Why liberals shouldn't turn against immigration

Dean Baker easily could have gone into academia, or sold out and gone on to make big bucks consulting in the private sector. Instead he did neither, choosing to work in the less prestigious, less-highly-paid world of liberal Washington think tanks. The reason is that he cared about the American working class, and wanted to dedicate his life to fighting on their behalf. This is extremely laudable.

However, that doesn't mean I always agree with Dean's ideas for helping the working class. In the last month, he has written a number of pieces with an anti-immigration slant. For example, on Feb. 2, he wrote this:

It's true that a declining population means that labor will be in shorter supply. That means that the least productive jobs will go unfilled. That is the way economies develop and the reason that half of our workforce is no longer employed in agriculture. In the U.S. this would mean that we might have fewer restaurants and the convenience stores won't be open all night. In Japan, perhaps they won't be able to find workers to shove people into the subway cars in Tokyo. What's the problem?... 

This is not to argue against immigration or immigration reform. The way we treat people who came to this country to work...is an outrage and is bad for the economy. However readers deserve a more serious discussion of the issues involved... 

Some immigration to the country undoubtedly provides economic benefits. However in nearly all cases there will be winners and losers. For example, a large flow of immigrants at the low-end of the labor force will hurt the people who have recently immigrated to the country. Some of us may not consider that a good thing.

On January 19, he wrote a piece saying that China's declining working-age population is good for wages (implying that the same would be true for us). And on February 26, he wrote an article for the Guardian, again calling for shrinking working-age populations:

There is no reason why the prospect of a stagnant or declining workforce should concern the vast majority of people... 

The people who hire help – the very same who also dominate economic policy debates – are terrified over the prospect that they will have to pay workers more in the future. 

But the rest of us can sit back and enjoy watching them sweat as ordinary workers may finally start to see their share of the gains of the economic growth of the last three decades.

The anti-immigration implications of this argument should be extremely clear. If shrinking labor forces lead to higher wages, then large-scale immigration will depress wages. Hence, we should consider immigration restrictions as a way to boost working-class wages. Some other liberal think-tankers, such as Jared Bernstein, have been writing pieces in the same vein. I call this view "labor protectionism".

There are four big, big reasons why I think that "labor protectionism" is a very bad idea, and should be dropped from the policy platform of American liberalism.

Reason 1: Agglomeration Economies

We are not living in an Econ 101 world, in which most people are employed as farmers, working the land. Instead, we live in a world with "increasing returns to scale", where economic activity gets a productivity boost from being concentrated. We also live in a world with transport costs, where companies want to put their offices and factories where the workers and customers are, and workers and customers want to live close to where they can get goods cheaply. These facts are summarized by Paul Krugman's theory of "New Economic Geography," for which he won the Nobel Prize.

But also, we live in a world in which knowledge economies are of extreme importance; when people cluster together in cities, their productivity goes up because of the accidental (and purposeful) exchange of knowledge and ideas.

For these reasons, the vast bulk of America's GDP is produced in extremely densely populated cities. Check out this infographic:



A casual glance shows that America's output is highly concentrated in big cities like New York and L.A. Studies confirm that living in a big city increases a worker's productivity by a substantial fraction.

If we were to kick a bunch of working-class people out of NYC, would NYC wages rise for the working-class people who were left, because of an artificial shortage of low-skilled labor? Maybe. But if we did this too much, NYC would start to lose its huge productivity advantage, and the tactic would backfire, reducing the wages of the working-class people who remained (to say nothing of the people who got kicked out).

Now realize that our modern world is very globalized. Much of our economy depends on trade (in fact, much more than the percent of GDP directly taken up by trade!). In the global economy, agglomeration economies mean that capital will flow to the country where the most productive workers and the densest, highest-purchasing-power markets reside. In other words, immigration helps us remain at the center of the global economy.

Reason 2: Politics

As you may have noticed, America's more liberal major party just won a historic electoral victory. This victory has many Republicans and conservatives fretting that liberalism has won in the U.S. But it's not the margin of Obama's victory that is producing this hand-wringing; rather, it's how Obama won. Specifically, Obama garnered 71% of the Latino-American vote and 73% of the Asian-American vote. These are the country's two fastest-growing demographics. If they continue to lean so strongly Democratic, the GOP is looking at a long, long time in the wilderness.

But why did Latinos and Asians abandon the GOP? The Republicans used to get a much larger share of both groups' votes. Many believe that the strongly anti-immigration views and policies of the Tea Party GOP in the last few years sent a clear message to Latinos and Asians that the GOP was a nativist party (the beginning of this was the grassroots conservative opposition to George W. Bush's immigration reform plan).

If, at this moment of triumph, liberals themselves were to turn anti-immigration, it would reverse the enormous gains and squander the historic opportunity that the GOP's nativist blunders have offered us. Think about whether the GOP or the Democrats would be better for America's working class. It's a no-brainer.

Reason 3: Offshoring

In today's globalized world, capital is highly mobile; multinational corporations can put their factories and offices wherever they want. If we did manage to push up wages here in America by restricting immigration, companies can partially offset this by relocating overseas. The offices and factories will go where the labor is. That is probably exactly what is happening with Japan, where population is declining and immigration is heavily restricted, but real wages have been flat or falling for many years.

In other words, offshoring cancels out much of the effect of immigration restrictions. Of course, we could try to block that effect by restricting offshoring and trying to trap capital here in the U.S. Labor protectionism would thus require actual, general protectionism in order to be effective. But actual, general protectionism would be bad for our economy, including our working class. We depend a lot on trade.

Reason 4: Social Security

The most enduringly popular liberal social program in the U.S. has been social insurance, a.k.a. Social Security. But as everyone knows, Social Security is put under extreme fiscal strain if the population starts shrinking. This would give political ammo to Republicans who want to destroy this cornerstone of the New Deal. Thus, immigrants - who are relatively young, and who almost all work - are key to preserving our system of social insurance in an age of low fertility among the native-born.

The Alternative: Focus on High-Skilled Immigration

There is a clear alternative to labor protectionism. Instead of focusing on decreasing the supply of working-class immigrants, let's focus on increasing the supply of high-skilled immigrants! Dean Baker, in fact, hits upon exactly this solution when he writes:

[A] large flow of very highly educated immigrants, such as doctors, can get the wages of these workers more in line with the wages of professionals in other wealthy countries and provide large savings in areas like health care.

This is absolutely right, but it receives only a small mention at the end of an otherwise distinctly anti-immigration post. Instead, support for more high-skilled immigration should be the centerpiece of liberals' approach to immigration.

Remember, even in Econ 101, what matters for wages is relative scarcity, not absolute scarcity. What that means is, for a given number of working-class population, every additional high-skilled immigrant raises working-class wages. So if we increase the absolute numbers of high-skilled immigrants, we will automatically raise working-class wages.

And if we shift our immigrant mix toward high-skilled immigrants, by allocating more visas based on skills and fewer based on family reunification, the effect on working-class wages will be even more dramatic. This is the upshot of the Atlantic article that Adam Ozimek and I wrote on high-skilled immigration in June 2012.

Brad DeLong once wrote:

I think the United States needs more immigrants--more people willing to take risks and work hard to seek a better life for themselves and their children, and illiterates from Chiapas seem to me as good as doctors from Calcutta.

I also like both varieties of immigrants. But if you care about the income distribution, the two are not equivalent. High-skilled immigrants are better for American working-class wages than low-skilled immigrants.

In conclusion: If liberals go the labor-protectionist route, the potential gains (in terms of working-class wages) are miniscule, or even negative, while the potential losses (in terms of political advantage for the Democrats) are absolutely enormous. But if people like Dean Banker drop the labor-protectionist angle and throw their political weight strongly behind a shift toward high-skilled immigration, the potential gains are very large and the potential losses miniscule.

The choice is clear.


Update: Dean Baker responds. I can't say I'm happy with what he writes ("As far as less-skilled immigration, I would want it sharply limited"), but hopefully the logic of my post will sink in over time...