The Trump Administration recently announced that it would assess a one-time $100,000 fee on each new worker in the H1B visa program (not for each annual renewal, as some originally thought). This marks an enormous increase over the prior arrangement, in which the fee was typically less than $5,000. Naturally, many in Trump’s base enthusiastically cheered the move, saying, “This is what I voted for.”

 

In the present post, I want to disentangle the various reasons one might support such a measure, in order to clarify the debate and facilitate considerations of the various factors at play. To be clear, I am not trying to dissuade the MAGA base from applauding the move. Rather, my modest aim is to help sharpen everyone’s rationale for supporting it.

 

“Saving US Jobs”: Foreign Workers vs. Automation 

The main justification I saw from the supporters of the move was that it would return the STEM jobs back to American workers, rather than temporary immigrants from (say) India who are willing to do a given technical job for far less than an American with comparable skills. For many, merely pointing out this fact seemed sufficient to demonstrate that Trump’s change was a good idea.

 

It’s this type of analysis that I want to run through a wind tunnel, as it were. To that end, suppose we change the context away from foreign immigration and make it about domestic automation. For example, suppose there’s a US trucking company that currently employs ten full-time American drivers. Now further suppose they change their fleet over to self-driving trucks, thus putting those ten Americans out of work. Would it be a good policy to assess this trucking company a $1 million fee for doing this? After all, they are effectively taking a job that used to be filled by an American worker, and filling it with a “foreign” worker (who’s from the country of Computer-stan).

 

When I posed this hypothetical scenario on social media, some of the MAGA fans did agree that such a fee would be beneficial. To those folks, I have no further argument. I think they’re mistaken, but at least they’re being consistent.

 

However, what I found more interesting was that many of the fans of Trump’s H1B policy would not want such a fee assessed on US companies that “eliminated jobs” through automation. They told me that with automation, the average productivity of US labor is increased, raising real wages for everybody. (I’m translating their plain English words into econo-speak, but that’s what they were arguing.) In contrast, they claimed, it lowers US wages when foreign labor is allowed into the country, and that’s why the two scenarios are totally different.

 

Let me take the rest of this post to explain why I think this latter strategy fails. In other words, I don’t think you can plausibly argue that purely on economic grounds, allowing foreign workers makes US workers poorer, while allowing automation makes US workers richer. To be sure, there are other, non-economic, reasons we might want to restrict the flow of foreign workers into the United States, but here I’m trying to disentangle those other considerations from the purely economic analysis.

 

The Impact of Self-Driving Trucks on US Workers

Suppose the average salary of a truck driver is $100,000. Then with the rise of ever-more dependent self-driving vehicles, suppose that a fleet owner could effectively “hire” a computer to drive a truck for the equivalent of (say) $20,000 per year. (In other words, there would be a higher purchase price for a self-driving truck versus a conventional one, and so there would be an implicit price for the self-driving capabilities that, we suppose, works out to $20,000 per year.)

 

Now if US drivers had no better alternative, they would simply have to accept a wage cut and receive $20,000 per year rather than $100,000. However, in practice most of these workers have alternative options, and so they would leave trucking and go into them. For example, maybe the typical worker would leave trucking and $100,000 per year, and go into construction, initially making $60,000 per year. In the short term, the displaced truckers will definitely be hurt by the innovation, because after all they could have gone into construction before. The reason they were driving a truck is that that had been their best option, on the eve of the innovation.

 

Now initially, the drop in expense is largely captured by the owners of the trucking fleet. They can deliver their cargo to their customers on the same terms as before, but now they save $80,000 per year for every driver who left. Their gains are clearly higher than the losses of the truck drivers, who each only lost $40,000. (In fairness, we would also want to include the impact on the original construction workers, whose wages would also be depressed because of the influx of new competition from the displaced truckers.)

 

If we include the impact on all parties, in a scenario like this the total gains to the winners outweigh the total losses to the losers. That’s one sense in which “average output” goes up, and “the economy” is more productive because of the innovation.

 

But we can say more than that. Over time, so long as there is open competition, other trucking companies will likewise switch to self-driving vehicles. With the drop in costs, these companies will be able to afford cuts in the prices they charge for shipping cargo. Ultimately, margins will be restored in the trucking industry to their previous levels, such that the benefits of the innovation will have been passed through to the customers.

 

Thus, when the dust settles, the long-run impact of self-driving 18-wheelers will be cheaper products for all consumers. That means all workers will see higher “real wages” because a given paycheck will go further at the store. This particular innovation will still on net make the old truck drivers worse off, and possibly even the construction workers whose wages were depressed by the new competition, but all other workers will strictly benefit.

 

And now, finally, if we generalize this process for innovations across many industries, we see that in general, innovation makes everybody better off. For example, the displaced truck drivers will benefit from automation in agriculture, where self-driving tractors will allow for lower food prices. So while the truckers may wish to live in a world where the government didn’t allow the introduction of self-driving 18-wheelers, they would be hurt if the government banned all self-driving vehicles.

 

 

The Impact of H1B Foreigners on US Workers

The same type of analysis applies to US wages when we consider the introduction of foreign workers. If a US firm brings in workers from (say) India who are willing to do a particular job at half the salary of a comparable US worker, that definitely hurts those particular Americans in the short run. But it obviously helps the shareholders of the company right away, and with competition in the long run those savings flow to all US consumers.

 

To put the matter crisply, if you first run through a hypothetical scenario in which a US company “eliminates jobs” via automation, you can then do the same scenario in which the company “eliminates jobs” by hiring an H1B worker. If you reach a different conclusion in the two scenarios, I want to encourage you to specify exactly what drives the difference.

 

For example, let me address one obvious retort: In light of the shocking accident in which an illegal immigrant made a reckless U-turn and killed three people, a MAGA supporter might understandably argue, “It doesn’t help the average American to bring in cheaper truck drivers from India, because they don’t have the same standards as US drivers. Even if your apples are a cent cheaper, that’s swamped by the higher chance of getting killed on the highway.”

 

To reiterate, I understand why that would be a plausible reaction to the very specific hypothetical I sketched above. But is that going to work in general? After all, there will be scenarios where self-driving vehicles cause deaths that wouldn’t have occurred with a human driver. (Hopefully, these will pale in comparison to the safer driving due to lack of fatigue, intoxication, etc. on the part of the computer drivers.) Indeed, if we banned airplanes, then nobody would ever die in a plane crash. That doesn’t mean it would be a good policy.

 

On the particular tragedy of the U-turn, there are many different things that went wrong. After all, there are plenty of Americans who shouldn’t be driving an 18-wheeler, and the system should be such as to keep them from getting behind the wheel. If the argument is that an influx of foreign workers is incompatible with the existing regulatory / training procedures, fair enough. Again, my point isn’t to make the reader change his or her mind, but just to sharpen the argument.

 

Conclusion

I have subtitled the present post “Part 1 of 2” because I want to field popular objections to my arguments as laid out above. (I already have a few in mind, that I received on Twitter when I first posted my analogy.) Critics should feel free to send me their feedback at: Robert.murphy@infineo.ai and perhaps I’ll address your objection in the next post!

 

 

Dr. Robert P. Murphy is the Chief Economist at infineo, bridging together Whole Life insurance policies and digital blockchain-based issuance.

 

Twitter: @infineogroup, @BobMurphyEcon

 

Linkedin: infineo group, Robert Murphy

 

Youtube: infineo group

 

To learn more about infineo, please visit the infineo website