The best picture of Marco Rubio that does or can ever exist. (Phil Coale/AP)

Sen. Marco Rubio (R-FL) got in hot water earlier this week when one of his aides told Ryan Lizza that "there are some American workers who, for lack of a better term, can't cut it." He was promptly denounced by all sides for the remark. The American worker, after all, is also the American voter. The only problem is that the aide was totally right.

First, let's be clear about what exactly was said. The broader context was the Rubio aide's rebutting the idea that we shouldn't adopt immigration reform because of high unemployment. Ryan Lizza has posted a screenshot of the transcript:

Via Ryan Lizza.

So the question to answer in evaluating the aide's statements is not "Are there jobs that Americans won't do?" As Jonathan Chait notes, Americans will do all kinds of painful, disgusting work if you pay them enough:

The issue is whether adding more immigrants to the population will make unemployment increase. While it's true that most any job will be done if the wage is high enough, many jobs simply won't exist above a certain wage. For instance, while many Americans would pick strawberries or gut catfish for $50-an-hour, in practice, raising wages that high would just mean we import strawberries and catfish from countries with far lower wages. Similarly, in a world without immigrants willing to clean homes for fairly low wages, more people would simply clean their own homes.

The key question is whether immigrants in jobs like that are complementing the work of Americans, or substituting it. Substitute goods are ones that perform the same function, or roughly the same function. Margarine and butter are the classic examples of substitute goods. You use both of them to add fat to stuff. When butter gets more expensive and margarine doesn't, more people buy margarine. If immigrants are substitutes to native-born workers then the more immigrants there are the fewer native-born workers employers will need.

Complementary goods, by contrast, go together, and when the price of one goes down, demand for the other goes up. Peanut butter and jelly are complements; when peanut butter gets cheaper, people tend to buy more jelly to go along with it. So the question is whether native-born Americans and immigrants are more like margarine and butter or more like peanut butter and jelly.

If the former, then adding immigrants just adds people to compete for the same positions, driving down wages. That's not a slam-dunk case against immigration; even the most avid proponent of the idea that immigrants and native-born workers are substitutes, Harvard economist George Borjas, finds that immigration increases wages for Americans overall, though he finds wage reductions for high school dropouts and college graduates. And obviously immigration increases the wages of immigrants. But there are real losers in a substitutability scenario.

Immigrants specialize in manual labor, like the migrant workers in this scene from PT Anderson’s "The Master" do, freeing up native workers for more intellectually stimulating work. (The Weinstein Company)

Not so in a complementarity scenario. That's why Bocconi University's Gianmarco Ottaviano and UC-Davis's Giovanni Peri caused a stir in 2006 when they estimated that immigrants are actually complementary to American workers. There's micro-level data to back Ottaviano and Peri up. The Current Population Survey finds big differences in the occupations of native-born vs. foreign-born workers. The economist Jack Strauss has found that cities with larger Latino immigrant populations see lower poverty and higher wages among African Americans, evidence that immigrant workers and native-born African Americans complement each others' skills, rather than substitute for each other.

Ottaviano and Peri have also looked at specific task specializations to prove their point. They, along with the University of Essex's Greg Wright, have found that within manufacturing, immigrants tend to perform low-complexity tasks while native workers perform more communicatively and cognitively intense work, resulting in very little competition between the two. Peri and Colgate's Chad Sparber found that, across the economy at large, immigrants tend to specialize in manual or physical tasks while native workers specialize in communication-heavy work.

"Workers’ skill endowments imply that immigration reduces the compensation paid to manual tasks and increases the compensation paid to communication tasks," Peri and Sparber write. "The complementary nature of the two skills and the reallocation of native workers toward communication tasks favor wages paid to native workers. The effects compensate (in part or entirely) for the depressing effect of immigration on the wage paid to manual tasks."

And it doesn't only compensate for the wage effects of immigration. It explains why the unemployment impact is so small. The CBO projects that the Gang of Eight immigration bill will raise unemployment by, at most, 0.1 percentage points between 2014 and 2020, and have no effect thereafter. An analysis by Rob Paral & Associates for the Immigration Policy Center found no statistically significant relationship between unemployment rates and immigration at the regional, state, or county level.

To be sure, the claim that U.S. and foreign workers aren't perfect substitutes is disputed by some reputable economists. And there are clearly some jobs for which they are perfectly substitutes. It's not clear that there's any reason why either native or immigrant workers would be better suited to washing dishes, for example. But the number of Americans who aren't capable of higher-paying work than dish-washing is, mercifully, quite low. If more Americans work in fields where their labor is complemented by immigrants than in fields where Americans and immigrants can be substituted for each other, then substitutability isn't a reason for Americans as a whole to oppose immigration.

Moreover, even if all immigrant labor, regardless of the sector, were perfectly substitutable for native labor — a claim that even some economists who doubt Ottaviano and Peri's conclusion don't make — it's worth remembering that the number of jobs aren't fixed. Immigrants draw paychecks, but they also pay rent, buy groceries, use laundromats, go to the movies, and so on and so forth. Those activities all increase economic growth, creating jobs both for other immigrants and for native-born Americans. That's yet another force preventing immigration from increasing unemployment.

What Rubio's aide said was impolitic. A statement along the lines of, "Immigrants take crappy jobs and free up Americans to take more intellectually challenging ones," would probably poll better. But the central point of Rubio's aide, that the effect of immigration reform on unemployment is, as a result, minimal, is absolutely right.