Since being rolled out under the Obama administration, “Countering Violent Extremism” programming (CVE) has been a contentious subject in the United States. Pitched as a way of working with local communities to prevent “radicalization,” such programs have existed for years in European countries, where they have also been controversial. In 2015 former President Barack Obama held a CVE Summit at the White House to announce that the U.S. would be introducing similar measures, an announcement that was met with trepidation from many sides. While rightwing activists derided the programs for advocating partnerships with American Muslim organizations, many of those organizations were themselves suspicious of ill-defined “outreach initiatives” that often seemed more like covert surveillance and intelligence-gathering activities.

In an interview earlier this week with The Atlantic, George Selim, the Department of Homeland Security’s top official for domestic CVE, confirmed that he was resigning from his role  news that was first reported with celebratory tones on the website Conservative Review. In his comments to The Atlantic’s Peter Beinart, Selim said, “There were clearly political appointees in this administration who didn’t see the value of community partnerships with American Muslims” — a change in approach that would make it impossible for him to do his job. Earlier this May, it was also reported that the CVE Task Force Selim headed would have its funding eliminated by 2018, fulfilling a promise by Trump administration officials.

The Atlantic article about Selim’s departure noted that CVE programs were often viewed with suspicion by American-Muslim groups and civil liberties organizations, which considered them stigmatizing. But there were other aspects to CVE programs were more troubling. Efforts to fight terrorism by preventing “radicalization”— itself a very broad and ill-defined concept — had little empirical basis and often led to suspicion being cast on entire categories of people for completely innocuous behavior. Although CVE programming was often sold as an alternative to arresting people and engaging in other forms of “hard” counterterrorism enforcement, the reality was often different. Controversial law enforcement tactics, including the use of sting operations and agent provocateurs, continued unabated, while the existence of CVE programs that were often fronts for conducting surveillance led to the widespread fear and the chilling of constitutionally protected free speech.

“CVE has always been designed to extend surveillance into social services and has always threatened to inflate prosecutions by on-ramping, rather than ‘off-ramping,’ vulnerable folks who pose no real threat of committing or aiding attacks,” said Shannon Al-Wakeel, executive director of the Muslim Justice League, a civil rights organization that has been highly critical of CVE initiatives. “The Trump administration sets a high bar for explicit bigotry and targeting, but CVE has always been wrong-headed, dangerous and, in spite of its messaging, probably more abusive in its tactics than anti-Muslim critics gave it credit for.”

Although it was clearly effective at antagonizing American-Muslim communities, many law enforcement experts say that the millions of dollars the government spent on CVE didn’t seem to serve any actual purpose in identifying or capturing real terrorists. Last year the FBI rolled out an online video game intended to help teach young people how to avoid becoming radicalized, one of many measures introduced to widespread public skepticism.

“CVE has never made any sense to me. It seemed to be a Democrat response to anti-Muslim Republican demands for a counter-radicalization program,” says Michael German, a former FBI agent and a fellow with the Brennan Center for Justice’s Liberty and National Security Program. “Rather than just pointing out that radicalization is a flawed concept and focusing instead on anti-violence programs, they tried to create a counter-radicalization program with softer language.”

A 2015 report by the Brennan Center about CVE found that “a thinly sourced, reductionist view of how people become terrorists has gained unwarranted legitimacy in some counterterrorism circles,” leading to the development of dubious programs trying to identify individuals vulnerable to becoming terrorists. These programs were often based on junk science, identifying factors related to religious observance and lifestyle, including quitting smoking and drinking, growing facial hair, or wearing foreign clothing as leading indicators of terrorist inclinations. A 2007 New York Police Department report promoted the view that radicalization proceeded along a “conveyor belt” that led individuals down the path toward terrorism, effectively identifying even being a practicing Muslim or a member of a Muslim immigrant community in the United States as a first step toward radicalism. The NYPD report became the source of controversy when a 2011 Associated Press investigation revealed that the department’s Intelligence Unit was conducting mass surveillance on Muslim communities in and around New York City. An ACLU lawsuit over the surveillance led to the NYPD removing the report from its website, but the theories it promoted remain influential.

While the NYPD report promoted bogus theories that helped inspire dubious CVE programming, other confidential studies conducted by law enforcement pointed to deeper problems that were fomenting terrorism. One such study by the FBI’s counterterrorism division, published by The Intercept last year, found that foreign U.S. military operations were the most commonly cited motivation in 200 terrorism cases analyzed by its agents, ranking far above any other factor. Atrocities believed to have been carried out by American military personnel in foreign conflicts often trigger feelings of moral outrage in individuals who identify with the victims of those atrocities, leading some of them to engage in violence. The findings of the report aligned with those reported by terrorism experts. such as Marc Sageman, who describe terrorism primarily as a geopolitical phenomenon that would be difficult to address through even genuinely benign “community outreach” programs.

If it’s true that terrorism is more about politics and war, it is unclear what anti-radicalization video games or the monitoring of people’s clothing preferences or facial hair were ever intended to accomplish. Rather than an improvement, however, President Donald Trump’s cuts to the programs likely mean that even worse policies for dealing with domestic terrorism are coming down the pike. Regardless, it remains unlikely that CVE will be fondly remembered by the communities it targeted.

“Many people in government and in the counterterrorism world who pushed CVE programs in public would admit to me in private that they knew they were not helpful,” German said. “At best they would say, ‘It’s better what the Islamophobes or the FBI would do.'” German went on, “I don’t know anyone who believes there is any evidence CVE has been helpful in any way, except for the individuals who have made careers out of it.”