America’s immigration problem raises a huge set of thorny issues. At a theoretical level, it is difficult to articulate, let alone implement, the ideal immigration policy. While there are compelling arguments in favor of the basic norm of free trade, an open immigration policy could lead to massive political dislocations. Allowing the free flow of goods across borders is quite unlike allowing people to do the same. Goods do not put potential burdens on educational, health, and social service institutions; they do not participate in political activities, lobby to become citizens, or vote. By the same token, goods do not bring with them entrepreneurial skills and professional expertise like that possessed by immigrants; and, at the other end of the economic ladder, they help to fill many low-wage positions. Anyone who thinks they can come to a categorical judgment on immigration policy has not thought hard enough about the problem.

These difficulties work themselves into the fabric of our current immigration law. Right now, enforcement of immigration law is entrusted to ICE, the eerie acronym for the U.S. Immigration and Customs Enforcement. ICE is responsible for border protection and dealing with aliens already in the country. Its portfolio of duties involves the enforcement of over 400 different federal statutes. To get the barest sense of the massive size of this apparatus, just leaf through the major provisions of the Illegal Immigration Reform and Immigrant Responsibility Act (IIRIRA), which addresses border control, smuggling, document fraud, employment restrictions, and, with renewed urgency given the imminent arrival of the Trump administration, the “inspection, apprehension, detention, adjudication, and removal of inadmissible and deportable aliens.”

Every cranny of such statutes is laced with hidden complexities and potential constitutional problems. Just this past week, in Jennings v. Rodriguez, the United States Supreme Court grappled with the vexing question of what kind of statutory and constitutional protections are or should be given to immigrants who are held in detention for long periods of time without a review of their status. The Ninth Circuit decided that all these detainees had to be released after six months of detention if the government did not prove by a preponderance of evidence that their detention was warranted. It is unlikely  that this aggressive bit of judicial intervention will be sustained. But no matter the outcome of the case, the underlying difficulty remains. How does ICE deal justly and efficiently with the tens of thousands of individuals in custody at any one time, even at current enforcement levels?

To add to the complications, the detainees are wildly heterogeneous. Some are stateless persons, who by definition cannot be deported to their home countries. Some are persons with strong but uncertain claims for refugee status. Others are legal permanent aliens who are threatened with deportation for the commission of minor drug offenses that would merit only modest punishment if committed by citizens. Handling these cases in the criminal justice system is made infinitely more delicate because a plea bargain that might easily be reached with a natural-born citizen could precipitate a deportation proceeding against a similarly situated alien, thereby putting enormous pressure on prosecutors and defense attorneys to avoid penalties that could trigger a deportation that neither side wants.

In light of these difficulties, the prudent practical approach is to give strong weight to the status quo ante, by implementing only those changes that will make some clear improvement to the immigration situation, without upsetting America’s fragile consensus on immigration policy. One sensible place to start is with the liberalization of the H-1B visa program, which could help expand the competitiveness of American businesses both at home and abroad. These visas represent only a small fraction of the overall immigration system. The far greater risk to the system comes from the bellicose remarks of President-Elect Donald Trump, who has vowed repeatedly to step up enforcement of the immigration laws. Thus, in his November 13 60-minutes interview with Lesley Stahl, Trump minced no words when he said:

What we are going to do is get the people that are criminal and have criminal records—gang members, drug dealers—where a lot of these people, probably two million, it could be even three million, we are getting them out of our country or we are going to incarcerate. But we’re getting them out of our country, they’re here illegally. After the border is secure and after everything gets normalized, we’re going to make a determination on the people that they’re talking about who are terrific people, they’re terrific people but we are gonna make a determination at that. But before we make that determination, it’s very important, we are going to secure our border.

These statements reveal that Trump has paid scant attention to the nature of the immigration problem, the probable impact his policies will have on aliens and citizens alike, and the huge financial, logistical, and constitutional obstacles that stand in the path of his proposed program.

Start with the nature of the underlying problem. Trump makes it appear as though there were some vast new threat from immigration, most notably from Mexico. But the facts reveal a different story. The overall rate of immigration into the United States stands at about 3.1 immigrants per 1,000 per year, trending slightly downward from 2000 onward. Set against the backdrop of a declining birth rate inside the United States, foreign immigration acts as a useful counterweight, which among other things is necessary to prop up the generous entitlements supplied to senior citizens under Social Security, Medicare, and Medicaid. More concretely, the net rate of immigration from Mexico, which was very high, totaling 600,000 in the boom years of 2006-2007, was sharply negative in the recession years of 2012-2013, at minus 600,000, and roughly neutral since that time. The same picture has applied globally since 2009, with net immigration and outflow of illegal aliens about constant, with between 300,000 and 400,000 in each group per year. I use the term “illegal” instead of the more fashionable “undocumented” for two reasons. First, it is the statutory language; and second, it expresses an accurate statement about the legal position, which the word “undocumented” fails to to capture. It is not possible to be a legal illegal alien. It is possible to be a legal but undocumented one.

No matter the terminology, however, Trump’s broad allegations also fail to make any sense in dealing with aliens with criminal records. As noted before, the risk of deportation creates a powerful deterrent against the commission of crimes for any illegal alien. Over the past 25 years, we have seen a rapid decline in the overall and violent crime rates of the early 1990s, both when illegal immigration was high and after it declined. A similar decline is found among all subgroups of illegal aliens. The recent upsurge in criminal activity in the post-Ferguson era has nothing to do with immigrant populations. It is more likely that, in some key urban areas, it stems from the so-called Ferguson Effect. Hillary Clinton’s well-noted campaign charges of institutional racism against the police only worsened the situation. But no matter how one views that controversial issue, gang members and drug dealers make up at most an insignificant fraction of the illegal aliens in the United States. In any event, they are already subject to deportation under current rules. It is also worth noting that the net removals of aliens increased in the Obama years, both for criminal and noncriminal aliens. That number was 392,000 in 2010, and it is hard to see how the system could double or triple that figure in the next several years without a massive commitment of resources, which in turn would produce vast dislocations in a domestic economy that depends heavily on illegal aliens to keep things moving. A recent McKinsey study offers a powerful endorsement of the massive social gains that occur globally through immigration.

In light of all of this, it is hard to fathom how any major shift in immigration enforcement policy would be a net good for this nation. At one level, mass deportations would undermine local economies, reduce tax revenues (especially all sales and excise taxes), militarize our cities, and cost a fortune. Wholly apart from the economics are the potentially catastrophic social consequences of turning upside down the lives of these illegal aliens.

The threat of the Trump program is also manifest in the outspoken statements of key Democratic mayors. These city leaders plan to run “sanctuary cities” in which they will not cooperate with ICE officials in enforcing immigration laws, even as they stand to lose millions in potential federal aid. In one sense, the effectiveness of this program is limited, because it is clear as a matter of constitutional law that the federal government has full power to enforce its immigration laws without the cooperation of local officials, who in turn are bound not to interfere with those federal efforts. As the 2012 Supreme Court decision in Arizona v. United States makes clear, the United States can preempt all local laws that are inconsistent with federal policy. Arizona was an Obama administration victory because it preempted Arizona laws that wanted to step up enforcement of the control of illegal aliens beyond the level of federal norms. Turnabout is fair play, so now states cannot, under any Trump administration reforms, interpose their authority against that of the federal government. On the other hand, Trump should know that these expanded federal efforts will surely falter as the United States is not under current law entitled to dragoon local officials into funneling individuals into the ICE system. A little cooperation could go a long way.

It is critical in this situation for the new Trump administration to back off its confrontational policy and instead seek a more incremental and balanced approach to the issue. Ironically, one hugely important step in dealing with the immigration problem is to liberalize our trade policy with Mexico and other Latin American nations from which large numbers of illegal immigrants come to the United States. Increased trade will have two key effects. First, it will improve the economic situation in other countries, which in turn will reduce immigration into the United States and also induce some illegal immigrants to return to their home countries. In addition, free trade will increase exports to foreign nations, thereby increasing job opportunities for citizens and aliens alike. Trump’s quixotic campaign will at the very least block these improvements, and worse could serve to further inflame political passions and partisan divisions that only make the task of intelligent immigration reform even more difficult.

© 2016 by the Board of Trustees of Leland Stanford Junior University