Donald Trump’s immigration crackdown has entered a new phase after the 26 November shooting of two West Virginia National Guard members in Washington, D.C., which authorities are investigating as a targeted attack. One Guard member, Sarah Beckstrom, 20, died from her injuries, while Andrew Wolfe, 24, was critically wounded. Federal authorities identified the suspect as Rahmanullah Lakanwal, an Afghan national who had worked with US-linked forces in Afghanistan and entered the United States in 2021 through Operation Allies Welcome following the American withdrawal. Trump responded by announcing much broader immigration restrictions, including a review of immigration cases involving people from countries his administration considers security risks, a suspension of asylum decisions and a halt to processing certain Afghan immigration applications. The administration has framed the measures as necessary after a serious failure of immigration screening, while critics argue that one individual’s alleged crime should not become evidence against entire nationalities or migration programmes. The difficult policy question lies between those positions: a government has a responsibility to update security policy after a failure, but good risk management requires identifying the mechanism that failed rather than simply expanding restrictions to everyone who shares characteristics with the attacker.
The Afghan context makes this unusually complicated. When Kabul collapsed in August 2021, the United States evacuated more than 120,000 people in a matter of weeks, including Americans, citizens of allied countries and Afghans who had worked with Western forces or were considered vulnerable to Taliban retaliation. Normal immigration processing can take months or years; the evacuation needed to occur in days. The Biden administration therefore used humanitarian parole and accelerated screening to admit many Afghans while additional vetting continued. That created an unavoidable speed-versus-screening trade-off. Leaving people in Afghanistan longer allowed more time for background checks but exposed former American partners to potential Taliban retaliation. Evacuating quickly reduced that danger while increasing the probability that incomplete information accompanied some arrivals. Public policy rarely offers zero-risk options; governments choose which risks they are more willing to bear.
The relevant question now is whether the Washington shooting reveals a systematic weakness in that process or an individual case that no realistic screening system could have predicted. Background checks are good at discovering existing records, known organisational affiliations or intelligence warnings. They are much worse at predicting what a person with no disqualifying record may do several years later. This distinction between screening risk and behavioural risk matters enormously. If investigators discover that information available in 2021 should have prevented Lakanwal’s admission, screening procedures clearly require correction. If no such information existed and radicalisation or psychological deterioration occurred later, stricter entry vetting might not have prevented the attack. The appropriate policy response depends on the causal mechanism.
This is why national-security policy should resist the base-rate problem. Millions of immigrants can live peacefully while one individual commits an extraordinary crime. The crime receives enormous attention precisely because it is unusual, but visibility can cause people to overestimate how representative it is. Conversely, dismissing the incident as statistically rare would also be inadequate because governments have a special obligation to investigate low-probability events with catastrophic consequences. The rational response is neither collective suspicion nor complacency. It is targeted updating: determine which information systems failed, improve intelligence sharing, review genuinely comparable cases and concentrate investigative resources where evidence indicates elevated risk.
Trump’s broader suspension of asylum decisions illustrates how quickly a security incident can become a platform for a much larger policy agenda. The administration already wanted substantially lower immigration and stricter asylum rules before Wednesday’s attack. The shooting therefore changes the political feasibility of policies whose underlying objectives pre-date it. This is common after crises. Governments possess policy preferences waiting for an event that increases public willingness to accept them. The policy may still be justified, but analysts should distinguish measures designed specifically to fix the demonstrated failure from measures that use the failure to advance a broader ideological programme.
Afghan migrants present a particularly strong case against indiscriminate assumptions because many entered the United States precisely because they had assisted American forces, diplomats or organisations during the twenty-year war. Some interpreters and local partners accepted extraordinary personal risks because they were told the United States would not abandon people who supported its mission. Treating the entire group as presumptively dangerous after one alleged attack would therefore create a foreign-policy cost beyond immigration. Future local partners in another conflict will observe how previous partners were treated after the war ended. If helping America creates a long-term risk of abandonment or collective suspicion, cooperation becomes harder to obtain. Credibility towards wartime partners is a strategic asset.
This does not mean Afghan applicants should receive weaker security screening because of their service. The opposite is possible: governments can honour commitments while applying rigorous individual vetting. The principle should be individualised risk rather than nationality as a proxy for guilt. Nationality can sometimes contain relevant statistical information when state institutions are weak and records unavailable, but it is an extremely blunt screening instrument. Two people holding Afghan passports can have completely different histories, affiliations and risk profiles. The more government can replace crude proxies with individual information, the more security it can obtain with fewer unnecessary restrictions.
The asylum suspension raises a separate issue because asylum is designed precisely for people claiming persecution who may arrive without ordinary immigration documentation. A functioning system needs to distinguish genuine refugees from economic migrants and security threats quickly. America’s asylum process has instead accumulated enormous backlogs, meaning applicants can remain for years while cases are decided. That creates incentives for weak claims because simply entering the process can provide a long period of residence, while genuine refugees wait unnecessarily for certainty. The long-term solution is therefore not merely making asylum easier or harder. It is making adjudication faster. A decision reached in several months produces stronger enforcement and stronger humanitarian protection than a decision reached after several years.
The same principle applies to immigration more broadly. Trump is politically strongest when he points towards genuine failures of state capacity: an uncontrolled border, unresolved asylum cases or screening systems unable to integrate information across agencies. His critics are strongest when enforcement becomes collective or disproportionate. These positions do not need to remain mutually exclusive. A state can maintain a tightly controlled immigration system while admitting substantial legal migration. In fact, stronger enforcement can increase public support for legal immigration because citizens become more confident that the rules determine who enters.
This is the legitimacy dividend of administrative control. Immigration becomes politically explosive when voters believe governments have lost the ability to distinguish legal from illegal entry or low-risk from high-risk applicants. Restoring that distinction can reduce pressure for indiscriminate restrictions. The best immigration system therefore is not necessarily the one admitting the fewest people. It is the one where admissions reflect explicit democratic rules and government possesses enough administrative capacity to enforce them.
The Washington attack will understandably intensify American fear because the victims were National Guard members deployed in the capital, making the incident appear symbolically directed at the state itself. But symbolic power should not determine evidentiary standards. Investigators still need to establish motive, background and whether any wider network existed. Policy designed before those facts are known risks solving the wrong problem. Urgency is not the same thing as information. Governments sometimes need immediate precautionary measures, but permanent policy should follow investigation rather than precede it.
There is also a broader lesson from Afghanistan. The chaotic 2021 evacuation forced America to make immigration decisions under conditions created partly by the speed of the Afghan state’s collapse. The consequences of foreign policy therefore do not stop when troops leave. Refugee commitments, veterans, intelligence relationships and migration pathways can continue for decades. Ending a war ends one category of expenditure while creating others. This is another reason intervention should be evaluated through its full lifecycle cost, not merely the period in which soldiers are deployed.
Trump will almost certainly use the attack to strengthen his argument that immigration policy should place security ahead of humanitarian commitments. The political appeal is obvious because governments exist partly to protect citizens. But security policy becomes strongest when it remains discriminating enough that resources are concentrated on actual threats. Reviewing thousands of cases may be rational if investigators identify a common screening vulnerability. Reviewing millions of people primarily because of nationality can consume investigative capacity that could have been directed towards individuals presenting stronger warning indicators.
That is the opportunity cost often missing from “maximum security” rhetoric. Every person investigated requires investigators’ time. A system generating too many false positives can become less secure because genuine signals disappear inside enormous volumes of low-risk cases. Intelligence quality therefore depends partly on selectivity.
The correct standard after Washington should be demanding but simple. Establish how the suspect entered, what information existed at each stage, whether later warning signs appeared and which agencies possessed them. Correct any institutional failures revealed by that timeline. Continue rigorous screening of Afghan and other applicants. Prosecute the alleged attacker through ordinary law. Protect America’s commitments to people who assisted it where they themselves present no evidence of wrongdoing.
The shooting is a tragedy and potentially a security failure. Whether it justifies a much broader transformation of immigration policy depends on what the investigation discovers. A capable state learns from exceptional events without allowing exceptional events to erase the distinction between the individual who committed an act and the millions of people who did not.
