China’s tariffs on imports are more than double the United States’ (calculated on a weighted-average basis). Critics of open trade cite this disparity and similar ones with other nations to justify America’s current trade wars. America should have a level playing field, they say.
By this, they seem to mean that the U.S. government should apply the same tariffs to other countries’ imports that those countries apply to U.S. exports. But a level playing field can be very, very bad because the gain to domestic producers from raising tariffs is more than offset by the loss to domestic consumers.
To better understand this, we need to think about supply and demand curves. If you took Econ 101, you likely remember the graph showing how a transaction tax reduces both producer and consumer surpluses by discouraging some transactions—a deadweight loss. The tax also reduces harms exchanges that do take place because some of the consumer and producer surpluses are transferred to the government.
Analyzing a tariff is more complicated because the tax affects only imports. To understand it, first consider what happens when a country opens itself to competitive world trade. The blue curves in the figure below show the equilibrium when there are only domestic supply and demand, resulting in price p (domestic only) and quantity demanded Q (domestic only). However, when the market opens to world supply (the green curve), a new equilibrium results. The price shifts to p (free trade) and the resulting quantity demanded at that price shifts to Q (free trade-world). At the new, lower price, domestic production can only provide Qs (free trade-domestic) while imports provide the rest.
Domestic suppliers’ dislike of the opening of their home markets is understandable: it reduces both the quantity they sell and the price they receive. As a result, these producers’ surplus shrinks from region A + B to just A. But consumers gain immensely from trade; they receive the transfer of domestic producer surplus B along with new surpluses C and D.
The next figure shows what happens when the country introduces a tariff. The equilibrium price rises from p (free trade) to p (tariff), with the difference between them equaling the tax. Domestic producers increase the quantity they sell in their home market to Qs (tariff-domestic). Imports make up the balance of the quantity demanded, Qs (tariff-world_, but that amount is much less than it was under the free-trade equilibrium.
As a result of the tariff, consumer surplus falls significantly. Region A is transferred to domestic producers. Region C is transferred to the government in the form of tax revenue. Region D is a deadweight loss of would-be truck sales that are forgone because of the higher price. Region B is also a deadweight loss resulting from resources in the domestic economy shifting from higher-valued industries to production of the tariffed good.
For a real-world example of this, consider the U.S. “Chicken Tax,” a 25% tariff on light trucks. It was imposed as part of a 1960s trade war over European barriers to U.S.-produced chicken (hence the name). The chicken barriers ended long ago but the light truck tariff remains in place today (though most SUVs have been exempted from it since 1989) and it applies to countries like Japan and South Korea that weren’t even involved in the 1960s trade war.
If you’ve shopped for a pickup, you’ve seen the effect of the Chicken Tax: prices are high and competition is limited because few imported pickups are on the market. As a result of the tax, some would-be American consumers forgo buying a truck—the deadweight loss. Others do buy a truck but have reduced consumer surplus because of the higher price.
Tariffs often have punitive intent (especially in a trade war) and thus are set very high. That means the resulting price increase is large, as is the deadweight loss and the decrease in consumer surplus. Hence, consumers want low, and preferably no, tariffs. It’s far better for consumers—and perhaps exporting producers—if tariffs in trading countries are low and uneven rather than higher but even.
Back in 1992, the weighted-average Chinese tariff was 32% while the weighted average U.S. tariff was 4%. As of 2016, those tariffs had fallen to 3.5% and 1.7%, respectively. So yes, China’s tariffs are more than double the United States’. But that is far better than the countries having a level playing field of tariffs of 10%, or 20%, or 32%.
Thomas A. Firey is a Cato Institute senior fellow and managing editor of Cato’s journal Regulation.
READER COMMENTS
Warren Platts
Sep 30 2019 at 6:45pm
Your tariff diagram leaves out “box E”–the terms of trade gain–that happens when foreign suppliers lower their prices in response to the tariff. Typically, E > B+D, improving net welfare for the tariff-imposing country. VATs generate even better terms of trade gains, no doubt explaining the explosion of VATs even as tariffs have been reduced.
As for chickens, the EU still bans most American chicken imports because American poultry producers use chlorinated water to rinse chicken carcasses because that is the more sanitary procedure. The Europeans consider that to be a short cut that allows less sanitary procedures. Nonetheless, studies show there is no difference in the food safety between American or European poultry. Thus the EU is still up to its old mercantilist tricks; non-tariff regulatory barriers are the preferred weapon of choice these days.
And as for the success or failure of 25% tariffs on pickup trucks, it remains fact that U.S. pickups are the best in the world–expensive or not. I call that a success.
Jon Murphy
Sep 30 2019 at 10:53pm
The terms of trade gains do not exist in real life, nor really even in theory. See Krugman et all 2018, Humphrey 1987, and McCloskey 1980 for examples.
Furthermore, a “level playing field” argument is not the same as an optimal tariff. Indeed, an optimal tariff necessarily pre-supposes an unlevel playing field.
Jon Murphy
Sep 30 2019 at 10:58pm
Furthermore, the optimal tariff theory does not apply in the current trade situation for two main reasons:
First: China has retaliated with its own tariffs (something with the model explicitly does not allow)
Second: Trump has subsidized American farmers harmed by the aforementioned retaliations
Given #1 and #2 above, if the Trump Administration was following an optimal tariff model, we would expect to see tariffs fall. Instead, the exact opposite has happened: the tariffs have risen and risen.
Since models are about behavior rather than conditions, to best evaluate the model is to judge based upon behaviors (which is to say outcomes). Even if we grant some of the framework conditions necessary for the optimal tariff model to come into play (such as the existence of monopsony power by the imposing country), it does not logically follow that those necessary conditions are also sufficient conditions to justify the model. The outcome of this current trade war, even though the US does likely have some market power, is more akin to the standard trade model prediction that any tariff >0 will be a net welfare reducer (indeed, this is what the empirical studies all find). Thus, that model is more appropriate here.
Thaomas
Oct 1 2019 at 7:55am
I agree that it is almost impossible to believe that President Trump’s tariffs on Chinese goods are “optimal,” but as it has been a while since I taught trade theory, could you spell out the how optimal tariff theory works in the presence of (optimal?) retaliation and tax-subsidy compensation of the exporters harmed by the retaliation?
Jon Murphy
Oct 1 2019 at 8:15am
The optimal trade theory assumes no retaliation on the part of the imposed upon nation and no political distribution of the government tax revenue in the form of subsidies (there are other assumptions, such as the political system is costless to use, that are also damning to the optimal tariff Theory, but those are irrelevant to the point I am making).
Any net welfare surplus achieved through the optimal tariff (which I hasten to add for readers is a result of an accounting quirk and not an economic scientific reason) is small. If the foreign country retaliates or if there are consumer losses that are subsidized by the government, those net welfare gains disappear very quickly. In the case of retaliation, the optimal tariff just be recalculated and it would be downward revised, not upward, to account for the now-lower exporter surplus and government revenue. All of the papers and books I cite above focus on this problem in their rejection of the optimal tariff as both an explanation of observed outcomes or a viable policy tool.
Warren Platts
Oct 2 2019 at 12:45pm
That is not true. You are projecting onto the theory the assumption that the starting position must be one of complete free trade where one’s trading partners have committed to a policy of unilateral free trade–an unusual circumstance. The usual circumstance is a world of 200 nation states existing in an all-against-all trade war where tariffs are the normal state of affairs. In that case, each country will be better off pursuing an optimal tariff strategy.
In the case of the United States, it is Trump that is doing the retaliating. Given that the ROW–and China in particular–have been engaging in unfair trading practices for decades, the USA will be made better off under an optimal tariff strategy regardless of what other countries do.
Indeed, what optimal tariff theory demonstrates is the folly of a policy of unilateral free trade. Ironically, free trade cannot be free: it must be enforced because in a world of free trade, it is to the advantage of any country to cheat. And enforcement is a difficult thing to do in a world consisting of 200 sovereign nations existing in a state of virtual anarchy.
Thus it is no surprise that today’s global arena is anything but free. Tariffs are down worldwide, but countries have simply installed VATs to extract terms of trade gains from their trading partners, and non-tariff barriers to protect their workers. The few countries that have done neither–such as the United States for the most part–are either chumps or what McCloskey called “magnanimous Albions,” depending on one’s point of view.
Jon Murphy
Oct 2 2019 at 1:17pm
“The terms of trade argument [aka optimal tariffs] against free trade has some important limitations, however…For big countries like the United States, the problem is that the terms of trade argument amounts to an argument for using national monopoly power to extract gains at other countries’ expense. The United States could surely do this to some extent, but such a predatory policy would probably bring retaliation from other large countries.
…
The terms of trade argument against free trade, then, is intellectually impeccable, but of doubtful usefulness. In practice, it is more often emphasized by economists as a theoretical proposition than actually used by governments as justification for trade policy.”
-Krugman et. al 2018
“Except for Torrens and Bickerdike, these same economists [who theorized on the optimal tariff] also specified the basic shortcomings of optimum tariff theory. The theory, they noted, assumes unrealistically (1) that foreign countries will not retaliate with tariffs of their own, (2) that elasticities of supply and demand in foreign trade are not so large in the long run as to render the tariff ineffective, (3) that the optimum tariff rate can be precisely identified and skillfully administered, and (4) that politicians can resist pressures to raise tariff rates above the optimum level. None of these assumed conditions, they felt, were likely to be realized in practice. ”
-Humphreys 1987
“Finally, it hardly requires emphasis that throughout the preceding analysis we
have stuck rigidly to statics, and ignored completely such important dynamic
considerations as the probability of retaliation and the chance of protection in an important country precipitating a world slump.”
-Graff 1949
See also Lerner 1944, Kaldor 1940, Scitovszky 1941, Robinson 1946, Kahn 1947.
Alternatively, you could do the math yourself and see retaliation reduces the welfare gains.
Oh. Good to know that the Chinese increases in tariffs on soybeans and Trump holding off on tariff increases to save Christmas and the like aren’t actually occurring. Good to know that the Administration is just willy-nilly handing out subsidies to farmers for no reason.
Thaomas
Oct 3 2019 at 7:41pm
I agree that the standard way of deriving an “optimal” tariff assumes no retaliation. I meant how would it work with retaliation, (maybe some Cournot reaction function)? [My guess is that unless the retaliating country B can also impose a optimal tariff on item which the tariffing country A exports with less than infinite elasticity (and in which case why hadn’t B done so already?) that retaliation harms only country B, not Country A. But I’ll leave that to someone who does this for a living; I don’t anymore. :)]
I do not think the benefit from the “optimal” tariff is an accounting quirk. It results from the tariff pushing down the border price of the import, a terms of trade benefit to the importing country or in Trumpian words the exporting source “pays” some of the tariff.
I hasten to add that I in no way think that it is practical to make trade policy on the basis of “optimal” tariffs. To begin with it requires knowledge of the foreign country’s supply curve.
Jon Murphy
Oct 3 2019 at 8:31pm
Ah, my mistake! I misunderstood your question.
Long story short, both countries would end up harmed by retaliation. This was actually a central point of Gary Becker’s dissertation The Economics of Discrimination, where he modeled discrimination as an “optimal tariff” and shows how retaliation harms both parties.
Using a public choice model, retaliation also causes the optimal tariff to be recalculated, which in turn leads to more resources spent on that (the political process is not costless), and subsequent retaliation by the other guy, leading to more recalculation, etc.
The net benefit is. The only reason the optimal tariff is welfare-enhancing (not welfare reducing like every other tax) is because the costs to foreigners are not included into the calculation.
Don Boudreaux
Oct 1 2019 at 11:13am
Jon & Thaomas: There’s an elementary point that deserves here to be highlighted, which is this: The optimal tariff is a revenue measure, not a protective one. The optimal tariff has nothing whatsoever to do with ‘optimally’ protecting this industry or those workers from foreign competition. A successfully implemented optimal tariff results in foreigners paying to the home-country government an amount of tariff (i.e., tax) revenues greater than is the welfare loss suffered, as a result of the tariff, by home-country consumers.
Indeed, because the successful use of an optimal tariff increases the amounts that foreigners pay to citizens of the home country, it results in an increase in home-country imports relative to home-country exports. Put differently, the very purpose of an optimal tariff is to arrange for home-country citizens to receive a greater amount of imports in exchange for their exports – a purpose that could not be more different from the protectionist goals of Trump and his supporters (such as Mr. Platts).
Jacob Egner
Oct 1 2019 at 11:59am
Don: Wonderful comment. I’m really glad you take the time to share things like that with us. Thanks. –Jacob
Warren Platts
Oct 2 2019 at 12:01pm
I think it was Commissioner of Customs and Salt Duties Adam Smith himself who said that tariffs can raise a lot of revenue for the sovereign and provide a large measure of protection for the home country’s workmen. The two goals are not mutually exclusive. Most 19th century American protectionists would have agreed.
But what we are learning the 21st century is that when there are zero restrictions on capital inflows, for the USA, anyways, tariffs alone cannot end the trade deficit. In the case of China, what they have done is devalue the yuan to the point where it has completely offset the effects of Trump’s tariffs, on average. Hence the trade deficit has hardly budged, and, if the CPI is any guide, there has been no measurable cost to consumers on average. Yet tariff revenue has doubled.
That entails that practically all of the increased tariff revenue is because of the terms of trade gain: the tax incidence of the tariffs pretty much all falls on the foreign producers. That is good news for U.S. taxpayers, and to the extent that this throws sand into the gears of the Chinese economy, and considering that the CCP considers themselves to be in a state of “unrestricted warfare” (超限战, literally “warfare beyond bounds”) with the USA, that can only be a good thing.
But Dr. Boudreaux is correct to the extent that while Trump’s tariffs may be generating tax revenues that exceed any deadweight losses, the tariffs per se are doing little for the protectionist goal of ending the trade deficit. To secure the latter goal will require “tariffs” on foreign capital inflows as well as imported goods, such as the market access fee proposed by the incipient Baldwin-Hawley bill.
Jon Murphy
Oct 2 2019 at 1:18pm
This is incorrect. Cf Books IV and V of the Wealth of Nations. Couple this with his letters as Comissioner to Lord Auckland.
Don Boudreaux
Oct 3 2019 at 9:19am
Warren Platts:
I ask this question in all sincerity: Have you actually read more than a few select sentences or paragraphs written by Adam Smith?
Jon Murphy
Oct 2 2019 at 2:04pm
Joan Robinson’s metaphor about dumping rocks into one’s own harbor is as apt as ever.
Thaomas
Oct 3 2019 at 8:11pm
In no century will a change in tariffs change the trade deficit except as they change the saving-investment relation. To think otherwise is one (of many) of President Trump’s solecisms.
The failure of tariffs to change the trade balance does not, of course, mean that they cannot raise revenue or that they cannot raise incomes of firms producing substitutes for the tariffed goods, ie be “protective” of those goods, just that the same GE effects that prevent the trade balance from changing (except as they change the saving-investment relation) will “unprotect” non-tariffed tradable goods.
Ron
Oct 2 2019 at 10:56am
Pro-tariff people claim they are protecting American jobs. What do you have to say about that?
Jon Murphy
Oct 2 2019 at 1:19pm
One thing we see over and over again is protectionism does not protect jobs. See the current layoffs in the steel industry
Thaomas
Oct 3 2019 at 8:31pm
Tariffs can protect, indeed “create,” jobs in the directly protected industry. Raising total employment is something only the Fed can do and that only if there is unemployed labor. At full employment the additional people employed in the “protected” activities have to come from the “unprotected” activities. Jon’s example is of a loss of unemployment in the unprotected or disprotected steel using industry,
[The simplest way of seeing this is to imagine an economy that produces two goods, one that is mainly exported, one that is a close substitute for and imported good, but any GE model will do. Starting with full employment and a zero trade balance a tariff will increase output of the import substitute and decrease output of the exported good. Prices adjust so as to keep employment full and the trade balance at zero. The tariff has been “protective” of the import substitute and can have raised revenue without changing total employment or the trade balance.]
Thomas Firey
Oct 4 2019 at 1:24pm
Hi Ron; fair question.
Looking at my second figure, we see that tariffs HARM (some) domestic jobs. Remember region B is deadweight loss resulting from resources shifting from other industries to the protected industry as a result of the tariff. Part of that resource shift will be labor.
Are the “new” protection-created jobs “better” (e.g., better-paying, higher-number, more pleasant, targeting especially “deserving” workers) than the old, lost jobs? Maybe, maybe not. But for sure some domestic jobs will be lost because of the tariff.
Comments are closed.