse the pressure before reverting once attention wanders. And the strategy asks people to absorb the harm in the meantime, the deportations and detention and violence, on a promise that things trend better. Wait. Be patient. Vote for the lesser evil one more time. How long is anyone supposed to wait while the machine keeps grinding them down?

Co-option and incorporation

Systems neutralize threats by absorbing them. The radical demand gets watered into a policy proposal, the proposal gets negotiated into a modest pilot, the pilot gets underfunded into failure, and when it fails the whole idea is discredited. The other route runs the opposite direction: the reform gets implemented in a form that strengthens the thing it was meant to challenge. Community policing was supposed to soften aggressive law enforcement; it extended police reach into more of daily life. Restorative justice in schools sometimes just opens a new channel into the criminal legal system.

A system rarely needs to destroy its critics. It only needs to absorb them, turn them into stakeholders, hand them a seat at a table built to keep real change off the agenda.

Why "It's Working As Designed" matters

Here is the load-bearing claim. The systems producing harm aren't broken. They're doing what they were built to do.

Mass incarceration isn't a failure of the criminal legal system; it's a feature. Immigration enforcement isn't malfunctioning when it deports long-term residents and separates families. That's the design. Economic arrangements that generate enormous inequality aren't glitching. The inequality is the output.

Once that lands, the futility of reform follows. You can't fix a system that's achieving its intended outcomes by doing more of the same. There's no humane mass deportation, no gentle carceral state, no equitable form of exploitation. The problem was never implementation, never a shortage of good people or clever policies inside the machine. The problem is the machine.

What doesn't work

So if reform doesn't work, what does? First, the things that reliably don't.

Waiting for the right election. Courting the center. Trading away core principles to build coalitions with people who don't share the goal. Banking incremental wins that leave the structure intact, and throwing parties for symbolic victories while the material conditions get worse. Trusting institutions whose job is to preserve the status quo. Expecting the people who profit from the arrangement to give up their advantage on request, or believing that one more clear explanation will make the powerful see the light.

We know these don't work because we've run the experiment for decades and the numbers keep moving the wrong way.

The organizing alternative

What does produce change is organized people building power outside institutional channels. Direct action that imposes real costs. Mutual aid that demonstrates an alternative is possible. Solidarity across the divisions the system manufactures to keep people atomized and weak. Not lobbying, not white papers, not waiting for permission, but the capacity to make demands that can't be ignored and to back them with collective action.

This isn't a blanket rejection of reform. Some reforms materially improve lives and are worth fighting for. The catch is that the fight has to sit inside a larger strategy aimed at building power, not stand in for one. The goal isn't a system that works better. It's enough power to replace the system with something that doesn't need constant reform, because it wasn't built on exploitation to begin with.

The horizon problem

Reform struggles run into a horizon problem. The horizon of the possible gets drawn at whatever looks achievable inside current political constraints, and those constraints feel like physics. They aren't. They're artifacts of a particular distribution of power.

Abolitionist organizing pushes the horizon out. It makes demands that look impossible inside the current frame precisely because the frame is the thing producing the harm, and in doing so it changes what counts as thinkable. The eight-hour workday looked impossible until people organized and won it. Chattel slavery looked permanent until it wasn't. Apartheid looked eternal until it fell, not through patient negotiation but because people made the status quo too expensive to maintain.

Building alternatives

The other half of the work is construction. You don't only tear systems down; you build things that meet people's needs better than what you're fighting. Community defense networks that keep people safe without police. Mutual aid that delivers material support without the welfare state's dehumanizing paperwork. Care structures that respond to harm without a cage. Cooperatives that spread power and resources more evenly.

These aren't only protest tactics. They're prefigurative: the world you want, built inside the shell of the one you have. And they make an argument by existing, that the systems we're told are necessary aren't.

The long game

None of this is fast. Power takes time to build, consciousness takes time to shift, durable alternatives take time to stand up. This is generational work, not a three-election plan. The reformist path isn't fast either, though. It's a treadmill of small gains and large reversals, crisis and recovery, and abolitionist organizing at least has a coherent theory of how change actually happens. And while the long work runs, people are already doing mutual aid, giving direct support, reducing immediate harm. That part isn't nothing.

What success looks like

Success isn't the right people elected or the perfect bill passed or the powerful talked into being less powerful. It's collective capacity that doesn't need their permission, solidarity that can't be split, alternatives that end your dependence on systems built to harm you, and power your demands can't be brushed aside. Then it's using that power not to tune the existing system but to make a different one.

The choice

So here's where it leaves us. You can keep trying to reform systems that were never built to serve most people, asking nicely for crumbs from a table built on exploitation, betting the next election finally breaks the pattern. Or you can build something else, accountable to communities instead of capital, that treats people as people rather than problems to be managed.

Reform will keep failing. That's the job it was given. The open question is how long we keep pretending otherwise.