Raymond Richman - Jesse Richman - Howard Richman
Richmans' Trade and Taxes Blog
Tariffs Raise Costs on Dryers (Without Dryers Being Taxed?!?)
A few weeks ago a working paper by Aaron Flaaen, Ali Hortaçsu, and Felix Tintelnot got a great deal of attention. The headline treatment in many publications was along the lines Mark Perry at AEI gave it: "Trump’s washing machine tariffs created 1,800 US jobs, but at a YUGE cost to consumers of $820,000/job." The Washington Examiner opined that this study showed "Why protectionism fails."
But there are some strange things going on in the analysis of the paper. The authors acknowledge one such flaw in the abstract when they write: "We find that in response to the 2018 tariffs on nearly all source countries, the price of washers rose by nearly 12 percent; the price of dryers—a complementary good not subject to tariffs—increased by an equivalent amount."
Come again? Let's lay this out for a moment and pause to think about it objectively.
1. Tariffs were placed on washers. Washer prices went up by about 12 percent.
2. Tariffs were NOT placed on dryers. Dryer prices went up by about 12 percent.
Now let's apply some basic logical reasoning. Mill's method of agreement implies that if a cause produced an effect, then the cause must be present in both cases. Clearly the effect happened in both cases but the tariff was applied in only one case. Hence, the tariff cannot have caused the increase in price of the goods. Perhaps something else in the market dynamics involving washers and dryers such as industry consolidation, price fixing, or an increase in the cost of some input(s) caused the price hike. But what we can rule out, Mill's methods imply, is the argument that the price increase was caused by the tariffs. Indeed, the authors come close to acknowledging as much on page 18: "The price effects on dryers are initially puzzling as these appliances were not subject to any new tariffs."
When one digs deeper into the paper, there's more trouble. The estimated change in prices for washers is partly the result of factoring in a declining trend in washing machine prices that was steeper than that for other major appliances. Perhaps the increase is simply the result of the end of such temporary factors such as perhaps the end of a price war in washing machines? The figure on page 17 and the table on page 18 also make immediately clear that the summary that dryer prices went up by "a similar amount" is somewhat misleading. Dryer prices went up slightly but not significantly MORE in most estimates. Since the author's putative theory is that the fact that dryers are a complementary good meant that higher prices for washers somehow pulled up dryer prices (more on that theory in a moment) the effect on dryers ought probably be weaker.
The authors argue that the connection between the prices of dryers and washers arises because these are "complementary goods." Yet if they were simply complementary goods, the argument for dryer prices increasing would be even more wobbly. Think about the basic supply and demand analysis. If the price of washers is driven up by tariffs, then this should diminish demand for washers and also for dryers. And this diminished demand should drive down prices for dryers. Thus, the simple fact that these are complementary goods doesn't cut it. But Section 4.4 goes into more detail about what the authors mean. And the authors do have some traction here. For whatever reason, most washer and dryer pairs are priced at the same level. The matching model dryer is listed at the same price as its paired washer. This oddity involving the prices for the two appliances -- the fact that they are often sold as equal-priced pairs -- does arguably lead to linked prices.
The final headline number is also problematic. Most of the putative tariff costs to consumers clearly cannot be accounted for by the actual tariff rates imposed. The total tariff revenue collected according to page 26 from the washing machine and parts tariffs amounted to 82.2 million dollars. 82.2 million divided by the roughly 1800 US jobs the paper says were created comes to less than $46 thousand per job, not the $820 thousand number headlined in the news reports. Somehow tariff costs are being passed on to consumers to such an extraordinarily extreme degree that tariffs lead to a situation in which consumers are paying about 18 times more than the tariffs cost? Something's rotten in this set of calculations. It makes no sense.
Journal of Economic Literature:
Atlantic Economic Journal: