I have no idea if the Trump administration plans to impose a 20% tariff on Mexico. It seems now that the surprise announcement had to do with plans for a border adjustment tax, which is a different thing. But who knows? This is the Trump administration! Why are you asking me to make predictions?
Anyway, yesterday I mentioned that any tariff would impose serious problems on the American auto industry. The big problem is twofold: first, Mexican suppliers provide a large chunk of the inputs into American auto production. Second, many of those parts in fact cross the border multiple times. The end result is that a tariff would raise the cost of producing domestic cars by quite a lot.
Some figures: in 2012, the state of Michigan published a report on the Ford Motor Company’s Michigan Assembly Plant. The report found that the plant purchased about $3.1 billion from Tier I suppliers; of that, $831 million (or 27%) came from Mexico. (Roughly $3,800 per vehicle.) In 2014, the U.S. auto industry consumed $42 billion in intermediate inputs from Mexico (pages 10-11), approximately $3,500 per vehicle. So at first glance, the tariff would raise the cost of American-built small cars by $760 and all vehicles by about $710.
That is a lower bound, however. In the course of making a motor vehicle, auto parts cross the border as much as eight times. If the tax was applied at each crossing, the cost would go up commensurately. If you assume that components cross from Mexico four times, doubling in value-add each time, then the total cost per vehicle would rise from $710 to $1,300.
The cost increase would hit small cars more than large ones ... but consumers would see the price hike in the large car sector. The reason is that auto companies in the United States need to sell a certain percentage of small cars in order to hit their fleet fuel economy standards. They will therefore hold down the price on small cars, transferring the impact to their larger vehicles.
The problem is that the U.S. producers more larger vehicles than small cars.
How much would production shift back to the United States? I suspect less than the current studies predict. Why? Well, the Mexican peso would likely fall. The real exchange rate would go down and stay down. Mexican workers, who consume a lot of imports, would become poorer but they would also become cheaper. A 20% nominal decline in the peso means a drop from 22 per dollar to 26 per dollar. That is certainly conceivable. Hell, 30 per dollar is conceivable. A messy drop would potentially hurt the Mexican economy, if it turns out that Mexican companies have a lot of hidden dollar liabilities, but would nonetheless increase competitiveness.
And so, you would have a temporary increase in the price of large automobiles and light trucks, followed by a drop in Mexican costs. Some production would return to the U.S. (in order to avoid uncertainty) but in unsteady and unpredictable way. It would not be an economic apocalypse for the United States, but it would cause havoc across the Mexican economy with little (if any) employment gain in the USA and big losses for shareholders. And unless the tariff is extended to all countries, the big winners would be car manufacturers from Europe and Asia.
That said, this is all theory. Who knows what the Administration will do?