U.S. Customs personnel. (AP)

Gordon Hanson is a professor in the economics department and School of International Relations and Pacific Studies at the University of California at San Diego, where he holds the Pacific Economic Cooperation Chair in International Economic Relations. He is also the director of the Center on Emerging and Pacific Economies at UCSD. He specializes in international migration, trade and investment issues, and has written on many aspects of U.S. immigration policy, from the effect of illegal immigration on the economy to border security.

We talked about the latter topic Thursday. A lightly-edited transcript follows.

Dylan Matthews: An extreme fatalist would say that there's basically nothing that the U.S. can do to control the border, that as long as there's demand for jobs in the U.S., people will come. What do you make of that argument?

Gordon Hanson: I think it's a bit oversimplified. There are two excellent analyses of the impact of border enforcement on migration, each using data from the 1980s and 1990s, one by Pia Orrenius and one by Christina Gathmann. This work found that increases in border enforcement do deter people from crossing the border, but the effect is small. Do those studies, conducted when we had 4,000 border agents, apply when we have 20,000 agents or 40,000? It's not at all clear. Whereas previous increases in border enforcement left big gaps along the border, the recent and coming surges could possibly eliminate those gaps, causing the effectiveness of border enforcement to rise.

In general, I'm skeptical about the efficacy of huge increases in border security. At the same time, we have to be careful about what we think we know. The impact of enforcement on the decision to migrate looks pretty weak but we don't yet have analyses after the recent massive increases in border presence.

How much is technology changing this? Are drones making surveillance cheap enough that securing the whole border is finally feasible?

The relatively cheap technology of video surveillance cameras has had a significant impact on enforcement, but my sense is that these tools work best in urban areas, where agents have the capability of rapidly responding to information about migrant location. The deployment of surveillance technology is partly behind the dramatic reduction in crossings in San Diego, El Paso and other border cities. Video technologies are relatively cheap. However, the problem is that when you increase surveillance in cities people just stop crossing in urban areas and cross in remote locations, instead.

As I understand it, the current effectiveness of drones for border enforcement is limited. Part of the reason may be that to use drones in a meaningful way the Border Patrol would have to rethink how it deploys agents, with agent locations and movements being dictated by information coming from drones. There still appears to be something of a disconnect between drone operations and the activities of agents on the ground. Using drones properly could dramatically increase the effectiveness of the Border Patrol, but the cost of such a program could be enormous (given how expensive drones are to purchase and operate).

The Berlin Wall was a pretty effective border control mechanism. (AFP/Getty Images)

Are there other countries that have gotten good at border security, who the U.S. could learn from?

The former Soviet Bloc was very good at controlling borders. Such a reference sounds tongue in cheek, but if we're going to have 40,000 agents on the border, 700 miles of significant border barriers, and drones in constant circulation, how far are we from Soviet style border security? For the Soviet Bloc, border security had tremendous symbolic importance. The image of large outflows of emigrants seeking to escape the Soviet Union and its satellites would have been emblematic of the failure of those countries to be attractive places to live.

I would argue that our interest in stopping illegal immigration also has a significant symbolic component. An extra $30 billion a year, as proposed in the Senate deal unveiled this week, would take annual spending on border enforcement to around $50 billion. It's hard for me to believe that his level of spending passes any meaningful cost-benefit test.

How are we doing on ports enforcement? You don't hear about that as much as border security. How big a portion of the problem is that?

Most of the apprehensions data I've seen is between ports of entry. Border Patrol and Custom and Border Protection have been guarded about what happens at ports of entries. Yet, the total number of apprehension appears to be small, because you have so many eyes on the cars, trucks and people passing through.

Over the last 20 years, apprehensions plunged first in California and more recent in Arizona. Today, the battle is being fought in south Texas. I think we are too sanguine about the effectiveness of border enforcement, because we haven't faced a big test since the mid 2000s buildup. The poor state of the U.S. economy, and the Mexican economy’s recovery, has kept the incentive for illegal entry low. The only group showing are strong increase is Central Americans, with large numbers of migrants fleeing violence in Honduras and Guatemala.

How much progress have we made in terms of border security? We won't really know until we go through another economic cycle. I'm really curious to see what happens when the economic pressure for migration is back on. In the past, after each time we’ve increased enforcement, and then gone through an economic expansion, we subsequently lost our religious fervor about maintaining the intensity of border security. In the past, strong demand for labor has weakened our desire to maintain the effectiveness of border enforcement.

The 1990s offer an example of this phenomenon. The Immigration Reform and Control Act of 1986 (IRCA) came out during a similar moment of enforcement fervor, and was followed by the Border Patrol launching Operation Hold the Line and Operation Gatekeeper, which caused apprehensions in San Diego and El Paso to plunge. Yet, by the end of the 1990s, with sustained economic growth, President Clinton was effectively saying, "We're only going to focus interior enforcement on nuclear power plants, airports, and other locations central to national security." Somehow, the Border Patrol at the border was letting lots of illegal migrants cross the border.

How exactly did that happen? Did Congress pass laws telling Border Patrol to back off, did Clinton order them to back off, or what?

I think the process is pretty subtle. What do bureaucrats really dislike? Public attention and public criticism from people they deem to have power over them. Fear of this attention gets immigration authorities to back off.

One of my favorite examples comes from Georgia after it expanded cultivation of the  Vidalia onion. . High-end agriculture was a new thing in the 1990s. Vidalia onions had become popular, but they're labor-intensive to grow, which put pressure on farmers to hire lots of undocumented workers. Immigration agents, with their post-IRCA mandate to expand enforcement, began raiding farms.

The political response was immediate and caustic. Both of Georgia's senators several Georgia congressmen, and Attorney General Janet Reno said publicly that the Immigration and Naturalization Service was undermining the livelihood of Georgia agriculture. Soon after these complaints, the farm raids stopped. The lesson is that during good economic times if immigration authorities do what they are supposed to be doing they get in hot water.

But maybe things have changed. Previously, we had an implicit social contract with immigrants. We made it tough for them to cross the border but more or less left them alone once they’d entered the country, as long as they kept their noses clean and heads down.

We broke that contract after 2006, when President Bush expanded enforcement efforts in an attempt to create political support for immigration reform. President Obama has maintained the Bush approach, more or less, with interior deportations in the last five or six years at historic levels. When the economy fully recovers, will we maintain enforcement at its current intensity? I don’t know. History would suggest that we won’t. But one also has to keep in mind that opposition to immigration is much more organized today than in the past, with intense immigrant enforcement being a core pillar for many Tea Party politicians.

Vidalia Onions don't pick themselves. (Ryan Griffis / Flickr)
Vidalia Onions don't pick themselves. (Ryan Griffis / Flickr)

That's interesting, because I often hear more optimism about the potential effectiveness of E-Verify than of border enforcement.

The problem with E-Verify is that, given American sensitivities about privacy and civil rights, we make it hard for a program like E-Verify to work. We are unwilling to demand consistency across individuals in terms of birth certificates, driver's licenses, and other forms of identification. Inconsistencies make E-Verify prone to high false negative rates (that is, finding that people are ineligible for legal employment when they in fact are eligible).

If we are in a moment when hiring pressures aren't that intense, employers will tolerate E-Verify. But if job growth picks up, E-Verify would become more of a hindrance, creating frustrations for employers and political opposition to intense monitoring of worksites.

Also, E-Verify doesn't work for many occupations. It doesn't apply very well to housekeeping, yard care, home renovation, agricultural labor, and some construction jobs. What fraction of employment is accounted for by occupations that will be outside of E-Verify? I don't have a hard and fast number, but I'd be shocked if that share is less than about 35 percent of current employment of undocumented migrants. Large establishments will have a hard time hiring illegal workers, which is already the case, but in other significant swathes of the economy, it's going to be hard for E-Verify to make a dent.

I understand you have some criticisms about how we actually measure border effectiveness.

I was part of this National Academy of Sciences panel on estimating illegal border flows. Our charge was to address "How should DHS go about measuring illegal border flows, and are existing data adequate to ascertain how many people are crossing the border illegally each year?" Some of the panelists were scholars like me, who'd been studying immigration for many years. Others were statisticians with a deep understanding of data collection, surveys, and the statistical properties of administrative information. No one on the panel was impressed by the way that DHS measures illegal labor flows.

We simply don't know how many undocumented migrants cross the border. Pressure on DHS pressure has led them to use essentially manufactured numbers on “turn-backs” and “got-aways.” The resulting effectiveness rates they’ve produced aren’t terribly meaningful, due to double-counting of apprehensions and a host of other statistical problems.

More importantly, DHS estimates of their effectiveness rate are at odds with results from migrant surveys that ask individuals about apprehensions. We should approach estimates of DHS effectiveness with a healthy dose of skepticism.

That would seem to imply that there would be more stat-juking if we had a trigger mechanism, as in the Gang of Eight bill, where legalization provisions are tied to border security effectiveness.

In the end, defining effectiveness will be primarily a political battle. It DHS is required to reach a true apprehension rate of 90 percent immigration reform would never be implemented. So, what I’m guessing will happen is that DHS will define effectiveness measures that they can reach (or that politicians want them to reach). DHS will likely shield information about their activities, as they’ve done to date, complicating independent evaluation of their efforts.

Any final things you'd like to add before we wrap up?

Before the current round of immigration reform got off the ground, it seemed that DHS was on the verge of opening up access to its data on border apprehensions and enforcement. I've done a great deal of work DHS data on apprehensions and enforcement using their internal data. My interest in working with DHS was in ultimately producing academic studies that could be published in scholarly journals. Two years ago, or so, I thought we were close to getting permission to publish the results of our work. However, the environment has changed entirely. DHS won’t let me or others who’ve used their internal data on apprehensions release our work for public view. We’d learn much more about the effectiveness of border enforcement if DHS would release its data in the public domain. They receive free evaluation of their efforts by the scientific community and would surely gain access to more reliable and more defensible indicators of enforcement effectiveness. But DHS is going in the other direction. They seem intent on shutting down any chance of data access.

There's language being put in the immigration reform legislation that prevents the DHS from putting too detailed information on border apprehensions or enforcement in the public domain. I believe this is a terrible mistake. If we want to be able to evaluate the effectiveness of border enforcement, we need DHS to become more open about what it does.