The Washington PostDemocracy Dies in Darkness

North Carolina needed 6,500 farm workers. Only 7 Americans stuck it out.

Freddie Quell works at a cabbage farm alongside migrant workers in PT Anderson’s “The Master.” (The Weinstein Company)

When I talked to him about the economic effects of immigration last month, Center for Global Development migration expert Michael Clemens mentioned that he was working on research on agricultural migrant workers. That research is finally out, in the form of a report released by CGD and the Partnership for a New American Economy, a pro-immigration reform group started by Michael Bloomberg and Rupert Murdoch.

Clemens's study focuses on the agricultural industry in North Carolina, and more specifically on the North Carolina Growers Association (NCGA), which supplies manual laborers to North Carolina farms. The NCGA is the nation's largest user of the H-2A guest worker program, which is designed with agricultural workers in mind. Under that program's regulations, Clemens explains, NCGA "must submit an application to the US Department of Labor proving that it has actively recruited US natives and native workers will not take NCGA jobs."

That data is interesting, because it describes the labor market before any immigrant workers are recruited. That, as Clemens says, "allows us to assess the willingness of native workers to take farm jobs before they can even be offered to foreign workers, meaning that this study does not miss any impact caused by people who self-select out of an area or occupation because of competition with foreign workers."

That willingness, he finds, is basically nonexistent. Every year from 1998 to 2012, at least 130,000 North Carolinians were unemployed. Of those, the number who asked to be referred to NCGA was never above 268 (and that number was only reached in 2011, when 489,095 North Carolinians were unemployed). The share of unemployed asking for referrals never breached 0.09 percent.

When native unemployed people are referred to NCGA, they're almost without exception hired; between 1998 and 2011, 97 percent of referred applicants were hired. But they don't tend to last. In 2011, 245 people were hired out of 268 referred, but only 163 (66.5 percent) of the hired applicants actually showed up to the first day of work. Worse, only seven lasted to the end of the growing season:

For perspective, the NCGA was looking for about 6,500 workers in 2011. Even if every single referred applicant had been hired and finished the growing season, that would have filled only 4.1 percent of the slots available. As a consequence, the vast majority of the workers who start work at NCGA farms are Mexican H-2A visa holders:

And Mexican workers are far likelier to stick through the season than native-born workers. About 90 percent were still working five months along, compared to less than 10 percent of native-born workers:

The obvious counterargument to this analysis is that there are, by definition, no jobs Americans won't do no matter what. If agricultural labor paid $100 an hour, people would be lining up to take these jobs. So the alternative to foreign work, by this line of reasoning, isn't that these jobs are just disappearing, but them being offered at a higher wage.

It's not so clear that's the case here. For one thing, Clemens looks at a labor market where employers are only using legal labor, which under the H-2A regulations means they have to pay the same wage to native and non-native laborers (a wage set by the government, according to the type of work and the geographic area where it's being done), and thus can't benefit from paying lower wages to migrant workers. If even under that standard, Americans aren't applying for the jobs, even when they have automatic preference over foreign workers, that tells you Americans are really avoiding them.

But surely it's possible that, if H-2A didn't exist, the prevailing wage for agricultural work would rise above where it's estimated now, and thus attract more native workers, right? Sure, but it's unclear even that would make a difference.

For one thing, the share of workers completing the growing season didn't increase during the recession, during which wages for NCGA jobs grew relative to other wages in the economy. "The increase in unemployment mirrors a loss of wages generally in the economy — decreased availability of jobs everywhere else in the economy should be reflected in an increased interest in the NCGA farm jobs because people’s ability to earn income across the broader economy is more limited, making these jobs more valuable in comparison," Clemens explains. That suggests that similar in scale wage increases wouldn't spur increased demand for these jobs on the part of native workers either.

Further, if wages increased too much, these businesses wouldn't be profitable and all the jobs would disappear. Clemens notes that this appears to be the case for cucumber farming. "If wages were doubled, raising the hourly wage for collecting cucumbers from $9.70 to $19.40, it would be impossible for farmers to grow cucumbers profitably – the farmers’ labor cost per acre per year would be too high," he writes.

Neither Clemens nor the organizations putting the report out are exactly neutral parties in this debate, of course, but the data they present is fairly compelling. It seems clear that it would take a quite large increase in agricultural wages to get native workers to do these jobs, an increase that could very well put the farms in question out of business. Given that, making it easier to bring in agricultural workers appears to both dramatically benefit those workers and make life slightly more convenient for the industry in question.