The Devil Test

The issue of sanctions vs. tariffs has been raised before (here and here), and it's an issue worth pursuing: why are some restrictions on trade used as punishment, whereas other restrictions on trade portrayed as development? The reason I have trouble understanding those who see a clear distinction between selective tariffs (=good) and blanket sanctions (=bad) is that it doesn't seem to pass the Devil Test.

This test comes from Pete Boettke's "Is Economics a Moral Science"*, and refers to the use of value-neutrality in economic debate. Let's agree that angels want "good things" and devils want "bad things"; the angel wants economic development, the devil wants economic stagnation. Traditional political debate will focus on the worth of these two moral desires (i.e the intentions), but economics provides the value-neutral criteria to take those ends as given, and analyse the actual cause-and-effect. It provides constraints for their arguments, and potential for rational debate. It means we can argue about scientific right vs. wrong, not moral good vs. bad.

In this case restrictions to economic activity (i.e. tariffs and sanctions) are being advocated by both angels (people wanting economic development) and devils (people wanting economic collapse). There's an inconsistency here - someone must be wrong since angels and devils aren't allies.

So either the UN is really an angel (and secretly attempting to stimulate the North Korean economy and empower their citizens), or the European Commission is really a devil (and is fully aware that tariffs are harmful but are controlled by the vested interests who benefit from it). Which seems most likely?

The Devil Test suggests something's afoot, and notice the bias the underlines this whole issue. We're now looking at sanctions against North Korea, but the debate is already framed by central planners: the question is should we do X, or what's our response. Such an attitude is biased since it pre-supposes that deliberate intervention is the appropriate solution. It might well be, but it isn't necessarily so. It ignores the endogenous solutions that result from the people themselves, and even if we Westerners claim this as "our problem to solve" we should, at the very least, do so in partnership with the people:

"I have always felt that we should be guided by domestic oppositions in the application of sanctions. That's why sanctions were right against Poland in the 1980s, where Solidarity wanted them, and against South Africa, where the ANC wanted them, and are right against Burma today, where Aung Suu Kyi emphatically supports them."
Tim Garton-Ash "Histories of the Present" p.429

---

*Boettke, P.J. "Is Economics a Moral Science?" Markets & Morality 1(2) (October 1998): 212-19

Comments

[Jim]

This again. It seems to me you're not adhering to value-neutrality yourself, because you seem to be of the opinion that all trade restrictions = bad. Once you accept that they can sometimes be good - which is what experience and the evidence should tell you - as well as sometimes being bad, and that policymakers can make reasonable guesses as to which restrictions might be good and which might be bad, it should be clear that the restrictions a government selects for itself are probably more likely to be beneficial than those imposed from outside with the intention of harm. Also, think about how different sanctions - which are typically blanket embargoes on imports of something a country will have difficulty producing itself - usually are from voluntary restrictions - which are typically simple taxes upon imports of something the country *can* produce for itself - and the likely impact on consumptions patterns should again be different.

So in summary, I think the reason you're confused about this is because you need to get around your own priors and actually consider the evidence.

[AJE]

you seem to be of the opinion that all trade restrictions = bad

Well then you've misunderstood me. I'm not claiming that at all - I'm claiming that sanctions and tariffs are both trade restriction (and that whether I personally think they improve or reduce prosperity is irrelevent). In fact the quote at the end suggests the opposite of what you've accused me of.

the restrictions a government selects for itself are probably more likely to be beneficial than those imposed from outside with the intention of harm

Once again, you inject your own priors in by associating sanctions with "intention of harm". As I showed previously both tariffs and sanctions can be implemented with alternative intentions, therefore you're making a false distinction.

To keep this value-neutral it's not at all clear to me that internal tariffs will always and everywhere make a country better off than externally imposed trade restrictions. Indeed I believe that you yourself have advocated UK intervention in the trade policy of foreign countries precisely on the grounds that failed states can do themselves more harm than would occur under imposed conditions.

Are you really arguing that Western agencies should never intervene in the trade policy of foreign countries? Or are you just saying that they should only do so if they have the right intentions? (Road to hell etc...)

I do think my priors are irrelevent in this debate. I'm making an analytical assumption that tariffs and sanctions are analytically equivilent, and can therefore be analysed under the same framework (as "restrictions on trade").

The fact that you define them with regard to the intent to cause harm is precisely why we need a value-neutral debate: the intent of a policy os wholly irrelevent. All that matters is the effect.

So in summary, I think the reason you're confused about this is because you need to get around your own priors and actually consider the evidence.

As I've said, you've misunderstood what my priors are so they don't appear to be all that strong. Rather than insist on a value-laden debate based on rhetoric, and before it's worthwhile to look at any evidence, please explain why we can't discuss tariffs and sanctions as being analytically equivilent?

[Jim]

"please explain why we can't discuss tariffs and sanctions as being analytically equivilent?"

Simply because trade restrictions are not all the same. In fact, different trade restrictions can have very different real effects. Just because you choose to lump them all together in one analytical category does not mean they actually belong there. In my view, that's an analytical error, equivalent to saying that all taxes are the same. You can pursue this line all you like, but I think you're barking up the wrong tree.

[AJE]

You've misunderstood me because I'm not denying that "different trade restrictions can have very different real effects"

I just think it's nonsense to define those different effects as being whether it's a tariff or a sanction, as if they're two completely seperate categories.

[Jim]

"I'm not denying that "different trade restrictions can have very different real effects""

Good. And since you're able to make the distinction, and so am I, isn't it likely that policymakers are able to as well? And if they are able to distinguish different likely effects, isn't that going to make them choose different types of trade restrictions depending on the effects they want? And isn't that what we see in practice - that sanctions are typically different in scope and kind from voluntary selective trade protections?

To put it another way, once you accept that trade restrictions vary significantly in their effects, and these effects can be guessed at reasonably well, intentions become a much better guide to likely effect, but even leaving intentions aside sanctions and selective voluntary protection are usually clearly distinguishable anyway. It's not so much about observing the data through the prism of one's values as it is about the relevant actors involved being purposeful and not idiots.

[AJE]

And if they are able to distinguish different likely effects, isn't that going to make them choose different types of trade restrictions depending on the effects they want?

But unless we're assuming omniscience and benevolence these policy choices will have (a) unintended consequences and (b) hidden intentions.

even leaving intentions aside sanctions and selective voluntary protection are usually clearly distinguishable anyway

At what stage of this debate do you intend to actually justify this claim?

[Jim]

"But unless we're assuming omniscience and benevolence these policy choices will have (a) unintended consequences and (b) hidden intentions."

They may do, but since they are different policies, enacted by purposeful and observant actors, why should they have the *same* consequences?

"At what stage of this debate do you intend to actually justify this claim?"

At what stage do you intend to stop asking me to state the obvious? A typical sanction is a complete embargo on importing any quantity of a particular good, and it brings no revenue to the government of the 'importing' country. Now, is a typical trade tariff anything like this? No, it clearly is not - typically, it is a tax on the imports of a particular good, and it does bring revenue to the government. In neither design nor effect are they similar - and I really don't understand why this is so hard to grasp, apart from it fits handily with your general antipathy to trade restrictions.

[AJE]

why should they have the *same* consequences?

But there's a difference between observing something and controlling it. Accepting that policy has different effects doesn't prove that we can control those effects.

At what stage do you intend to stop asking me to state the obvious?

After you've done so!

In your example, what's the difference between a typical sanction and a tariff set so high that it prices foreign competitors out of the market?

Nothing.

And whilst we're at it, if you're claiming that the main difference between tariffs and sanctions is whether it raises government revenue, then according to your logic a quota is a form of sanction. And since quotas are as relevant as tariffs for this debate (i.e. internal voluntary restrictions vs. externally imposed ones), your whole argument falls apart.

Let me put this another way. A government believes the infant-industry argument, and therefore slaps a tariff on imports of that product.

According to your framework we can't begin to analyse the situation until we know the level of the tariff (i.e. is it set so high that imports are zero).

I'm claiming that we can make predictions about the likely effects without knowing the exact level of the tariff, since whether it's technically defined as a tariff or sanction they're both trade restrictions, and therefore both analytically equivilent.

[Jim]

"In your example, what's the difference between a typical sanction and a tariff set so high that it prices foreign competitors out of the market?"

In effect, not much - and the difference between such an extreme example tariff and the vastly more common type that doesn't completely price out imports is very large, which just goes to show that you need to look closely at each particular trade restriction rather than just lumping them all together as "analytically equivalent". Do you think that a 100% income tax is analytically equivalent to a 10% income tax?

"if you're claiming that the main difference between tariffs and sanctions is whether it raises government revenue"

I'm not claiming that. They are clearly different policy instruments and usually have clearly different effects aside from the impact on revenues.

"if you're claiming that the main difference between tariffs and sanctions is whether it raises government revenue, then according to your logic a quota is a form of sanction"

No, but a quota is actually an interesting example, because its real effect is entirely contingent on the particularities of the situation - i.e. whether the quota has been reached or not. Another reason why lumping trade restrictions together isn't particularly useful.

"According to your framework we can't begin to analyse the situation until we know the level of the tariff (i.e. is it set so high that imports are zero)."

Don't you think it's useful to know the level a tax is levied at? Isn't one that's set so high that income is zero a bit ... different?

"I'm claiming that we can make predictions"

No doubt you can predict away, I just don't think it'll be useful. For example, as you say above you don't know whether a tariff is going to have the same effect as a sanction - eliminating all imports - until you look at it. And whether imports are zero or not matters for a lot of things - if producers learn technologies from imported goods, for example.

It doesn't sound like we're going to agree on this. You think they're analytically equivalent, I don't. Next question please.