The Wage Penalty

In March, Harvard economist George Borjas reported on what may be a major discovery about the wages of the 8 million undocumented workers in the United States.

Borjas was researching the wage penalty for workers without legal status. This is the difference between the pay received by immigrants who have authorization to work and the pay received by unauthorized immigrants with similar characteristics — in other words, it’s a measure of how much employers can keep wages down by exploiting workers’ lack of legal status.

Many variables are involved, so different researchers have come up with different results, but generally the studies have found a wage penalty that ranges from 6 percent to as much as 20 percent.

Borjas carried this research a step further by trying to determine whether the penalty has varied over recent years. Applying a technique developed by the Pew Research Center, Borjas analyzed data from the Census Bureau’s American Community Survey (ACS) to track the penalty each year from 2001 to 2014. He found that while undocumented workers suffered a penalty of around 9 percent in the earlier years — a result in line with what other researchers had found — the penalty started to diminish after 2007. By 2010, the penalty for male workers had fallen to 5.7 percent, down from 9.1 percent in 2005. By 2014 it was down to 3.4 percent.

The wage penalty for being undocumented “fell dramatically in the past few years,” Borjas concluded, acknowledging that his analysis was just “a first step,” since it depended on complex factors.

Still, if his analysis is correct, it has important implications for immigration policy. It could influence the debate on whether “more fear” or “more rights” works best for righting the wage penalty, bringing everyone’s wages up, and making the economy friendlier for workers.

An Argument Against Legalization?

George Borjas is a reputable economist, and we should take his findings seriously. At the same time, he’s a conservative whose political views can influence his conclusions. He frames his findings on the wage penalty in a way that seems to bolster conservatives’ arguments for “more fear.”

At issue is the depressing effect that low pay for undocumented workers has on the wages of other people — citizens and authorized immigrants — who work in the same types of jobs. There isn’t a lot of agreement among economists on how significant this effect is, but it’s important enough to be a concern for the low-income workers competing with the undocumented for these jobs — a concern often expressed in the claim that “they take our jobs.”

If lack of legal status is a major cause of the low wages unauthorized workers receive, it follows that a legalization program would raise their wages, and that just as their low wages now exert downward pressure on other workers’ wages, a new legalization program would produce upward pressure on those wages.

In other words, an immigration amnesty would benefit both unauthorized and authorized workers.

We rarely hear this common-sense argument for legalization, but it does get into the media on occasion. For example, in March 2016 Forbes ran a column by right-wing British writer Tim Worstall advocating legalization on this basis. Hillary Clinton made the same argument in a little-noted remark during her electoral debate with Donald Trump on October 19, 2016.

With his new findings Borjas was apparently trying to undercut the argument. “The small magnitude of the current wage penalty,” he wrote in his March report, “suggests that the enactment of a regularization program may only have modest effects on the wage of undocumented workers.”

But even “modest effects” on undocumented workers’ pay would be enough to exert at least some pressure on the wages of other workers in the same jobs. And Borjas may be underestimating the size of the wage penalty.

This is because of an inherent problem with the method we use to determine the penalty — the comparison of undocumented immigrants’ wages to those of documented immigrants. Like native-born workers, immigrants with work authorization experience a depressing effect from competition with unauthorized workers, so the authorized immigrants’ wages are lower than they would be without the presence of undocumented immigrants. The result is that the actual wage penalty is almost certainly greater than what we get by comparing authorized and unauthorized workers.

We can try to compensate for this problem by weighting the statistics, but there are other factors that limit the usefulness of authorized immigrants as a control group. For example, because many Mexican workers are undocumented, employers may suspect authorized Mexican workers of using false documents and therefore pay them less than they would pay white native-born workers.

A 2010 paper by sociologists Matthew Hall, Emily Greenman, and George Farkas includes a summary of research about this type of discrimination. The authors conclude that “if the wages of documented Mexican immigrants are penalized as a result of illegal immigration, our legal status wage differentials are likely biased downwards.”

It is even harder to quantify the other costs of unauthorized workers’ wage penalty for authorized workers in the same fields. Take unionization efforts in factories with a large undocumented workforce. A raid by immigration authorities can be a major setback if it occurs during an organizing drive. Maybe only undocumented workers are detained, but they may have been the voters that would have provided the margin of victory for the union.

An Economic Policy Institute study found that “[o]n average, a worker covered by a union contract earns 13.2 percent more in wages than a peer with similar education, occupation, and experience in a non-unionized workplace in the same sector.” Unions are the most powerful engine for elevated wages, and they are seriously threatened by Trump’s immigration demands.

An Argument for E-Verify?

Despite his conservative politics, Borjas’s March paper seriously undercut the Right’s arguments in favor of a major anti-immigrant measure.

In addition to analyzing the 2001–2014 wage penalty by year, Borjas studied its progress in different states. During this period four states — Arizona in 2008, South Carolina in 2010, Mississippi in 2011, and Alabama in 2012 — began requiring private employers to use the federal government’s internet-based E-Verify program, which searches government databases to catch workers with false documentation. Borjas found that for male workers “the state-level restrictions on undocumented employment seemed to have increased the wage penalty by around 40 percent.”

The idea behind E-Verify is that workers will stop immigrating here without authorization if we make it harder for them to get jobs. The obvious problem with this is that many of the low-paying jobs undocumented immigrants are likely to take are already in the informal economy, where employers are free to ignore government requirements for taxes, minimum wage, and other labor protections — and for verifying a new hire’s legal status.

As of 2010, about 56 percent of undocumented workers were working off the books, according to an estimate by the Social Security Administration. And we can expect that the main effect of E-Verify is simply to push still more undocumented workers into the underground economy.

A 2011 study by the Public Policy Institute of California indicates that this is exactly what happened with Arizona’s requirement for employers to use E-Verify. The authors’ estimates “suggest that wage and salary employment of Hispanic non-citizens dropped by approximately 56,000” as the requirement was implemented in 2008 and 2009. Many of these were undoubtedly undocumented immigrants who left the state, or failed to move there, because of E-Verify. (The authors didn’t try to determine how many left for their home countries and how many for friendlier nearby states like California and New Mexico.)

But the report found that “non-citizen Hispanic self-employment increased by about 25,000” during the same period. In other words, nearly half the undocumented immigrants who disappeared from the state’s formal economy reappeared as self-employed contractors — and we can assume that most of these were working off the books.

Since jobs in the underground economy generally pay less, this probably explains a great deal of the 40 percent increase that Borjas found in the wage penalty for the four states with E-Verify mandates.

Ironically, support for E-Verify isn’t limited to anti-immigrant right-wingers. Many liberals also back the program, from the New York Times editorial board to Mother Jones columnist Kevin Drum. Even Bernie Sanders seemed to accept E-Verify in his 2016 election campaign platform.

An Argument Against Obama’s Policies?

But the most interesting question raised by Borjas’s findings is why relative wage levels for undocumented workers started rising around 2008.

Borjas considered a few possibilities. For example, by 2014 some 66 percent of the undocumented population had been living in the US for a decade or more. Could unauthorized workers be getting paid more due to factors like additional years of work experience or increased English proficiency? But if that was the case, new arrivals wouldn’t be getting the same relative wage increase as the immigrants who arrived earlier. Borjas’s analysis showed no such effect.

Borjas also considered the possibility that the cause was “a favorable shift in the legal environment regarding undocumented immigration in the past decade, particularly during the years of the Obama administration.” One example, although Borjas didn’t cite it, could be President Obama’s DACA program. This might have contributed to the change by making work authorization available to hundreds of thousands of young out-of-status immigrants. But Quartz’s Gwynn Guilford, one of the few journalists to cover Borjas’s findings, noted that DACA wasn’t instituted until 2012, four years after undocumented wages began rising.

A better explanation might be a “favorable shift” in the attitude of the native-born population towards immigrants, which could translate into better treatment of undocumented workers by employers and others.

Gallup tracking polls from the period show that popular views of immigrants did in fact improve, although there are many variations in opinions about specific immigration policies. In June 2002, some 52 percent of respondents considered immigration a “good thing,” with 42 percent saying it was a “bad thing.” By June 2013, some 72 percent of respondents saw it as a “good thing,” while just 25 percent described it as a “bad thing.” CNN/ORC polls for the more recent period, 2015-2017, show a similar increase in respondents supporting legalization of the undocumented population.

But why did the native-born population become more welcoming to immigrants, despite harsh anti-immigrant rhetoric from many politicians and much of the media? After all, this was also the period that included the rise of Donald Trump and the so-called “alt-right.”

One reason might be simply that the undocumented population had stopped growing and the majority had become long-term residents. The US population generally seems unaware of this change — after all, our media rarely stresses it — but many native-born citizens are personally affected by it. The undocumented are now members of the community, people that with whom citizens work, socialize, and raise families. They have lost much of the “other” quality that feeds xenophobia.

Or an Argument for Organizing?

Another factor, often overlooked, could be activism by immigrants themselves.

The nation was taken by surprise when immigrants mounted massive demonstrations in the spring of 2006. This upsurge in immigrant activism continued in a wave of organizing by the young activists often referred to as “Dreamers,” many of them openly declaring themselves “undocumented and unafraid,” in conscious emulation of LGBT activists who had raised public consciousness by “coming out.”

Back in 2006, commentators warned that immigrant activism would provoke a backlash — the same warning given African-American activists in the 1950s and ’60s. Politicians told immigrants that they should focus instead on supporting efforts in Congress for a bipartisan “comprehensive immigration reform.”

While the 1960s civil rights movement did in fact provoke a backlash, it also had the effect of making large numbers of whites conscious of race-based oppression within their society, and of impressing them with the courage and commitment of African Americans who risked so much, including their lives, in the fight for justice and dignity. Isn’t it likely that immigrant activism has had a similar effect?

There’s certainly evidence of at least some impact on the labor movement. City University of New York sociology professors Ruth Milkman and Stephanie Luce wrote in April that the immigrant protests in 2006 and the 2011 Occupy Wall Street movement were “two key turning points” in labor’s recent shift towards supporting “alt-labor” efforts — the nonunion, community-based labor organizations known as workers centers and organizing campaigns like the Fight for 15. These efforts have concentrated on low-wage workers, and immigrant workers have played a particularly active role in them.

The number of US workers centers “grew substantially in the aftermath of the Great Recession,” Milkman and Luce noted, from 167 in 2006 to over 200 by 2010 and to as many as 230 by 2013.

While the organizing campaigns haven’t resulted in higher union membership, they appear to have had an impact on wages. More than 2 million workers are estimated to have benefited from minimum wage increases in states, cities, and other localities since 2012, while nearly 1 million benefited from minimum wage hikes by large employers during the period.

In addition to the impact on wages, there’s a secondary, less tangible effect: the sense of solidarity that comes from participation in organizing drives which unite authorized and unauthorized workers in collective struggle.

Immigration reform efforts in Congress have repeatedly failed. The changes in public attitudes towards immigrants and the apparent rise in undocumented wages suggest that immigrant activism — we could call it “immigration reform from below” — has been more successful.

The cutoff year for Borjas’s research was 2014. We don’t know how undocumented workers’ wages have fared since then. It certainly seems possible that they could be falling now as immigration raids and Trumpist rhetoric drive immigrants back into the shadows.

But this isn’t just an academic question, and we aren’t just observers. The experience of the previous period suggests that what makes the difference — in immigrants’ wages, as well as everyone else’s — will be the level of activism and militancy embraced by immigrants and their allies.