by Ian Fletcher
We do, however, have one obvious starting point for reasoning out a tariff policy. If free trade is wrong because of the list of its flaws we have compiled, might the right policy consist in systematically fixing these flaws? There is, in fact, an entire school of thought that aims to “restore the lost innocence” of free trade in this way.759 Because many (not all) of these flaws consist in failures of the free market to work properly, the great attraction of this approach is that it satisfies people who are ideologically attached to free trade as a broken ideal that can be achieved after all, once it is repaired.
Can this approach work? Let’s look at the list of dubious assumptions underlying free trade to see if we can fix free trade by imposing policies that will make these assumptions hold true after all. To wit:
Dubious Assumption #1: Trade is sustainable.
Environmental sustainability is a problem intrinsic to the entire modern industrial economy. In the context of trade, the obvious solution is to tax trade that depletes nonrenewable resources or emits pollution.
Financial sustainability is, as analyzed at length in Chapter Two, achievable by controlling either trade or the countervailing financial flows that take place when trade is paid for.760
Problem fixable? Yes.
Problem fixable without ending free trade? No.
Dubious Assumption #2: There are no externalities.
For negative externalities like environmental damage, the obvious solution is to tax imports produced in harmful ways.
Positive externalities like technological spillovers can be addressed by tax credits for research and development. This is already U.S. policy to some extent, though without protectionism, this can just end up subsidizing research whose value is harvested by production abroad.761
Problem fixable? Yes.
Problem fixable without ending free trade? No.
Dubious Assumption #3: Factors of production move easily between industries.
This problem mostly comes down to labor. We can do little about it because the U.S. already has one of the most flexible labor markets in the developed world. We could expand palliative adjustment programs and worker retraining, but they have limited ability to solve this problem, for reasons analyzed in Chapter Three.762
Problem fixable? No.
Problem fixable without ending free trade? No.
Dubious Assumption #4: Free Trade does not raise income inequality.
Free trade raises income inequality in the U.S. because it lowers returns to the scarce factor of production (labor) and raises returns to the abundant factor (capital).763 It also impacts low-skill workers harder than high-skill workers.764 Although other nations have mitigated these problems somewhat with various nontrade policies, such egalitarian interventions are unlikely in the U.S.
Problem fixable? Yes.
Problem fixable without ending free trade? No.
Dubious Assumption #5: Capital is not internationally mobile.
The controls on international capital flows associated with a new Bretton Woods-style system of fixed exchange rates would help here,765 but this is not a free-trade solution unless we consider it free trade when we lack freedom in the financial flows that pay for that trade.766
Problem fixable? Yes.
Problem fixable without ending free trade? No.
Dubious Assumption #6: Short-term efficiency causes long-term growth.
As analyzed in Chapter Two,767 short-term efficiency can be downright destructive if people have short time horizons, something only authoritarian governments have been able to correct.
As analyzed in Chapter Nine,768 even with long time horizons, economic growth is largely about ladder externalities and related dynamics; most effective strategies for exploiting these are contrary to free trade because Ricardianism (and thus free trade) is about best exploiting immediate comparative advantage.769
Problem fixable? Yes.
Problem fixable without ending free trade? No.
Dubious Assumption #7: Trade does not induce adverse productivity growth abroad.
This problem is almost impossible to solve without abandoning free trade, as it concerns events in foreign nations not under our control. Ceasing to import so many goods, especially at the frontier of our trading partners’ technological capabilities, would obviously slow them down, but only somewhat.
Problem fixable? No.
Problem fixable without ending free trade? No.
Dubious Assumption #8: There are no scale economies.
The existence of scale economies is a fundamental fact of modern industry. They are not the product of any particular policy decision, so nothing can be done to make them go away, short of a return to premodern technological levels.
Problem fixable? No.
Problem fixable without ending free trade? No.
So it appears that we can’t plausibly hope to fix these eight problems without giving up free trade to some meaningful extent. The above list implies some policies to mitigate free trade’s harmful effects, but no fundamental solution that will redeem free trade as such.
IS THERE A NATURAL STRATEGIC TARIFF?
Here’s the nightmare that haunts all criticisms of free trade in this country: what if these criticisms imply that America needs a complicated technocratic tariff policy? This seems to be suggested by the complexity of the defects in free trade and by the fact that the nations which have most successfully repudiated free trade actually have complicated technocratic tariff policies. That would spell trouble, as the political difficulties of achieving such a solution in America are no secret. In the words of Alan Blinder, a member of Bill Clinton’s Council of Economic Advisors and former Vice Chairman of the Federal Reserve Board,
In Japan, industrial policy was and is run by a cadre of intelligent, respected, and powerful technocrats largely insulated from political interference and acting in the national interest. The United States, I am afraid, is too democratic for that. Political considerations would quickly overwhelm economic merits; industrial policy would more closely resemble life support for dying industries than incubation of emerging ones.770
It is, in fact, sorely tempting to take these political difficulties as an excuse to do nothing at all. The dangers of a special-interest takeover are not imaginary. But we can’t afford to quail at the challenge of making the politics work, as we are competing with rivals who have already done so. Like it or not, they have raised the bar for us. For the U.S. to concede that there exists an area of national policy this important that our rivals can master and we can’t is a decision in favor of voluntary national decline.
Billionaire investor Warren Buffett says that one of his criteria for investing in a company is that it must have a business that even a fool can run—because sooner or later a fool will. A similar philosophy should guide our construction of a tariff policy. We need a broad-based policy that can survive imperfect implementation and political meddling, a certain amount of which will be inevitable. We do not need an intricate, brittle, difficult policy that will only create work for bureaucrats, lawyers, and lobbyists. Among other things, any policy too complex for the public to understand will be beyond the reach of democratic accountability, the only ultimate guarantee that any tariff policy will remain aimed at the public good.
As noted in Chapter Six,771 one of the great puzzles of American economic history is how the U.S. once succeeded so well under tariff regimes that were not particularly sophisticated. This is where the idea of a so-called “natural strategic tariff” comes in. This idea says that there may be some simple rule for imposing a tariff which will produce the complex policy we need. The simple rule will produce a complex policy by interacting with the existing complexity of the economy. All the complexity will be on the “economy” side, not the “policy” side, so all specific decisions about which industries get protection, how much, and when will be made by the market. No intricate theory, difficult technocratic expertise, or corruptible political decision-making will b
e required.
There are obviously any number of possible natural strategic tariffs. The one we will look at here (which is probably the best) is actually the simplest:
A flat tax on all imported goods and services.
Prima facie, this is strategically meaningless because it protects, and thus promotes, domestic production in all industries equally. And if a tariff is going to win the U.S. better jobs, it will do so by winning us more positions in good industries (as defined in Chapters Nine and Ten). While a flat tariff would help with the deficit, which is good, it would provide the same incentive for domestic production of computer and potato chips alike, so it would not push our economy towards any industry in particular.
Or would it? The natural strategic tariff is a bet that it would. The key reason is this:
Industries differ in their sensitivity and response to import competition.
Although this is a complex issue, the fundamental dynamic is clear from the obvious fact that a flat tariff would trigger the relocation back to the U.S. of some industries but not others. For example, a flat 30 percent tariff (to pluck a not-unreasonable number out of thin air) would not cause the relocation of the apparel industry back to the U.S. from abroad. The difference between domestic and foreign labor costs is simply too large for a 30 percent premium to tip the balance in America’s favor in an industry based on semi-skilled labor. But a 30 percent tariff quite likely would cause the relocation of high-tech manufacturing like semiconductors. This is the key, as these industries are precisely the ones we should want to relocate. They have the scale economies that cause retainability, high returns, high wages, and all the other effects of good industries. Therefore a flat tariff would, in fact, be strategic.
TARIFF EFFECTS ON SCALE-ECONOMY INDUSTRIES
A natural strategic tariff would interact with different industries in surprisingly sophisticated ways. There is not the space here to discuss all of them, but the prime example is the fact that it would have different effects on industries that were at different points on their cost curves. The key here is that the cost curves of scale-economy industries tend to be concave. That is, they go down like this , not like this . This is not the place to go into the details, but a concave cost curve is actually guaranteed whenever production cost consists of an investment in capital (physical, human, or intellectual) plus an incremental cost for each additional unit of output.772 As a result, these curves have steeper slopes in their earlier than later stages. That is, they look something like the graph below, though their exact shape will vary:
The difference in slope between the early and late stages is the key. When a tariff is imposed upon an industry that is in the early stages of its cost curve, costs will fall rapidly with relatively small increases in output. So if we start at the far left on the graph above and move to the right a little, the curve goes down a long way. (See the graph on the next page.) Therefore, when a scale-economy industry in its early stages is given tariff protection and its sales increase as a result, it will enjoy a large cost decline. This induced cost decline will then improve its cost advantage over its foreign competitors by even more than the size of the tariff itself. This will result in further sales increases, further cost declines, and so on until its cost curve bottoms out. Therefore a flat tariff will, under these circumstances, trigger a virtuous cycle, and a fairly small tariff will produce a much larger ultimate cost advantage for the domestic producer. This advantage will outlast tariff protection and lock in retainability. (In the extreme case, this virtuous cycle will end only when the domestic industry has wiped out all its foreign competitors and become globally dominant.)
On the other hand, when a tariff is imposed on a mature industry (which, by definition, will be in the late stage of its cost curve), the slope of its cost curve will be relatively flat, so even fairly large increases in sales will not shift its costs very much. Therefore tariff protection will not trigger a virtuous cycle, and the domestic industry’s cost advantage will not greatly exceed the tariff itself. The net effect of strong impacts on some industries and weak impacts on others will be a bias towards stimulating industries that a) have concave cost curves and b) are on the early part of those curves. These will necessarily be nascent scale-economy industries: nascent because of b) and having scale economies because of a). In other words, the tariff will be self-targeting on precisely those industries we should want to target.
This mechanism is not perfect, and in the real world it would suffer a thousand quibbles, complications, and exceptions. Sometimes it would even backfire. But it is real enough to be worthwhile, as it is the effect on average that will matter.
This all resembles, of course, a classic infant industries tariff, but without the contentious problem of deciding which industries should be targeted or for how long. Strictly speaking, a true infant industries tariff would have the same effect as the natural strategic tariff, only more efficiently, as it would not waste any tariff incentive on mature industries, or on industries without concave cost curves. But as noted earlier,773 infant industries are only the first rungs of the larger phenomenon of the path-dependence of economic growth, so even a perfect infant-industries tariff would not be ideal.
TOO GOOD TO BE TRUE?
One is justifiably suspicious of cure-alls. So it is worth understanding why the natural strategic tariff is neither a gimmick nor too good to be true. It is neither a magic trick to make economic vitality appear out of nowhere nor a hammock to enable lazy and uncompetitive American industries to survive. Anyone supporting it for these reasons will be disappointed.
Fundamentally, the natural strategic tariff only works because it interacts with the existing competitive strengths of the U.S. economy. Specifically, it works because:
1. The U.S. is closer to being cost-competitive in good industries than in bad ones.
2. The U.S. domestic market is big enough to support scale economy industries.
3. A tariff has different effects on industries at different points on their cost curves.
It follows that if we do not cultivate the existing strengths of our economy, a natural strategic tariff will not do us much good. If implemented, such a tariff would be the cornerstone of trade reform, without which other measures will not work very well. But we would still need to fix our substandard education system, our crumbling infrastructure, and our short-termist financial system. We would still need to return to America’s Hamilton-to-Reagan tradition of industrial policy to some extent.774
THE REST OF THE WORLD
What about other nations? What should they do? The solution for most of them will not be the natural strategic tariff described above. If a Third World nation like Costa Rica, for example, imposed it, it would not push the Costa Rican economy towards good industries. The reason is that Costa Rica, unlike the U.S., is not closer to being cost competitive in good industries than in bad ones, as it is not richer than other nations in skilled labor, capital, or technological know-how. And because Costa Rica’s domestic market (smaller than Jacksonville, Florida)775 is too small to support, say, a full-scale aircraft industry, its domestic market is not a viable launch pad for significant scale-economy industries.
The right policy for a nation in Costa Rica’s position will be one centered on:
1. Avoiding trade deficits, asset sell-offs, and foreign debt, as discussed in Chapter Two.776
2. Avoiding free trade’s tendency to wipe out the most advanced sectors of developing nations, as discussed in Chapter Seven.777
3. Avoiding the authoritarianism associated with the WTO and related institutions discussed in Chapter Eight.778
4. Implementing the good newly-industrializing-country industrial policies discussed in Chapter Nine.779
The biggest choice developing nations must make is whether they are aiming to build up globally competitive industries and ultimately make the jump to the First World, or merely aiming to achieve comfortable (and by no means impoverished) mediocrity. Erik Reinert refers
to the latter as the lost art of creating middle-income countries;780 it is a concept basically nonexistent in mainstream economics. Such nations are inefficient by the standards of the global free market, but they are better off succeeding at the lower standard that they aim for as protected economies than tacitly aiming at world standards through free trade and failing. As noted in Chapter Seven, these nations aim to build up industrial sectors that, while inefficient, are still higher value-added sectors than the peasant agriculture to which their populations would otherwise be confined.781