The Obfuscation Report

Apr. 10, 2007 | It was the video that burned up the Internet on Friday -- Fox News hosts Bill O'Reilly and Geraldo Rivera, normally amicable foes, if not allies, screaming at each other on the set of "The O'Reilly Factor" with such ferocity they seemed likely to come to blows. At issue was the case of Alfredo Ramos, an illegal immigrant who allegedly killed two teen girls while driving drunk in Virginia Beach, Va. O'Reilly argued that the two teens would not have been killed had Ramos been deported, while Rivera contended that O'Reilly was "obscuring a tragedy to score a cheap political point." Rivera also claimed that "illegal aliens commit crimes at a lower rate than citizens do."

On Monday, Salon spoke with professor Robert J. Sampson, chairman of the sociology department at Harvard University and most prominent member of a new school of academics who say that, contrary to widespread public belief, immigrants may actually be the secret to decreasing crime in the U.S. Sampson et al. believe their research shows immigrants are less likely to commit crime than native-born Americans, and that immigration itself may actually play a role in lowering the overall crime rate. Salon asked Sampson to rate O'Reilly and Rivera as debaters, and to explain what his research says about immigrants and Americans' perceptions of them.

What did you think of Bill O'Reilly's arguments?

Well, I wasn't terribly surprised -- I think it's emblematic of a larger way in which these incidents are often perceived, and I think he was thrown off by being confronted with what I thought was a pretty logical argument on Rivera's part ... So basically the way I would interpret it is as follows: [O'Reilly] was interpreting this particular incident, which, of course, is horrible, only from the lens of thinking about the person as an illegal immigrant, rather than actually confronting the data, which show that in fact immigrants, illegal aliens, are disproportionately less likely to be involved in many acts of deviance, crime, drunk driving, any number of things that sort of imperil our well-being. And so what he was doing was starting with the category of the person -- illegal -- and inferring from that things that don't follow.

In my opinion, this is sort of typical ... unfortunately, for many Americans it seems that the old adage has basically been turned around. Believing is seeing. And I think in this case O'Reilly has a particular perception and belief [about a] category [of person], which then influenced the way he saw and interpreted this particular event, and Rivera called him on it, and he got upset, obviously.

But on some fundamental level, doesn't O'Reilly have a point -- this wouldn't have happened had Alfredo Ramos not been living in this country, right?

That's true. And you can follow that logic out for a lot of different things; when I teach my crime classes, I often say, "Well, we can basically eliminate crime, right? If we really wanted to, we could abort all male babies. That would reduce the crime rate to pretty much zero in the future."

So yes, you can think of counterfactuals -- if a category of persons were not actually here, then yes, the crime would not have been committed. But let's extend that logic: If the majority of people who are in the category of producing most drunk-driving homicides or deaths were not in the country then by definition the rate of drunk-driving deaths would be reduced. So who is that? Well, they're young people, disproportionately male, disproportionately white, mainly suburban ... The perception and the stereotype is what's driving the argument, not the data.

What have you found in your research about the relation of immigration to the crime rate?

In our research, which is based on over 10 years of data collection and analysis of a long-term study in Chicago, our findings tend to be quite similar to other research showing that first-generation immigrants have lower rates of crime, particularly violent crime. In particular, first-generation immigrants, that is, people born outside the country, are much less likely to commit violence, in our data about 45 percent less likely than third-generation immigrants. In turn, second-generation immigrants are about a quarter less likely to commit crime than third-generation.

So, in other words, native Americans, those born here and whose parents are born here, are the most violent and the most criminal. And that's not just our data, this is other data. Immigrants are less likely to be imprisoned relative to their numbers; Latinos, in particular, even though they enter the country being disproportionately poor, which would signal, based on everything else we know, that they would have a high risk for all sorts of negative outcomes, including [poor] health, low-birth-weight babies, incarceration, violence and so forth, [are less likely to be imprisoned]. They are doing rather well in many dimensions, and this has led to what is known in the literature as "the Latino paradox." And the paradox is just that -- even though they are disproportionately poor and have all kinds of risk factors, they are doing better in many dimensions. So that particular finding in our data I think is consistent.

Then one can look at all kinds of other data. I would point to two broad trends. One, if you look at the crime rate or the violence rate, or in particular, one which we can measure very well in the United States, let's take homicide, where in almost all cases there's a body, so we know how to measure it pretty well. Immigration was exploding, literally, in the United States, going up by the millions -- no one disputes this -- in the '90s. I've produced charts showing that as that was going up, violence was going down at a very rapid rate. In fact, the two lines are pretty much inverse to each other. Now, that doesn't prove causation, but it certainly shows that the common perception that as immigration goes up that crime will also go up is just not true.

And then if you look at the cities that are by all accounts immigrant inflows, or border cities in particular, such as El Paso [Texas] or San Diego, Tucson [Ariz.] or other cities, they are not our leaders in violence by any means. In fact, those cities have done quite well. It typically, historically, continues to be cities with high proportions of native Americans that have the high homicide rates, whether it be Baltimore, Detroit, Atlanta or Washington, D.C.

Do the trends for immigrants hold true for illegal aliens in particular? Was Geraldo Rivera right about illegal aliens committing less crime?

It's hard to break out that precise figure, because of the uncertainty. First of all, we're not even allowed, because there are certain restrictions placed on our research, to ask about someone's immigration status.

But ... it certainly would track in our data, in most data. If you think about it, the national trend, an over 50 percent increase in immigration flow over the last 10 or more years, has also been highly correlated with the influx of illegals, so you're finding an influx in both and so the pattern would be similar, and that's my read of the data. Similarly, in Chicago, the neighborhoods we studied that were immigrant enclaves, they were also where you found illegal immigrants ... So to the extent that the patterns hold that link immigration to lower crime, then I think it would also hold for illegal immigrants.

What is it about immigrants, do you think, that makes them less likely to commit crime?

There are several things that I would point to. One is just a simple selection factor, as we would call it in the academic literature. What that means is that the pool of people that are coming into the country are selected on certain characteristics, such as wanting to get ahead in the United States, and that's associated with working hard, keeping out of trouble, keeping their heads down. Then it's not terribly surprising, if you think about it, that people who are coming here to better their lives would not necessarily be picking up and doing crime right away. It's a selection factor that makes a lot of sense.

And this is not just from Mexico, but if you're thinking about immigration from around the world. That's another factor, I think, that in any story about immigration has to be emphasized, that this is now something that is going on in all ethnic groups.

Secondly, and related to it, there's less incentive to commit crime, and greater sanctions, because of course one can be deported, and one doesn't want to draw attention to oneself.

Third, I think there's a family structure relationship here, in that the immigrants, at least in our data, are much more likely to be in intact families. In Chicago, for example, the Mexican-Americans are more likely to be married even than whites, and family structure is related to the risk of certain outcomes among offspring in our data and [that of other researchers]. So the fact that there are more intact families is, I think, part of the explanation, which of course also points out an irony in the anti-immigrant onslaught from the far right. David Brooks has written about this. If one views family intactness as "family values," then one would be actually in favor of more immigration. That's an interesting irony there.

Fourth ... increasing immigration has been associated with changes in cities in particular. We've seen this rebuilding and economic boom in part, I believe, due to immigration. That is something that would lead to lower crime.

And lastly, which is a little bit more speculative on my part, but I think it's an interesting hypothesis to consider, is that a lot of violence in American cities is associated with subcultures of honor. The notion [is] that a lot of homicides, in particular, derive from disputes over honor and a sense of perceived insult. And that sort of honor system or cultural system, if it exists, then to the extent that is diluted by immigration, then one would expect lower crime rates. At the very least, you would think that immigrants are less exposed to that, because of the tendency to settle in immigrant enclave areas. And in fact another irony in all of this is that there are many scholars who attribute the subculture of violence and honor of violence in American society to Scottish-Irish dueling culture that was brought over centuries ago. Now, that's also sort of a speculative hypothesis, but there are a lot of people who buy into it, the notion that the culture of violence is brought over by the white immigrants, the Scots-Irish, who settled in the South, South Carolina and other areas, and then diffused into the inner city.

Have you looked back at previous immigrant groups and seen whether they had similar patterns in terms of crime?

Yeah, the data are pretty consistent with this, and if you look back at the criminological literature, even going way back, you can see this as well, where first-generation kids were doing better.

So again, it's not a new finding, and that's why I think that the question really as to why this is generating the kinds of responses that you see, and that this video kind of captures, the fury, the question really is why the fury?

I've argued, and some of our data on another topic, really, but it bears on this, shows that there are these deep perceptions that are linked not just to immigrant groups but also to racial minority groups in the United States. [These perceptions] are tenacious, and they hold regardless of what the data actually tell us. So, for example, we've done studies where we've videotaped very slowly, going down the street, each side of the block, videotaping the amount of disorder -- for example, the number of people drinking in the street, or the number of broken windows, graffiti, vacant houses and so forth -- and also measured the crime rate in the community, and then asked people about their perceptions of the exact same thing that we're looking at, and we can objectively measure through videotapes.

What we find is huge, huge effects of the racial and immigrant composition of the population and what people perceive to be a problem. The way it works is that the more there is a concentration of blacks, Latinos and immigrants, the more people perceive disorder to be a problem. That's pretty interesting, and a bit sad.

Furthermore, we found that [perception] to be true [among] all groups. This is not a story about, "Oh, whites are the bad guys who perceive these groups to be linked to disorder." All groups, actually -- Latinos and blacks and whites, Asians -- were all exposed to the cultural stereotypes in the society, and so what's happening, I think, is that there's this implicit and deep-rooted sense [among all groups] that low-income and immigrant groups are inevitably associated with certain characteristics.

What's the reaction been to your research?

It's kind of been all over the map. I've gotten gratifying support and confirmation in the sense of scientific data from other studies, and I've been able to give a number of talks and I think present the data in a way that has hopefully shed some light on the beliefs in the country that have, I think, gone some ways toward counteracting some of the stereotypes. And of course I've gotten lots of angry diatribes as well from people who are angry, don't believe the data, think that it's untrue, that I have some kind of sinister agenda. I've gotten vile e-mails, hate mails, threats, that kind of stuff as well. So that's not been pleasant, of course.

Is it that they just don't want to believe that what you're saying could be true?

I think that's the biggest part of it. There's almost a sense of outrage because it's obviously not true, and therefore how could I say it? That's the feeling among some.

I'm talking about serious reactions. I've gotten reactions on the order of somebody who wrote something like, "You need to pass your research out so Mexicans can wipe their asses in the streets where they're shitting." That kind of stuff, you can't take that seriously.