It's hard to get accurate numbers on immigration, harder than it was 100 years ago. Then almost all immigrants came across the Atlantic and arrived at ports, where they could easily be counted. No one bothered to see how many people crossed the Rio Grande and the Arizona desert, and there could not have been that many: The territory was almost entirely uninhabited. Today immigrants come from all directions and by all modes of transportation, including on foot, and many are here illegally. So no one can know the numbers for sure.
But we can try to get good estimates, and that is what Roberto Suro and Jeffrey Passel of the Pew Hispanic Trust have done in their just released report on immigration from 1992 to 2004. Here are some highlights:
Suro and Passel report that immigration remained at roughly constant levels, between 1 million and 1.25 million, between 1992 and 1997. It spiked to 1.46 million in 1998, 1.53 million in 1999, 1.55 million in 2000, and 1.39 million in 2001, the great bulk of that presumably before September 11. Then it continued at the levels of 199297: 1.17 million in 2002, 1.10 million in 2003, 1.22 million in 2004. In other words, when the economy heated up in the late 1990s, immigration increased significantly; when the economy cooled down, and new restrictions on entry were put into place, it went down. Nothing terribly surprising here. As Suro and Passel point out, immigration levels were highly responsive to the economic cycle in the heavy immigration years of 18401924, so they are responsive to the (much smaller) oscillations in the economic cycle today.
Illegal immigration continues at high levels: 38 percent of all immigrants in 199297, 42 percent in 19992000, 43 percent in 200204. There's a spike upward, by the way, in 2004, but the total number of illegals then was well below the levels of the immigration spike years of 19982000. Compared with the early 1990s, we have fewer legal permanent immigrants and legal temporary migrants and more of what Suro and Passel call "unauthorized migrants."
Where do immigrants come from? In all these years the answer is pretty much the same: about one third from Mexico, about one fifth from the rest of Latin America, about one quarter from Asia, about one fifth from the rest of the world.
Where do they go? Here there are fairly big changes. In the 1980s, by my recollection, about three quarters of immigrants headed to one of just six states: California, New York, Texas, Florida, Illinois, and New Jersey. In effect this was migration to just a few large metropolitan areas: Los Angeles and to a lesser extent San Francisco, New York, Houston, DallasFort Worth, Miami, and Chicago. Now immigrants are spreading out. The percentage going to these "big six" states fell from 66 percent in 199297 to 59 percent in 19992000 to 57 percent in 200204. California, New York, Illinois, and New Jersey are receiving significantly fewer immigrants than in the early 1990s; Texas and Florida, significantly more. There have been even larger increases in the numbers going to several other states in the South (especially Georgia and North Carolina) and the West (especially Arizona, Nevada, and Colorado). These are all places with big metropolitan areas that are growing faster than metro Los Angeles and New York.
The policy implications? The labor market is obviously drawing far more immigrants here than are permitted by our immigration laws. We need to improve enforcement and to change our laws so they can operate more in tandem with the labor market. Immigration is flowing, as it did in the 1980s and 1990s, to the metro areas that are growing fastest. If we want rapid growth, then we probably want levels of immigration approximating what we have today. One policy recommendation that I remember someone making around 1980 that we obviously don't need: creation of a government agency to advise immigrants which metro areas were growing and creating new jobs. The immigrants figure that out better than government ever could.