Over the weekend there was a series of bombings and attempted bombings in New Jersey and Manhattan (where I live). Authorities have identified and arrested one Ahmed Khan—a naturalized citizen who came to the U.S. from Afghanistan as a young boy—in connection with the attacks, which injured dozens of people in the New York area.

Republican presidential nominee Donald Trump was quick to seize on this incident as further proof of the need for racial profiling in the fight against terrorism—a call he made previously in the wake of the massacre at the Pulse nightclub in Orlando, perpetrated by Omar Mateen (who was also of Afghan heritage, albeit born and raised in the United States).

The intuitive appeal of this strategy is obvious: it seems like a “certain kind of person” tends to commit these acts. Let’s pay closer attention to “those people” and we can probably nip a lot of attacks in the bud. In fact, the solution sounds so straightforward that many perhaps wonder why on earth this practice is not already central to our law-enforcement and counterterrorism portfolio.

I will briefly answer that question.

1. Most in America’s Arab community
are Christians, not Muslims—and conversely, most American Muslims (and indeed, most of the world’s Muslims) are non-Arab.

If asked to pick a racial or ethnic group most closely associated with Islam, most would answer “Arab.” There is a sense in which this choice is natural: Islam originated in the Middle East, and when we talk about Islam it is generally in the context of Middle East invasions and occupations and their blowback. And indeed, a plurality (about 22 percent) of those arrested for jihadist terrorism in the U.S. have been people of Middle Eastern origin.

However, if police were to profile or otherwise discriminate against American Arabs, Christians would bear the brunt of these policies—just as Arab Christians (along with Sikhs, Hindus, and other ethnic and religious groups) are often targets of hate crimes intended for Muslims. Moreover, considering that about 78 percent of those arrested for Islamic terrorism thus far have not been Arabs, devoting greater time and resources to this majority-Christian ethnic bloc would distract law enforcement from the individuals who posed the greatest risk.

Here, the reader might think, “Most American Muslims aren’t Arabs? No problem. Authorities can just profile whichever group most American Muslims happen to be.” However, this also proves untenable: most American Muslims are African-American. However, African-Americans amount to a full 12 percent of the U.S. population, and more than 90 percent are non-Muslim. In fact, as with America’s Arab population, most African-Americans are Christian. So as with Arab Americans, if law enforcement focused resources on this group to prevent Islamic terrorism, these efforts would be not only unwieldy, but also ineffective.

2. Most American terrorism is carried out by people who were born and raised in America, not by immigrants.

Donald Trump’s discussions about racial profiling often occur in the context of a broader conversation about immigration policies. There is a pervasive fear that immigrants are responsible for the lion’s share of terrorism.

Allow me to alleviate that concern: most terror attacks are carried out by people who were born and raised in the United States. Even about half of jihadist terrorism is committed by U.S.-born citizens.

3. It is both difficult and unwise to profile Muslims.

Here, the reader may be tempted to say, “All right, forget about the race business, or even national origin—let’s just profile people based on religion.” But it is difficult to identify who is a Muslim, and determining which Muslims harbor jihadist sympathies is even more complex.  

One might think, “Well, we know a lot of Muslims congregate in mosques. Let’s surveil those!” (This is another suggestion Donald Trump has previously made.) There are two big problems with this.

First, a large share of the Westerners who gravitate toward groups like ISIS are new converts or the newly devout, rather than lifelong pious Muslims. They don’t regularly attend a mosque, and in fact tend to view mainstream Muslims as being ideologically compromised (which is one reason it doesn’t really matter if “moderates” speak out to condemn terrorism). In other words, although mosques happen to be sites where Muslims regularly gather, they are the wrong place to look for the Muslims most likely to commit terror attacks.

Second, subjecting Muslim communities to greater surveillance or infiltration by law enforcement has a chilling effect on cooperation between Muslim populations and the authorities—while simultaneously affirming the Islamic State’s narrative that Muslims can never truly belong, be accepted, or worship freely in “the West.” (Indeed, provoking hostility toward Muslims abroad is an explicit goal of ISIS’ terror attacks.) Both of these trends would render a successful jihadist attack more, not less, likely.

And more broadly, focusing disproportionately on Islamic terrorism actually makes it easier for non-Islamic terrorists to succeed. Prior to the Pulse nightclub massacre, most deaths from terrorism in the United States since 9/11 were justified by appeals to non-Islamic (at times, anti-Islamic) right-wing ideology. However, it would be a mistake to assert that law enforcement should profile people based on their political affiliation—such a strategy would fail for basically the same reasons that attempts to target people based on their religion or ethnicity do.

4. We tried it and it didn’t work.

Many of Trump’s proposals, including profiling Muslims and infiltrating Muslim communities, were attempted in the aftermath of the 9/11 attacks. The reason most of those policies were abandoned had little to do with “PC culture” and a lot more to do with the high costs and poor results of the programs. That is, law enforcement has come to discourage profiling for the same reason intelligence services have disavowed torture: it simply doesn’t work—neither in theory nor in practice. Donald Trump has been ill-served by his rogues’ gallery of national-security and law-enforcement supporters who take a contrarian stance on these issues, as would be the American people if these supporters were allowed to implement their policies.

However, the U.S. intelligentsia is as much to blame for this outcome as Trump (perhaps more so), given that so many who could provide him sound advice have instead attempted to distance themselves from the candidate or even worked toward his demise—as though Clinton and her advisors are not equally committed to doubling down on failed national-security policies.

This is highly irresponsible, and they may come to regret it. Because despite their antipathy—indeed, perhaps precisely due to “establishment” resistance—Donald Trump has a solid chance of emerging victorious in November, and all of these more “reasonable” experts who didn’t want to sully their hands or reputations will find themselves on the margins, watching helplessly as America once again has to learn the hard way that tactics like profiling and torture provide overwhelmingly bad intelligence while exacerbating terrorism. (In the case of torture, this is a lesson we have ostensibly “learned” several times before to no avail.)

Profiling is ineffective. Torture is ineffective. But sanctimonious armchair criticism is also ineffective. If you want to prevent a train wreck, get onboard the Trump Train and help steer.

Musa al-Gharbi is a Paul F. Lazarsfeld Fellow in sociology at Columbia University.