 Our current estimates from the Pew Research Center are 11 million. They represent about one-quarter of all the migrants in the U.S., so there's about 44-45 million in the United States. And 11 million are what we call unauthorized migrants. The number reached as high as about 12.2 million in 2007. Our estimates show it dropped about a million after the recession by 2009. And since then it's been stable with a little bit of a downward drift, but we estimate about 11 million right now. The basic logic of it is actually pretty simple. We have data from our censuses and surveys. There's a big survey that the federal government does called the American Community Survey that's used to estimate the total foreign born population. We make an estimate of how many people are residing legally in the United States as immigrants using principally data on people admitted to the United States. And basically we subtract the total minus the number in the country legally, and that difference is the number in the country without authorization. There's a little bit more to it than that, but that's the basic logic of the approach. When we generate the estimate we put what we call a margin of error around it, and it's based on the idea that it's from a sample. Our margin of error is that it could be 150,000 higher or lower. There are other kinds of potential errors in the estimate, and it could probably be as much as a million higher and maybe half a million lower, in my opinion. But our margin of error is 150,000. First let me be clear, Mexico is about half of the total. We just aren't sure it's over half anymore. So we estimate about five and a half million unauthorized Mexicans in the country. The next biggest country we estimate is El Salvador at 700,000. So Mexico is still by far the largest number. What has happened, though, is that the Mexico number used to be close to 7 million, and it's dropped considerably in the last seven years. I think there's a number of factors. The job situation in the United States, which has always been the primary factor drawing Mexicans to the United States, after the Great Recession we had in 2007, 2008 and 2009, there were no jobs for Mexicans. A lot of the Mexicans in the country at that point went home, particularly the ones who had recently arrived. And the number coming from Mexico has dropped considerably. It's lower than it's been in the last 20 years. And that's due to, first, the lack of jobs. And second, I think the enforcement at the border has made it more expensive and more difficult for Mexicans to come to the United States. And in northern Mexico, there's a lot of drug-related violence, so it's dangerous in Mexico, not just to cross the border, but just to get to the border. And I think those are the big factors of one other. Mexico's undergone a tremendous drop in fertility in the last 50 years. In 1970, the average Mexican woman had seven children, and now it's dropped to a little over two. And it takes a while demographically for that to work its way through the population, but what it means is that in the age group of the late teens to early 20s, there's not nearly as many Mexicans as there were 20 years ago. So there's not as many potential migrants either. So I think all of those factors are working together. Well, in the United States, I think the big holes in the estimation process have to do with how good our surveys are. How many people are left out of the surveys? We've been able to make estimates, and we're fairly confident of them, but it would be nice to have better information on survey undercount, survey coverage of both legal and unauthorized. The other thing I think would be very helpful is more direct data collection from people who are known to be unauthorized immigrants about their characteristics. It doesn't necessarily have to be a universal survey, but it would be helpful for those of us who are modeling what unauthorized immigrants look like to have more data collected from unauthorized immigrants directly. The experience I've had is that as we've been able to get more and better data on the immigrant population and how many are in the country, how many immigrants are in the country, not how many are undocumented are in the country, demographers have been able to improve the methodology to separately estimate legal and unauthorized immigrants. When I started doing this, which was 35 years ago now, there was a very wide range of estimates. People were estimating 6 to 12 million unauthorized immigrants in the United States, and at the time there were probably only 2 million in the country. So as we've had more data and better data, the estimates have gone up, but the range of estimates has narrowed, and we've finally reached the point where the high estimates and the low estimates are converging. So I think better data on immigrants and better data that researchers can use can improve the discourse about how many immigrants are in the country.