In February we estimated that 300,000 construction jobs could go to illegal immigrants as a result of the stimulus bill. We stand by this number as a reasonable estimate of how many stimulus-related jobs could go to illegal aliens.
Some have taken the view that it is impossible to know how many stimulus-funded jobs might go to illegal immigrants. This way of thinking misses the point of how an estimate can inform public policy. We would never argue that our estimate is precise, but instead, as our press release stated, this is an "estimate" of jobs that "could" go to illegal immigrants. In fact, the headline of our press release is followed by a question mark to emphasize that the number is an estimate of what could happen.
Just as we cannot know the exact amount of oil in Alaska or the exact number of homeless people, we cannot know the exact number of illegal immigrants who will receive stimulus jobs. But, as is the case in these and countless other examples, we do not have to throw up our hands and say the question is simply unknowable. We can, in fact, make a reasonable and plausible estimate of illegal immigrants and stimulus jobs based on the available data. There are, of course, several challenges in creating such an estimate.
To be sure, it is difficult to determine how much construction and infrastructure spending there is in the spending bill. It's 1,073 pages with construction and infrastructure spending spread throughout. The Association of General Contractors of America (AGCA) states the total is $134.6 billion. We feel our estimate of $104 billion is actually quite conservative. In fact, the AGCA states that there is also another $8.8 billion for school construction in the bill on top of the $134.6 billion they identify. The fact that they list education spending separately highlights a key problem when reading the bill — schools may be build and repaired by construction workers, but should such work be counted as a construction job or education job?
On the question of how many jobs are created for each $1 billion dollars spent on construction and infrastructure, we use research paid for by the Federal Highway Administration, which placed it at 19,584 per billion. We use this estimate because a significant share of the jobs were for highway construction. Stephen Fuller, a professor at George Mason University has his own estimates that show that each $1 billion in construction spending on an annual basis creates a little over 14,000 construction and construction supply jobs. Other estimates exist as well. If we use the AGCA estimate of construction with Fuller's lower job creation numbers it would still come to nearly 2 million jobs.
The key point is that we do have an idea of how many jobs will be created by the stimulus bill, even if there is debate over the exact number. We also have an idea of how much construction spending there is in the bill and we have an idea of what share of workers in construction are illegal immigrants. The fact that these things are all subject to debate does not mean we cannot make a plausible estimate.
One final point is worth mentioning. Some who do not study the demographics of immigration seem to think that the recent decline in the illegal immigrant population makes it impossible to estimate how many illegal immigrants remain in construction. There is certainly good evidence that the illegal immigrant population has declined recently. But, the available data indicates that their share of construction workers has remained about the same. The January 2007 Current Population Survey (CPS), collected by the Census Bureau, showed that 28 percent of construction workers were foreign born. The January 2009 CPS, collected just two months ago, showed it was 26 percent. Given the margin of error in the survey, these two numbers are not statistically different. The CPS is one of the primary data sources used by researchers to measure illegal immigration. Our research and that of the Pew Hispanic Center shows that somewhat over half of the foreign born in construction are in the country illegally. If the illegal share of construction workers had declined dramatically, then the overall foreign born would have declined dramatically in the new 2009 data. Of course, the CPS does show a substantial loss in the overall number of construction jobs, but the losses seem to have hit US-born workers and foreign-born workers equally.
It's worth noting that the federal government publishes estimates of not only the size of the illegal population, but where they live, from what country they came, what year they arrived, etc. Others have looked at illegal-immigrant tax payments, employment and use of social services. Our analysis simply builds on this work.
If a blogger or reporter gives out our estimate as a hard and fast number, as if it were the number of students in a local high school, then that is a mistake. But if the numbers are treated as an estimate of jobs that could go to illegal immigrants, then it can help to inform the public discourse. This is the whole purpose of any policy-oriented research in areas where information is limited.