[Immigration, Labor]

Immigration's Effect on American Labor.

I run and work on a small residential construction crew, my people make about 50% what I made as a crew member 20 years ago, while there are other factors at work, the biggest and most addressable one is cheap illegal labor. Neither my crew, nor I can afford much sympathy that doesn't address our plight. If that sounds hard, try making a living doing this.
--Chuck Butcher, Democratic Candidate for Oregon District 2

"Finally, comprehensive immigration reform requires a temporary worker program that will relieve pressure on our borders. This program would create a legal way to match willing foreign workers with willing American employers to fill jobs that Americans will not do."
--GW Bush, Saturday

With regard to immigrant labor, two views depend on very different realities about what it's like in the labor market. Dubya hews the popular bidnez line, in which cheap labor is great for Latin Americans and US citizens because the former get better jobs than can be found in their home country, while the latter enjoy cheaper prices and competitive businesses. But folks like Chuck have seen an entirely different reality: because illegal aliens have no protection and must work at exploitative wages, employers are less willing to pay fair salaries, and worker salaries across the board are driven down. And, in the marketplace, businesses that pay fair wages put themselves at a fatal disadvantage.

So which is it?

According to the nonpartisan but mostly anti-immigrant think tank The Center for Immigration Studies, Chuck's view is the clear winner:

• Of the 900,000 net increase in jobs between March 2003 and 2004, two-thirds went to immigrant workers, even though they account for only 15 percent of all adult workers.
• In just the last year, 1.2 million working-age natives left the labor force, and say that they are not even trying to find a job.
• The decline in native employment was most pronounced in states where immigrants increased their share of workers the most.

Ah, but will Chuck's view hold up to more Neutral scrutiny? Yes, as two Harvard economists, George Borjas and Alexander Katz, demonstrate. They find the volume and effect of Mexican labor (to which their recent study was confined) to be a serious effect on American labor. The following is from a Robert Samuelson article on their research:

Among men, about one in 20 U.S. workers is now a Mexican immigrant; in 1970, that was less than one in 100. The vast majority of Mexican workers lacked a high-school diploma in 2000 (63 percent for men, 57 percent for women). Only a tiny share had college degrees (3 percent for men, 5 percent for women). By contrast, only 7 percent of native-born U.S. workers were high-school dropouts and 28 percent were college graduates in 2000. Mexican workers are inevitably crammed into low-wage jobs: food workers, janitors, gardeners, laborers, farm workers. In 2000, their average wages were 41 percent lower than average U.S. wages for men and 33 percent lower for women....

For today's Mexican immigrants (legal or illegal), the closest competitors are tomorrow's Mexican immigrants (legal or illegal). The more who arrive, the harder it will be for existing low-skilled workers to advance. Despite the recession, immigration did not much slow after 2000, says Camarota. Not surprisingly, a study by the Pew Hispanic Center found that inflation-adjusted weekly earnings for all Hispanics (foreign and American-born) dropped by 2.2 percent in 2003 and 2.6 percent in 2004. (Pdf of the original study here.)

When Borjas and Katz dig directly into the question of whether low-paid immigrants lower the wages of native laborers, they uncovered an interesting quirk in the data. They begin with past studies, which have tended to support the Bush thesis, that immigration is good for Americans and has no effect on native laborers:

There is a great deal of dispersion in the findings reported by the various studies in thisempirical literature. Nevertheless, there is a tendency for the estimated cross-city correlations to cluster around zero, helping to create the conventional wisdom that immigrants have little impact on the labor market opportunities of native workers, perhaps because “immigrants do jobs that natives do not want to do.” It would seem, therefore, that a fundamental implication of the standard textbook model of the labor market—that an increase in supply lowers wages—is soundly rejected by the data.

But Borjas and Katz find troubles with the way these statistics were measured (averaging national data), and Borjas has a developed a new measure. His findings? Not only did immigration lower low-skill native wages, but it didn't improve high-skill native wages, either, as promised by Bush and Co.

The second column of Table 11 shows the predicted labor market effects when the supply shocks are given by equation (1). Mexican immigration, which is predominantly low-skill, accounts for all of the adverse impact of immigration on low-skill native workers. It is also worth noting that the earnings of college graduates would have fallen by 3.9 percent if there had been no Mexican immigration, as compared to the 3.8 percent decline that occurred with the actual flow. In other words, the influx of low-skill Mexican immigrants barely improves the wage of high-skill workers.

There are a number of other reasons why illegal laborers damage native laborers. As Chuck mentioned in his comments to yesterday's posts, immigrants disproportionately use social services and education while contributing little to the tax base. Because low-wage American workers depend on those social services, this is a double whammy--wages are falling, and so are available services.

The guest-worker program is good for large business owners, who can cut costs and achieve a competitive advantage with low-wage labor. But for American workers, there's no upside. Bush is wrong about immigrants taking only jobs no one will do; worse, because they're willing to do that work so cheaply, American workers have seen their own wages fall, just as Chuck predicted.

This post is getting long, so I'll defer thoughts about what can be done until later.