Scott Sumner  

Growth, or jobs for Americans?

Alan Blinder's Bad Article... Scott Alexander Calls Out the ...

The New York Times has a good article on the trade-off between growth and jobs (a theme I have repeatedly discussed):

Donald J. Trump wants us to dream bigger about the economic future. We shouldn't be content with the roughly 2 percent annual growth that has been the norm this century, he has said. And he thinks he can bring about the kind of robust growth of 4 percent or more that was commonplace in decades past.

But the closer you look at the math of economic growth, the more you see the inherent contradictions in trying to make that happen. The two strategies that would most directly help achieve that goal clash with other planks of Mr. Trump's economic agenda.

Economic growth can happen two ways: More hours are worked, or more economic output is generated from each hour of labor.

Higher productively often leads to job loss:

[S]ome of the very innovations that have helped improve workers' productivity are also the main culprit behind the decline of well-paying manufacturing jobs.

These advances have made the United States richer -- with jobs like designing software systems and taking medical images. But that's not much solace for former manufacturing workers who haven't found lucrative or rewarding work in growing fields, a group that is at the core of Mr. Trump's campaign appeals.

It's not just manufacturing. Consider one innovation that could plausibly become a reality in the years ahead: trucks that drive themselves. Over time, that will make the United States economy more productive and raise incomes.

But if you are one of the 1.7 million long-haul truck drivers in the United States, making an average of $42,500 for a job that doesn't require advanced education, it should be concerning. It's plausible to imagine a majority of those jobs going away over the next decade, which will be a boon for the countless industries that rely on trucks to bring in supplies and distribute finished goods. That will raise G.D.P., the broad measure of economic growth.

The other way to raise growth is immigration. Consider this story about the opening of the world's highest bridge:

The Beipanjiang Bridge took three years to build, cost more than $146.7 million dollars, and stretches to be 4,396 feet long. Its four-lane roadway sits nearly two thousand feet over the Beipan River, which it's named after, to connect the Yunnan and Guizhou provinces in southwest China, People's Daily Online reported. It surpassed the 1,837-foot-tall Sidu River Bridge for its world record title.

We know what you're thinking: $146.7 million is cheap! And it is! After all, when the Golden Gate Bridge spanning San Francisco Bay was built in 1933 it cost $1.5 billion (in adjusted dollars), and the brand new New Bay Bridge, on the other side of town, ended up costing a whopping $6.7 billion when it was finished in 2013.

In fairness, the two Bay Area bridges are both longer and wider. Nonetheless, the Chinese bridge would have certainly cost far more than $147 million if built in the US.

The best way to rebuild our infrastructure would be to bring in Chinese construction firms and Chinese temp workers, who would live in prefab dorms while providing America with first-rate infrastructure. But we won't do that, because American workers would complain. Hence we will not end up with first-rate infrastructure. Despite all the talk, we will fail to rebuild our infrastructure.

Unlike Canada, this country has not yet accepted that the things that would provide faster economic growth (trade, immigration and automation) are exactly the same things that threaten to take jobs away from domestic workers. This is why I believe that growth will remain low over the next 10 years, even if there are occasional spurts of above 2% growth. We say we want faster growth, but we simply do not want to do the things that would be necessary to achieve that goal.

It's even worse in southern Europe. They say they'd like to have the sort of healthy economy that Germany has, but their governments are not willing to adopt German style economic reforms. Indeed they don't even like it when the Germans tell them how to do it.

You'd think that people would appreciate good advice.

COMMENTS (19 to date)
bill writes:

The cost of infrastructure in the US is exhorbitant. It doesn't have to be this way.

K Banaian writes:

I'm trying to understand this point about the cost of the Chinese firm coming in to build a bridge in the U.S. The cost of the labor in China is a function of the opportunity cost of workers in China. When they come to the U.S. to build a bridge, doesn't their opportunity cost rise? Or do we assume the only thing they can do here is work on that bridge, or else we send them home? I don't think you mean to assume that. (And I know you didn't -- it's in the Yahoo story.)

This is not to mention the opportunity cost of the materials or the investment capital.

It's almost certainly true that opening up public construction contracts to international bidders would lower cost, so I don't think your basic point is lost with this observation. I would prefer your point be a little cleaner, even if it means we can't make it more precise.

Scott Sumner writes:

Banaian, In the example in this post I assumed they were guest workers brought here specifically for the infrastructure project. That sort of thing is common in Asia.

Tom West writes:

The post raises a good point about the trade-off between jobs and growth. But even taking the article at face value, I think a number of people are expressing concern at the social costs of higher growth.

Maybe most of the Trump voters were delusional and think they can have their cake and eat it too, but at least it sounds like many of them would be willing to trade growth for less societal change.

(At least that's how I interpret the Trump vote.)

Thomas Sewell writes:


As almost all of Trump's picks for agency heads appear to be anti-regulation/government control, what kind of drag on GDP growth would you say the current regulatory state has and how much do you think that might change under new leadership?

Personally, I think that has the potential to more than outweigh whatever poor trade/immigration, etc... ideas Trump has which he can get through Congress, but I haven't dug into it enough to do any sort of a reasonable job of quantifying it.

Still, I've thought the Obama Administration's focus on increasing the regulatory burden has been a big contributor to keeping GDP growth low post-recession, so thinking along those lines would seem to imply a major reversal in that may be enough to make a 1-2% GDP growth difference.

What do you think?

Thaomas writes:

There are a lot more infrastructure projects that pass an NPV test today than there is willingness by governments to undertake them. True there would be even more if we used the MC of Chinese labor and project management, but I think that failure to use BC analysis is the bigger constraint.

Thomas Sewell writes:

Hmm.... taking a few minutes to do a little digging, it appears a good estimate is about -0.8% annual GDP growth change from impacts of the regulatory state over the last 30 years or so. Which is a pretty significant difference and well worth doing, but less likely to be enough to make a 1-2% difference over the long term.

ColoComment writes:

"-0.8% annual GDP growth change from impacts of the regulatory state over the last 30 years or so."

But if most of that increase in regulatory burden happened within, say, the last 10-15 years, and it was primarily those newer regulations (EPA's newer regs. on coal industry; Keystone XL Pipeline authorized; EEOC diversity reporting; SarBox, etc.) that were eliminated or lightened, mightn't that change the result?

Scott Sumner writes:

Lots of Toms today!

Tom, You said:

"At least that's how I interpret the Trump vote."

That's possible, but my point was that Trump campaigned on the idea of making America great again by sharply increasing economic growth. So there is tension between the growth promises and his other promises.

In addition, countries that reject neoliberal policies do not end up doing any favors for their working classes. So I think the voters are mistaken if they assume that making our labor market more like Greece will help blue collar workers.

Thomas, It's really hard to say, given that we don't yet know what polices will be enacted. The markets seem to expect a slight pickup in growth, but certainly not a sustained increase of 1% to 2%. Perhaps a few tenths of one percent. And even that would require that Trump abandon his promise to bring back manufacturing jobs, etc.

Thoamas, I don't agree, I think the unwillingness to use foreign labor is a far bigger constraint on infrastructure spending, perhaps an order of magnitude bigger constraint.

Thomas, And keep in mind that 0.8% is the total estimate. Obviously Trump doesn't plan to unwind all of those regulations, far from it. On the other hand, tax reform could also help.

The recent Carrier move is effectively an INCREASE in regulation.

arqiduka writes:

True, employment and growth seem to be inversely correlated in our day and age and it will be interesting to see where will the new administration draw the line.

Perhaps they have an easy way out though: just crack down hard on (low-skilled) immigration and you could in theory get more employment/higher wages for domestic workers without sacrificing much growth potential (I know, there’s a ton of “buts” and “what if”s there, but in the main it should hold). Not a silver bullet but worth noting.

john hare writes:

Crack down on low skill foreigners and you eliminate the production they create. If you don't believe that, try running a company with low skill American workers. All the small construction companies in my local area need useful help, and can't get much of it locally. My four man (at the moment) company has been trying to hire for a couple of years now without success. Starting pay is $12.00 for no experience. In my area, (central Florida) that is considered high.

I can hire bodies, but bodies don't produce. They have to be willing to learn, think, and do the actual work to be useful. Hiring non-productive people steals wages from productive people as well as the company. I should use the term anti-productive as the skilled people have to stop producing to attempt to get them to produce.

arqiduka writes:

@ John Hare,

I cannot comment on the local situation but will venture a guess that, absent a largish pool of immigrant labor, wages will just have to increase to the point of attracting (decent) locals. Of course, by “locals” I mean Americans, not necessarily townsfolk.

You will surely counter that many companies just can’t afford to pay much more that the actual going rate. Maybe not all can, but those that survive will be able to produce nearly as much as before. In short, you can “cheat” at the employment-growth game by letting the losses fall to labor, as long as it is immigrant labor.

john hare writes:


I can and do pay a lot more for productive labor. Most of my competitors do as well. There is a severe shortage of American workers willing to do the hard jobs at any rate of pay. Company average was nearly $50k including helpers, and the people in my crew are originally from Mexico. You don't get $30.00 an hour quality people by tripling the wages of $10.00 an hour people.

Wages have increased, and it's not working.

Mike W writes:

Unlike Canada...

What has Canada done to reconcile that the things that would provide faster economic growth (trade, immigration and automation) are exactly the same things that threaten to take jobs away from domestic workers?

Thaomas writes:


But the regulations on emissions from coal, reduces current GDP but avoids future harm from CO2 emission. The optimal tax/subsidy from any given regulation is not always zero.

ColoComment writes:

@Thaomas Ok, I'll give it a shot to explain my thinking (and please feel free to correct me where I'm in error!)

"but avoids future harm from CO2 emission."

That begs the question (that is, you are making an unproven assumption.) You assume future harm with no countervailing benefit. You also assume that avoidance of that assumed harm outweighs the cost to obtain it, as well as that doing something is "better" (for whom? for what?) than simply preserving the status quo.

Pure speculation here:
It could very well be that future emissions of CO2 at the current levels would prove, on the whole, more advantageous than harmful (e.g., lead to an increase in planetary flora, an increase in arable lands of the earth, or some such.) How does anyone know what is the optimum atmospheric proportion of CO2?

Further, the U.S. has already reduced carbon emissions beyond what the EPA originally required. We have the cleanest coal-burning power plants available with today's technology. Even absent more stringent regulations, energy industry relative market prices are encouraging the replacement of coal with natural gas for energy generation: why not simply let the markets do their thing? Why is government manipulation of the energy market the preferred choice of actions?
And, finally, don't economic precepts suggest that the last increment of any regulation proves the most costly per unit of benefit? When is it time to stop and call it a success?

ColoComment writes:

"The best way to rebuild our infrastructure would be to bring in Chinese construction firms and Chinese temp workers, who would live in prefab dorms while providing America with first-rate infrastructure. But we won't do that, because American workers would complain."

Huh? Importation of Chinese firms and *temp* workers would reduce the cost of American infrastructure construction? GMAB.

Even before American *workers* would have the chance to complain, you'd have to get past, among many other things, the EIS & permitting processes; the licensing of the construction firm with all the attendant WC, UEI, and bonding payments made; the EEOC, OSHA and DOL regulations for temp labor; social diversity requirements; and at the end of it all, you'd still have to pay wages in accordance with Davis-Bacon.

And please don't try to build "infrastructure" anywhere near an Indian reservation, or you'll also be shut down by claims of destruction of spiritual ground and artifacts, and a Army Corps of Engineers that will knuckle under to said claims, notwithstanding that you've complied with all required procedures.

Insofar as quality of Chinese construction, please Google "Chinese faulty construction" & check out some of the links, like this one:

Bernie638 writes:


Chinese construction firms and Chinese temp workers would be cheaper, but wouldn't reduce the cost of rebuilding our infrastructure by enough to make a real difference. Consider how much any of these projects spend on compliance costs (environmental regulations, legal requirements, accounting, lobbyists, etc.) that don't change by having a foreign workers. If anything replacing an American lawyer or accountant with a Chinese one would increase costs (lower productivity due to unfamiliarity with our laws).

Please look at the California high speed rail project update reports SB1029. Some of this contract was won by a foreign firm (Spanish not Chinese), and had some numbers for the costs of hiring local workers (approximately 250 MM in 2014 out of a 67B project).

Alternately you could calculate the savings backwards. Most projects track person-hours worked. Pick a difference between American and Chinese pay and multiply by person-hours. Then compare to the total budget for the project.

Reducing regulatory costs is the quickest way to lower the cost of infrastructure, and there would be job losses, just not blue collar jobs.

pyroseed13 writes:

This is an interesting post, but I'm somewhat of a Cowenian pessimist on this topic. I think there are potential growth opportunities from tax (particularly on the corporate side) and regulatory reform, but I'm not really buying the argument that more trade and immigration would unleash this swell of "growth opportunities." Trade is already relatively free, and our past trade deals have had only modest effects on GDP. TPP would help because we would be trading with relatively high-skiled, high-wage countries, but I'm not sure what the magnitudes of those effects would be. As for immigration...well...sure, if we adopted a points system and only brought in people who we think are likely to contribute to economic growth, than sure. But I'm not really seeing the positive effects in just bringing in more low wage laborers, other than for the laborers themselves. As another commenter pointed out, the Davis-Bacon acts as well as red tape are the main obstacles to improved infrastructure. Why not focus our efforts on reforming those?


Return to top