Sen. Ted Cruz’s recent call to “patrol and secure Muslim neighborhoods” brought widespread scorn, including from some strange bedfellows. Lambasting Cruz’s statement were not just civil liberties groups and President Obama but also the NYPD, GOP candidate Gov. John Kasich and Fox News’ Bill O’Reilly. The consensus: Singling out Muslim communities would backfire, and persecuting Muslims could end up radicalizing them.

It’s not hard, though, to see why Cruz thought his proposal might be a political winner. Just look at the rationale he gave on CNN:

If you have a neighborhood where there’s a high level of gang activity, the way to prevent it is you increase the law enforcement presence there and you target the gang members to get them off the streets.

He’s right about one thing: Targeting gangs is not radical, it’s the status quo. We even call them “terrorists.” California’s 1988 STEP (Street Terrorism Enforcement and Protection) Act pioneered criminalizing gang membership and “enhanced”’ penalties for gang-related crimes.

Hillary Clinton may now regret using the word “super-predator” during the 1990s crime wave, but targeting gangs is still widely accepted. Thirty-two states now have stronger penalties for gang-related crimes, according to a study by the National Gang Center.

Swapping “Muslim neighborhoods” for “gang” ones might have struck Cruz as logical, even mainstream. Yet to commentators across the political spectrum, it seemed obvious that treating all Muslims as potential terrorists would be alienating and counterproductive. Indeed, a growing body of empirical evidence supports that conclusion.

Being Muslim doesn’t make you a terrorist. Being poor and black doesn’t make you a gang member. But being treated like one might. 

What’s surprising is how few recognize the same fatal flaw in anti-gang measures. Just as living in a Muslim community doesn’t make you a radical Islamist, living in a poor community of color doesn’t make you a gang member. And yet that’s how these communities are often policed.

Police keep databases of suspected members, but often include people based largely on physical appearance. Looking like a gang member — and often just being a young man of color — is enough to raise your chances of arrest and imprisonment. Anti-gang laws increase punishment for these “false positives” too. Those accused often take plea-bargain deals to avoid potentially devastating, “enhanced” gang-related sentences.

For parents, spouses and children in “gang communities,” such targeting can inspire indignation and suspicion, and discourage cooperation with law enforcement. And would they be wrong? Evidence suggests that mass incarceration in the United States has strong racial bias and been disastrous for affected communities.

If the criminal justice system is going to treat you as a gang member, you might as well become one. 

For targets themselves, the effect can be even more perverse. If the criminal justice system is going to treat you as a gang member no matter what you do, you might as well join. At least you will get some protection, especially in prison, where gangs rule.

Indeed, if you expect incarceration, you would probably be willing to go to great lengths to please prison-gang leaders. As David Skarbek has shown, California prison gangs use this leverage to govern and tax street gangs.

My own research shows that poorly targeted anti-gang initiatives increase prison gangs’ power on the street. If crackdowns were truly well-targeted, discriminating between those who obey gang orders and those who do not, they would discourage gang obedience. But they aren’t. If your chances of incarceration go up regardless of whether you follow prison-gang orders, you have more, not less, reason to obey. And as prison conditions worsen with overcrowding, prison gangs’ protection becomes even more valuable.

The bottom line: Just as targeting whole Muslim communities would increase the power of radical Islamist leaders, targeting “anyone who looks like a gang member” increases gang leaders’ power.

Mass incarceration ended up increasing gangs’ power

Consider what happened in California. The Mexican Mafia prison gang first took control of Los Angeles’s Latino street gangs after incarceration rates spiked following the passage of STEP and related initiatives. The Mexican Mafia has used this power to coordinate violent turf wars against black L.A. gangs and rival Latino gangs in northern California — who are now coordinated by their own respective prison gangs.

These perverse outcomes, though, pale in comparison to the disastrous effects of anti-gang initiatives in Latin America.

Across Central America, “mano dura” anti-gang laws transformed neighborhood gangs into franchises of prison-based “maras” operating nationwide extortion networks. In Brazil, mass incarceration policies swelled the ranks of São Paulo’s sophisticated Primeiro Comando da Capital (PCC) prison gang.

In both cases, officials doubled down on gang repression, further spiking incarceration rates. Yet this only gave the maras and the PCC more influence on the street. They deployed this power strategically, orchestrating crippling terror attacks on urban targets to force concessions from officials. Chillingly, new prison gangs in far-flung parts of Brazil have begun to copy the PCC’s terrorist tactics.

Note: The figure above is adapted from my study of prison gang projection of power. A list of sources appears in the study’s appendix.

U.S. prison gangs have not yet used their street power for terrorist purposes. More broadly, researchers have demonstrated many important counterproductive effects of mass incarceration. Political momentum for criminal justice reform is growing.

But scaling back anti-gang laws remains a hard sell. And therein lies the irony of the Cruz episode. Those who oppose such reforms can see clearly why the same policy of poorly-targeted policing, applied to Muslim communities, could be catastrophic.

Benjamin Lessing is assistant professor of political science at the University of Chicago.