The New York Times' lead story this morning made an interesting case for what it sees as a decline in illegal migration from Mexico to the U.S.
But it played down a central element of the picture: the fact that the U.S., with its widespread poverty and huge wages gaps between the rich and the poor, is rapidly getting to be more like Mexico than in the past, so the Mexican poor no longer have as many reasons to want to come here. Being sensible, they stay home.
In addition to that, the article frays a bit when parts of it are examined by a microscope.
The report by Damien Cave was obviously months in the making, and I, for one, am grateful that the Times still engages in expensive investigative reporting. The report of the truly remarkable drop in Mexican birth rates was particularly valuable, as were the reports of expanding educational and economic opportunities in that country.
The whole tone of the long article, however, was that improving conditions in Mexico were the main reasons for the drop in emigration, and that U.S. enforcement activities had relatively little to do with the trend. So, the article seemed to be hinting, the U.S. need not do anything drastic like really enforcing the immigration law. If that is the implication, I disagree with it.
When the Times does a big story, like this one, it always makes sure that there are photographs and charts to illustrate and strengthen it. Today's report was no exception.
Across a full page inside the paper was a chart headed "A Decline in Illegal Immigration from Mexico." It looked like a mountain range in profile, and ran from 1950 until 2009, an impressive time series with a steep and steady decline from 1999 to 2009.
If, like most readers, you just let the image sink it, it appeared persuasive. It may well reflect a bit of reality, but a careful reading of the fine print and the use of a little math make it far less convincing.
First, let's see what we are measuring. What the graph estimates is number of Mexicans entering U.S. illegally for the first time and not the accumulated Mexican illegal alien population in the U.S. (It's like big business talking about the H-1B program in terms of 65,000 first-time admissions a year, not the probable H-1B population of 650,000 I estimated earlier this year.)
This then is a flow estimate, and flows of people are always more volatile than total population figures; if you chart flows, you are sure to get more mountains and valleys than if you study total populations. And total populations are much more important than yearly fluctuations in flows. So the whole chart must be taken with a grain of salt, which is not included in the Times recipe.
Next you need to read the fine print in the text. The chart is not based on anything as massive as the annual arrests of illegals at the Mexican border, which are also down in the recent years. It is based on a series of annual interviews, over the years, with 800 to1,000 households, I think in a single Mexican state, where emigration has been a major factor over the years.
Then you have to read the fine print in the chart itself. The range, along the left side of the chart is from (implied) 0.0 percent to 1.0 percent (which is stated); when you have one percent of 1,000, what do you have? Well you have 10 individuals. The chart, in short, shows the first time illegal emigration from these 1,000 households as ranging from zero to 12.
Supposing both the mythical Carlos and Pedro, young cousins from two of the surveyed households, both come down with terrible colds at the start of the migration season and do not make their first migration that year, well, the percentage would drop from say 0.5 percent to 0.3 percent and the chart would show a very sharp drop.
One should not project migration or population trends using such a tiny base, or better, one should use several statistical methods, with this as simply one of them.
Another frayed piece of the picture is in Cave's rather heightened emphasis on a (really tiny) reason for the decline in illegal migration: the expansion of the H-2A nonimmigrant program for workers from Mexico. The text adjacent to one of the lesser graphs in the article, says: "nearly 250,000 Mexicans received H-2A (visas for agricultural guest workers) in the years between 2006 and 2010 – a 75 percent increase over the previous five-year period."
Interesting, but 50,000 a year in annual admissions is not terribly important when you realize that the total estimated illegal Mexican population in 2009, according to the Office of Immigration Statistics was 6,650,000, or 133 times the number of H-2A annual admissions from Mexico.
Further, my experience with the post-WWII Bracero program would suggest that the 50,000 coming in one year consisted of about 40,000 repeaters at most, and 10,000 or so new people. (I was Assistant to the Secretary of Labor for Farm Labor in the mid-1960s.)
Incidentally, Cave said that the H-2A program was "one of the few visa categories without a cap."
Hardly. There are, as we have noted, oodles of programs allowing nonimmigrants to work in the U.S. with a variety of letter and number combinations (F-1, H-1B, H-2A, H-2B, J-1, L-1, O-1, R-1 etc.), most with little or no protections in place either for the alien workers or for the resident workers they often displace. Some parts of the H-1B program for high tech workers, and all of the H-2B (non-skilled, non-ag workers) program, do have ceilings, the rest do not. It is the programs for immigrants, a totally different category, where numerical ceilings are more likely.
In summary, an interesting piece of work but its effect – if not its purpose – is to say, "Don't worry America, things are taking care of themselves."