This is not going to get Mexico to pay for the wall

 Well, now we know Trump’s big leverage over Mexico: force people to show ID in order to make money transfers.

Really, that’s it. That’s the big idea. So let’s briefly unpack it.

First, it would take some time. The Trump administration would need to go through the regulatory process. So this “Day 3” bullshit it just that. 

Second, it would then immediately hit the courts under multiple grounds, not least NAFTA. So more delays.

Third, let’s say it happens. Suddenly, undocumented immigrants can’t make money transfers via Western Union. Well, they would immediately move towards other means. And illegal transfer organizations would spring up. Bitcoin? Well, yes, but as this World Bank report from 2005 discusses, also “ethnic stores, travel agencies, moneychangers, hawala-type systems, courier services, hand-delivery, etc.” Some of which would, in 2017, use Bitcoin to actually send the money. 

This would all raise costs for the migrants sending money home, but it would not stop the flow. So the Trump administration would have to go after them. That can be done, but more delays. (Here Carlos will mention that the administration will be facing massive bureaucratic pushback, slowing things further.) 

At this point the Mexican government has been chuckling at Washington’s blundering for about two year. But finally, somehow, the screws go in. Then what?

Well, to figure out what the Mexican government will do, you need to look at Mexican politics. Half of all remittances go to eight states. Two of them are the D.F. and Mexico State. Those are rich places; no elected officials are going to come under pressure if remittances dry up. Of the top-ten states, few households receive remittances: 10% in Michoacán and Guanajuato; 7% in Guerrero. These households would take roughly a 15% hit to their incomes, with some knock-on effects for local economies. (Page 115.) One good counterfactual might be the Great Recession, which put a big brake on remittances from the United States.  This Banxico paper shows the results. They were unpleasant.

But would they be so unpleasant as to get the Mexican government to pay to avoid them? Uh ... no. First, it is not like Mexican governments are bending over backwards to address rural poverty now. Second, even if all the affected households decided to vote against the incumbent government, the swing would be less than 5%. Third, the swing by people not affected will be more, since, you know, they would be the ones paying the taxes to build the border wall they don’t want.

Note that the above completely avoids any Mexican nationalism. Which is, in fact, a thing and means that no government would pay. So we are back to cancelling NAFTA and ... yes, this is in the proposal ... raising visa fees. Because “even a small increase in visa fees would pay for the wall.”

OK. The U.S. issued 1.5 million non-immigrant visas to Mexicans (including border crossing cards) and 82,476 immigrant visas. Let’s say the U.S. issued $17 billion in 30-year bonds to pay for the wall. You could finance the resulting $804 million annual expense by raising visa fees a mere $502 per visa. A small increase, indeed. I mean, it is possible that demand might go down a tad if the modal visa fee went from $160 to $662.

I am wasting my time, obviously.

Comments

[shah8]

Hey now, I profit, I profit! This is as good as a standup routine...

[J.H.]

What if the visa fees were doubled instead of quadrupled with the expectation that the increase in visa fees would pay off the debt over 60 years instead of 30 years?

[Noel Maurer]

Who knows? There are no bonds dated that long. And we have no idea what the demand elasticity is for U.S. visas.

My proposal is tongue-in-cheek. I'm not sure that the U.S. has ever issued hypothecated bonds. The Panama Canal bonds were actually general-obligation, despite the name. A Trump administration could "pay" for the wall however it wanted.

The CBO uses a 10-year cash flow window, which is also silly. But hey. Let's assume that El Presidente wants to build the wall in eight years. He needs $17 billion. So $2.1 billion per year, under paygo rules. $2.1 billion divided by 1.6 million visas: a mere $1,300 extra per visa!

We're dancing on pinheads, here. Hey! That's a good insult for El President Aspirante. "Pinhead Trump"! Or maybe Tiny Dancer.

[Noel Maurer]

Anyway, for anyone interested in the current state of the border barrier:

https://www.fas.org/sgp/crs/homesec/R43975.pdf

http://www.gao.gov/new.items/d09244r.pdf

https://www.washingtonpost.com/politics/the-6526-miles-of-fences-in-the-southwest-border/2015/07/17/e4e4e674-2c40-11e5-a5ea-cf74396e59ec_graphic.html

Short version: the areas where Trump wants to build a wall are already covered with fences. (Trump has said that he does not want a wall where there are natural barriers.) The 2006 Secure Fences Act mandated double-fencing, but Congress blanched at the cost and repealed that requirement in 2008. End.