« Debt, Default and Leadership | Main | Opposition to two FTAs: Canada/US 1987 vs Canada/EU 2013 »


Feed You can follow this conversation by subscribing to the comment feed for this post.

If the supply of imported goods slopes up, then the tariff revenue paid by foreigners commonly exceeds the deadweight loss of domestic consumer surplus. Then the tariff is a very, very good thing for the imposing country. (Even though the world as a whole ends up poorer.)

Mike: Yep. That's the "optimal tariff" argument. It's a way of getting foreigners to pay Canadian taxes (or exploiting our collective monopsony power). But it's funny you rarely hear the same argument being used to justify export tariffs (where we could exploit our collective monopoly power). The optimal tariff argument makes more sense in the US than in Canada, simply because we're a smaller player in the world economy, so the "small open economy" assumption is more reasonable. (Though maybe not for some of our exports).

Nick - I think your student has a good point, though, or a point anyways. Suppose the only cost difference between Canadian and world producers is labour costs. You can think of two states of the world:
a. the government imposes a tax, and uses it to finance a Working Income Tax Benefit which, by subsidizing Canadian wages, reduces Canadian costs to world costs
b. the government imposes a tariff equal to the difference between Canadian labour costs and world labour costs.

Your response might be that I've fallen for the lump of labour fallacy. You might accuse me of believing that there are only so many jobs in the economy, and a firm's marginal cost curve in a world of minimum wage and other legislation is not a true reflection of the opportunity cost of its labour and other inputs. You might say that, in truth, if we weren't protecting industries, the labour in those industries would be productively employed elsewhere. You might ask me to count the black deadweight loss triangles.

To which I would respond: "I've got some nice Havana Club in the cupboard, would you like some?"

Awesome trapezoids, b.t.w.

This is all correct as far as it goes, but there are problems if you use this as a model of the actual economy.

The point of the tariff is the increase in Canadian production. This obviously has macro effects that this model ignores. The immediate question is whether the gains from increased Canadian production (employment, supplier revenue, other possible multipliers...) offset the losses of Canadian taxes.

This model also would have problems if Canada were a developing country. Tariffs have been used as part of a successful industrial policy (as in Korea). Of course, more often than not industrial policies cause more problems than they solve.

The big troubles with tariffs would seem to me to be retaliation and misallocation. The tariff escalations of the early 1930s are the classic example of the effects of the former, and Japanese agricultural tariffs of the latter.

It would also be useful to consider the common case where the tariff is so large that imports are 0. The US sugar tariff is a good example here. You get a transfer from the public to the sugar producers and the misallocations of corn ethanol and high fructose corn syrup.

I think you could fairly easily extend the model to show some of this. For instance, the trading partner could have a corresponding export diagram in the same tradeable. There could be an additional pair of retaliatory diagrams, and there is loss of the gains from comparative advantage.

It's more complicated, but it models a much larger share of the real world losses. It may be too complicated, but I have a strong prejudice toward models with clearly visible effects, rather than just theoretical ones, even those which are in their own terms provably correct.

trading partner could have a corresponding export diagram in the same tradeable. There could be an additional pair of retaliatory diagrams,

electrons will go in one direction, but not for long. Coulombs like circuits. Same difference for foreign trade. When you put import tariff, you simultaneously slow exports. Does factory owner like trade surplus towards greater profit for hoarding his gold? More of slightly protective import tariff? Do his workers prefer trade deficit? More imports of cheap tennis shoes and gourmet foods from Sicily? Doesn't matter. Either way tariff slows the circuit. With less circuit you get less division of labour thus less efficiency. As economy becomes more complex with time, division of labour has that much more potential to generate mega-progression of paradigms and quantum jumps to higher orbit.

When you look at Baltic Dry Index you will be sad. If you sell iron ore for steel ship construction or produce float-fuel at your refinery, you will be double sad. We got to get this World back into shape, got to snag few of politicians and give them whu'fo'.

Sorry, "could have a corresponding export diagram in the same tradeable" is an editing/proof reading error. That, of course, should be "in a DIFFERENT tradeable".

For instance Wikipedia says about the Smoot-Hawley

"In May 1930, the greatest trading partner, Canada, retaliated by imposing new tariffs on 16 products that accounted altogether for around 30% of U.S. exports to Canada.[15] Canada later also forged closer economic links with the British Empire via the British Empire Economic Conference of 1932. France and Britain protested and developed new trade partners. Germany developed a system of autarky."


"Using panel data estimates of export and import equations for 17 countries, Jakob B. Madsen (2002) estimated the effects of increasing tariff and non-tariff trade barriers on worldwide trade during the period 1929-1932. He concluded that real international trade contracted somewhere around 33% overall. His estimates of the impact of various factors included about 14% because of declining GNP in each country, 8% because of increases in tariff rates, 5% because of deflation-induced tariff increases, and 6% because of the imposition of non-tariff barriers."

It may be true that "With less circuit you get less division of labour thus less efficiency", but it's important not to lose sight of the economic and political distortions. You get resource misallocation and a political constituency that benefits from it, and like cancer the problem seems to spread.

Here is an actual diagram for sugar

S(us) represent the US producers’ supply schedule. S(world) is the world supply (assumed to be perfectly elastic). D(us) is US demand for sugar.

We'll leave the cost of the quota + tariff to the consumer as an exercise to the reader.

From the Steve Keen website:

"I also recommend reading Nick Rowe's post "What Steve Keen is maybe try­ing to say“, which is the first con­certed (and very accu­rate) attempt to put my endoge­nous money approach in a form that Neo­clas­si­cally trained econ­o­mists can under­stand. This could be the start of a real dia­logue in economics."

Frances: having thought it over last night, I finally came up with the right response: Yep, our trade theorist colleague Rick Brecher has published a lot of papers saying that a tariff can be the second-best(?) policy response when there is some can't-be-gotten-rid-of thing (like minimum wage law or efficiency wage) that causes wages to be "too high". But IIRC his models always have some second factor apart from labour?

And then I would say "yes please!" to a Havana Club anejo.

Thinking about how to shade in the trapezoids was actually the hardest part of writing this post. Mankiw does it by labelling each point, and then referring to area ABCD, but you can't see that so well. I normally shade in each area in a different colour, with diagonal/vertical lines, but that gets messed up too with the areas overlapping. I now think I'm going to use this method when teaching, because I think you can see it more clearly.

Peter N: Yep, this is a partial equilibrium model. Implicitly there's a second good, apples, that Canada exports. With just two goods we could draw a PPF and indifference curve, and a budget line with a slope equal to world prices, a do a general equilibrium analysis showing free trade is best. But then we need to introduce a third good, for government spending, which means a 3-D diagram, which is hard. But you get the intuition from the partial equilibrium diagram. Free trade is better than autarky because it results in more banana consumption (less apple consumption) and less banana production (more apple production). A tax on bananas distorts consumption but not production. A tariff on bananas distorts both consumption and production.

Yep, it felt good to get that response from Steve Keen. Mostly I totally disagree with him. But here I think he is onto something, and it's something similar to what I've been onto.

So that's what Rick is always going on about....

How would the model look like if we assume the government spends all revenue from the tariff as "banana vouchers" which will allow consumers to pay the higher prices imposed by the tariff? Sounds nonsensical, I know, but would it not be a possible way to rephrase the question of the student that may be helpful here?


The sugar graphic I posted is a quota, not a tariff, so the calculations are different. You can see the effect is a huge transfer from consumers to producers. For sugar, alone it's billions of dollars. It might be interesting for students to compare the effects of tariffs and quotas.

The externalities are also ugly.

How would you answer a student asking you about tariffs that countervail foreign subsidies? We correct for misallocation but The Economist would fault you as they think that "All tariffs are bad for consumers" and that's the end it. Which is true if you are a 19th century rentier.

Odie: take the second diagram, make the tax negative, so Pw+t is less than Pw, then impose a tariff as in the first diagram, and you get the answer. Too much consumption of bananas, and too much domestic production as well. 2 welfare loss triangles.

Peter: I usually do compare tariffs and quotas (but ran out of time this year). Yep, the quota is like a tariff, where the revenue goes to those who get the quotas.

Jacques Rene: Does it matter to Canada why Pw is what it is? If the Rest Of the World subsidises bananas, that will make Pw lower. But an import tariff would only help the ROW escape the effect of its silly policy, while causing a loss in Canadian total surplus. If the ROW gave Canada free bananas, should we say "sorry, but we can't accept that generous gift, because it is costing you too much"?

Nick: then why are we arguing about inefficiencies if we don't mind them?

Jacques: in this mode our social welfare function only incorporates Canadians. The rest of the world is just a black box, a mysterious device that transforms exports into imports.

Jacques Rene: I would have given the same answer as Alex. The arguments for tariffs are normally made from a nationalistic perspective, so counter-arguments against tariffs are normally made from that same perspective.

But suppose we don't take a nationalistic perspective. Suppose we look at global welfare. An import tariff would be the first best corrective to an export subsidy by the ROW. The net effect is just like a lump sum transfer from ROW governments to the Canadian government. It wouldn't be a first best corrective to a ROW subsidy on production of bananas, because ROW consumption would still be too high, but I think you could argue it works OK as a second best.

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad