Border security is a powerful argument used by critics of immigration. Even those who embrace immigration tend to make reform conditional on securing the borders.

The immigration bill approved by the Senate is a perfect example. The first section is devoted to border security and the third to internal enforcement. The bill mandates more funds and bureaucracy for this purpose and conditions the long transition to permanent residence for undocumented immigrants on the apprehension of 90 percent of those trying to sneak in.

One problem with making a conditional connection between border security, a very legitimate concern, and immigration, is that “security” can refer to different things here: the principle that the violation of the law by newcomers is unacceptable and the threat of terrorism.

The U.S. government has been toughening the border protection and enforcement apparatus for decades. The amnesty signed by Ronald Reagan in 1986 called for such measures. Beefing up controls and enforcement is also a reason why in the 1990s, under Clinton, many seasonal Mexican migrants decided to remain in this country. President Obama spent $18 billion enforcing border security last year with the help of 21,000 border patrol personnel, sophisticated technology and a fence that is several hundred miles long. An obvious reflection of shifting perceptions, immigration, once a brief of the Labor Department and then the Justice Department, is currently under Homeland Security.

This has not been enough to stop illegal immigration. The disconnect between the law and reality usually gives rise to black markets. This has been the case with immigration. Ironically, the market has been a better regulator than the authorities since, due to economic conditions, between 2005 and 2010 net immigration dropped to zero. Which is not to say people have stopped coming in. Precisely because the decades-long effort to strengthen controls has been less than effective, in the last four years another 700,000 people have been added to the undocumented list. The evidence cries for a system that connects supply and demand more realistically.

About half of those who try to sneak in are apprehended. Is it realistic to think that the system will be able to catch 90 percent? No. A country’s obligation to uphold the rule of law and fight terrorism cannot hang on the ability to seal off the borders and enforce police-state controls, which is what, taken to its logical conclusion, the security-based approach of some immigration critics would mean. Can a country accustomed to getting on average sixty million overwhelmingly law-abiding documented visitors seal off its borders without threatening the very rule of law it defends, its economy, and its relationship with the outside world?

A recurring argument posits that the terrorist attacks on U.S. soil prove that security is the paramount issue regarding immigration. This is a red herring since the attacks entailed a failure of intelligence and basic law and order. Of the roughly six million visas issued by the U.S. ever year, a very tiny minority are potentially problematic. What will prevent consuls from issuing them to terrorists, or preempt an attack if they get one, is not immigration policy but counterterrorism policy. Mohamed Atta, a 9/11 ringleader, had had contacts with terrorist organizations for years but was allowed to reenter the United States despite having overstayed his tourist visa.

One third of the forty-eight al Qaeda operatives who committed crimes in the U.S. between 1993 and 2001 were residents or citizens, while another third had visas. They obtained them because the countries where they originated their travel were not on the list of sponsors of terrorism. In a world in which terrorism is a moving target, every country is a potential source. How would, say, a million border agents and a tiny quota allowing foreigners to work and live in the U.S. have stopped these people from coming in or obtaining residence/citizenship?

Even a fortress America would not have prevented the attack on the Boston marathon since the perpetrators were raised in this country. Again, this was an intelligence and law-and-order failure: one of the terrorists had been questioned by the FBI. They were not linked to terrorist organizations and they developed their radical beliefs in the United States. Is it reasonable to expect immigration policy to anticipate who will develop radical anti-U.S. beliefs decades after settling in this country? Can immigration policy prevent anybody from learning how to build an explosive online?

Surely a flexible, realistic immigration policy that does not generate illegal entries is a better way of making sure everyone operates within the law and of helping counterterrorism policy separate the law-abiding wheat from the terrorist chaff.