When countries do have trade surpluses or deficits, they reflect shifts in savings/investment propensities, and are essentially unrelated to globalization

See Random thoughts on globalization by Scott Sumner.

"Here's Scott Alexander:

An article by Freddie deBoer in this month's Current Affairs proposes "Journalistic Self-Outsourcing". DeBoer notes that lots of journalists and intellectuals suggest that protectionism and other anti-globalization policies are immoral. For example, Zack Beauchamp of Vox calls Bernie Sanders' skepticism of free trade "screwing the global poor"; Brad deLong calls the same "a call to keep China a society of poor subsistence rice farmers as long as possible - keep them poor, barefoot, uneducated, and by no means allow them to work at any of the high-value manufacturing occupations we want to keep in the United States."

Here's how Fredrik deBoer begins his response to DeLong:

We have entered another phase of journalists, raised in affluence and currently enjoying at least middle class incomes -- who are thus, according to their own moral calculus, very economically privileged -- telling Americans devastated by the collapse of the uneducated labor market that their poverty, marginalization, and hopelessness is Actually Good, because people in Bangladesh can now move from absolutely abject poverty to slightly-less-abject poverty. Provided the sweatshop where they work doesn't collapse on them. And provided they are willing to endure a nightmare of nonexistent labor power, terrible health and safety standards, total impunity from their bosses, and for the women, an atmosphere of near-constant sexual threat and exploitation.

I'm not familiar with deBoer, so I have no idea whether he's being dishonest or if he's simply unaware of global realities. But he's got things exactly backwards. The gain in living standards in recent decades (due to neoliberalism, including globalization) is one of the best things that has ever happened. Yet look at the adjectives that deBoer uses:

Americans devastated by the collapse

and

people in Bangladesh can now move from absolutely abject poverty to slightly-less-abject poverty

The losses to certain segments of America, while large in an absolute sense, are utterly trivial compared to the gains in living standards seen in places like Asia. And yet deBoer makes it seem like it's the Asian gains that are trivial. Does he not know what it means to see per capita income rise from say $1000 to $2000, or $5000, or $10,000? Or is he just trying to score debating points? I don't know. I basically agree with Scott's take on all this, but I'd like to comment on this paragraph in Scott's long post:

Again, there's a contingent argument otherwise. If you're really pro-globalization, you might believe that it's impossible for the Chinese to take all our jobs, in the same way that the Luddite Fallacy says it's impossible for robots to take all our jobs. The more Chinese take manufacturing jobs, the more Americans will have lots of money which will encourage new service jobs that the Chinese can't easily take. If this is Beauchamp's argument, he could say that we should globalize all the jobs that can be globalized, including his if possible, and then he will just move to an unglobalizable job. Since his job hasn't been globalized yet, maybe he's already in an unglobalizable job, so people should just stop bothering him.

I don't think that's a useful way to frame things---it makes it seem like rich countries simply outsource tradable goods jobs to poor countries, which is not at all what's going on. When thinking about the impact of globalization on the US it's useful to view the US trade account as balanced. You might find this claim strange, as we actually run a large deficit. So let me explain in 4 steps: 

 1. I am fairly confident that for the developed world as a whole, trade is roughly balanced. The rise of manufacturing in developing countries does not cause significant trade deficits in the developed world. Some developed countries, especially English-speaking countries such as the US and Australia, run trade deficits, while other developed countries such as Germany, Japan and the Nordics run big trade surpluses. But overall you have a roughly balanced trade account for rich countries. Thus even if you believed that trade deficits destroyed jobs in manufacturing (I don't), the argument would be very hard to apply to the developed world as a whole, because we export about as much as we import. So when we import more, we also export more. If there are fewer jobs in manufacturing in the developed world (and there are) it's due to automation, not trade.

2. I hope you are with me so far, as the preceding argument isn't particularly controversial. The next step may be controversial. When countries do have trade surpluses or deficits, they reflect shifts in savings/investment propensities, and are essentially unrelated to globalization. Because the US invests more than it saves, it would run a big trade deficit even if China were not a big factor. In fact, before China became a big factor we did have big deficits with places like Taiwan, Japan, Germany, etc. Germany has labor costs comparable to the US, but doesn't see their manufacturing jobs being taken away by China. They run trade surpluses because they save more than they invest. 

3. I was unable to get good data on manufacturing, but for what it's worth one source claimed the US exports $1.4 trillion in manufacturing goods each year, and various sources had total manufacturing output in the US as just above $2 trillion (I am not certain this data is correct.) The point is that workers in the export part of the manufacturing sector are actually helped by trade liberalization. Given a trade deficit of say 3% of GDP, any big boost in imports will lead to an equally big boost in exports, and in jobs manufacturing those exports. Less that 9% of American workers are in manufacturing, and most of those seem to be in sectors helped by trade. 

4. To summarize, trade deficits are not a problem, and don't cause a higher unemployment rate. But even if I am wrong about deficits, globalization is not causing our deficits, as indicated by the fact that the entire developed world has been heavily impacted by globalization, but has roughly balanced trade. It's (almost) all about automation.

Some commenters claim that economists are hypocrites because they don't advocate outsourcing their own jobs to Asia. DeBoer makes a similar accusation against DeLong. I'm not sure that's true, as I see increasing numbers of Asian job candidates at the AEA meetings, looking for academic jobs, and I approve of that trend, as do most economists that I know. On the other hand, I suppose economists might be opposed to an otherwise desirable policy change that immediately took away 100% of our jobs, so perhaps in that sense the accusation is true. But beside the point. I am pretty sure that DeLong doesn't complain when factory workers fail to advocate the outsourcing of their jobs.

Our models predict that trade will hurt people in some industries, and that the losers from trade will lobby Congress to prevent open markets. If I were an American textile worker I'd be pissed off right now. And yet cynics who contrast the views of economists and unemployed textile workers are missing the point. The really interesting divergence is between economists and those non-economists who have the same economic interest as economists.

Doctors, lawyers, plumbers, and electricians, are all helped by free trade, as their jobs are not threatened by imports, and they gain as consumers. Just like economists. However their views on trade are very different from the views of economists, in two ways:

1. Most people believe that if you like a domestically built car almost as much as an import, the patriotic thing to do is to buy the American car. Economists believe that if the imported car makes you a tiny bit happier, the patriotic thing to do is to buy the BMW. (Or at least good economists, who probably account for at least 37% of the profession.)

2. Most people view trade and automation as being radically different issues. Technological progress is viewed as good; trade is viewed with some suspicion. Economists see the two issues as being quite similar; creative destruction that improves overall welfare at a cost of painful dislocation for some workers and companies. Just imagine if Trump started railing against technological progress---even his supporters would be perplexed. Belief in "progress" is almost a religion in America. And yet it's automation, not trade, which is destroying millions of manufacturing jobs all over the developed world. Again, for the developed world as a whole, trade is roughly balanced. It's automation that destroyed factory jobs in the developed world.

The bottom line is that it's too easy to dismiss the views of economists by pointing to their comfy position, unthreatened by Chinese imports. For better or worse, we really do have a radically different view of how the world works. I can respect someone who goes full bore Luddite. I won't agree with them, but I can respect them. I cannot respect anyone who becomes an anti-globalization activist without also railing against technological progress. Instead, I assume they don't understand the issue---they haven't yet absorbed the message in Paul Krugman's Pop Internationalism. They aren't rejecting comparative advantage and trade; rather they don't yet understand the concept.

PS. I am not certain about my trade account claim. I found some recent current account data that seemed to show a net surplus for the developed world, defined as Western Europe, plus the US, Canada, Australia and the rich parts of East Asia. I'm pretty confident that if the trade balance is smaller than the CA balance, the difference is not dramatic enough to affect my claim that trade is roughly balanced for the developed world. In any case, the Swiss will tell you that service exports are nothing to sneeze at.

Update: Tyler Cowen has a new post reminding us that reduced employment in manufacturing is about automation, not trade. However contrary to the title of his post, China is not really deindustrializing (prematurely or not), it is becoming more efficient at a rapid rate--the mark of a successful economy. Manufacturing output in China, already the world's largest, continues to rise rapidly."