by Ian Fletcher
Agriculture and other bad industries occasionally do exhibit innovation. But this is generally only in the production process, not the product itself. And innovations, when they come, tend to come from outside bad industries. Agriculture, for example, has benefited from genetic engineering and improved tractors, but created neither of these innovations itself. As a result, innovations do not establish virtuous cycles of innovation → more profit → more innovation inside bad industries. One telltale sign that a formerly good industry is turning bad is that product innovation exhausts itself and the industry turns to process innovation. And when a bad industry turns downright terrible, even process innovation exhausts itself and the industry just seeks cheaper labor. One can trace this process in individual industries over time. The shoe business, for example, began as a First World handicrafts industry, was mechanized over the second half of the 19th century, and began moving to the Third World in the 1950s—just as its productivity growth flattened out.
Which industries are good and bad changes over time as the technological frontier of the world economy evolves. The textile industry was good for a rising economy like Britain in the early 19th century, but is not good for developed nations today. These “has been” industries either migrate to developing nations or gently stagnate in place. As an economy accumulates more and more of them (as Britain’s did in the early 20th century) this is a clear sign that economic decline is around the corner.
No nation can plausibly hope to have its entire economy consist only of good industries, as some bad industries are sectors one can’t live without. One cannot go without haircuts, even if the productivity of barbers is no greater than 200 years ago. And even stagnant sectors contribute some output and employ some people. But the more of a nation’s economy is in good industries, the stronger its economy will be today and the better its growth prospects tomorrow.
ENTIRE ECONOMIES STUCK IN BAD INDUSTRIES
The poor and slow-growing economies of the Third World are that way because they are predominantly composed of bad industries and the path-dependence of industry entry traps them there. They are stuck in industries that have no increasing returns, no technological advances, and no ladder externalities. These problems mutually reinforce each other, ramify over time, and are the ultimate basis in hard economics for the well-known phenomenon that the rich get richer. (Karl Marx correctly observed that this happens, but he mistakenly thought that it was because the rich were exploiting the poor. Sometimes they do, but that’s not the fundamental problem.) The United Nations Development Programme has thus estimated that the income gap between the top fifth of the world’s nations and the bottom fifth was 3:1 in 1820, 7:1 in 1870, 11:1 in 1913, 30:1 in 1960, and 74:1 in 1997.681
Ironically, the gap between poor nations and rich ones was actually much smaller in the 19th-century heyday of colonialism than it is today. But this makes perfect sense: 150 years ago, there was relatively more to grab in colonies; there’s no point in an advanced nation conquering Rwanda today, when its own per capita output is 75 times greater. Insofar as colonialism, traditional or modern, overt or covert, is a deliberate economic strategy, it is about locking subject nations into bad industries.682 For example, the English government stifled Ireland’s nascent industrialization in 1699 by banning its exportation of woolen cloth outside the British Isles.683 Ireland obediently specialized in agriculture and even became a successful exporter. But it continued to export food even when its own people were starving. It had the capacity to produce nothing else.
Colonialism per se isn’t the problem here, as the economic mechanisms that do the damage are perfectly capable of operating in nations that are politically independent. (We noted Spain as an example of this in Chapter Six.) And not all colonies have been subjected to this treatment. The danger of being trapped in bad industries was, in fact, well understood by a number of small, raw materials-rich colonial nations which managed to avoid this fate. Australia, Canada, New Zealand, and (oddly enough) South Africa are the classic examples. All these nations were beneficiaries of the British Empire getting burned trying to make the U.S. a banana republic, as the imperial authorities allowed these colonies to raise tariffs against British goods in order to pursue their own industrialization.684
THE PATHOLOGIES OF BAD INDUSTRIES
Free trade does not automatically assign nations good industries. This is the fundamental problem. Acting according to their immediate comparative advantage, it can just as easily assign them bad ones. This may be optimal in the short term, but if a nation’s comparative advantage today is in producing bananas, then it will be stuck with roughly the same productivity 30 years from now.685 That isn’t true in industries like computers or automobiles. And in the presence of a large wage gap between nations, low-wage nations will automatically tend to attract bad industries under free trade, as here will lie their immediate comparative advantage.
The interaction of free trade with bad industries is toxic in a wide variety of ways, not all of them obviously trade-related. For example, free trade tends to exacerbate all the “bad habits” of modern agriculture, as the attempt to extract more returns from a diminishing-returns industry to keep up with declining terms of trade generates a relentless squeeze. This tends to increase the amount of land under cultivation and undermine conservation programs. It tends to force intensive use of pesticides and fertilizers. It tends to replace diversified operations with large-scale feed lots and monocrop agriculture. It tends to reduce specialty crops to commodities. And it tends to place absentee owners in control, undermining family farming and rural communities.
The First World palliates (not the same thing as solving) these problems with agricultural subsidies because it can afford to. This has the unfortunate spillover effect in the Third World of generating a tidal wave of cheap exported food that destroys farm jobs the same way manufactured imports destroy factory jobs in industrial countries. Given that the combined agricultural subsidies (including hidden ones such as cheap water) of the U.S., EU, and Japan equal almost 75 percent of the entire income of Sub-Saharan Africa, it is no accident that African farmers, for example, cannot compete.686 Once they can no longer support themselves on the land, they have no choice but to seek urban, mainly slum, life. When a Third World nation converts its food production to export and becomes dependent on imported food, it becomes vulnerable to volatility in its export markets. Bubbles in commodities such as biofuels make this worse, as when the bubble ends, it is impossible to convert back to food production in time to avoid food riots.
Raw materials extraction is the other sector notorious for bad industries. It tends to harbor many of the same pathologies as agriculture, plus a few nasty quirks of its own. Raw materials like oil notoriously breed parasitic elites composed of whomever manages to establish political control of the spot where the oil comes out of the ground. Unlike the elites of manufacturing-oriented economies, they contribute little in managerial or technical skill to the economies they dominate. They can get away with misgoverning their countries in ways that would ruin the productivity of a manufacturing-oriented economy. They have no need to share widely the wealth derived from the raw materials they extract (except with local warlords, security forces, and the police), and little incentive to reinvest more than a fraction of that wealth in the industry itself.
During the Cold War, much opposition to capitalism was motivated not so much by literal hatred of private property (let alone actual love of communism) as by the deep-seated fear that advanced industrial modernity was a closed club of the United States and Western Europe.687 Other nations, it was feared, could never break in, but would remain eternally trapped in bad industries—which would then guarantee their poverty and political subordination. So socialism was the only way out, with the USSR as its ultimate geopolitical anchor, even if obviously extreme as a literal economic model to imitate. But once a nation understands the above mechanisms of underdevelopment—better yet, how to manipulate them through protection
ism and industrial policy—abandoning capitalism entirely shows itself to be an unnecessarily radical solution. Japan and its followers in East Asia understood this, which is a big part of why they were so staunchly anticommunist during this period. Other parts of the world did not, and thus found socialism considerably more interesting.
GOOD AND BAD INDUSTRIAL POLICY
If free markets and free trade aren’t always best, this necessarily opens up the possibility that some other policy might be better, if properly designed and implemented. This, at bottom, is what makes successful protectionism and industrial policy possible.
It is no accident that when reviewing purported free-trade success stories around the world, one often finds protectionism and industrial policy right under the surface. In Brazil, for example, the steel and aircraft industries are legacies of past import-substitution policies;688 in Mexico, motor vehicles are; in Chile, grapes, forest products, and salmon. In fact, of the top 20 exporting corporations in Chile in 1993, at least 13 were creations of a single government agency, the Corporación de Fomento de la Producción (CORFO).689
Over the last 40 years, there have been two key laboratories of protectionism and industrial policy: East Asia and Latin America. As recently as the early 1970s, both regions were at similar levels economically, and Latin America was actually much richer at the end of WWII.690 And yet East Asia has succeeded economically, while Latin America has stagnated since about 1975. (The above examples are happy exceptions.) Protectionism and industrial policy clearly come in both effective and ineffective varieties, and neither concept deserves an uncritical endorsement.
We are now in a position to understand why some kinds of each work and some don’t.691 In the words of Dani Rodrik, both regions employed the “carrot,” that is, tariffs, industrial subsidies, et cetera, to help their industries. But only East Asian governments were politically disciplined enough to employ some needed “stick” as well, i.e., measures to prevent their industries from merely converting this help into immediate profits, not long-term upgrading of their capabilities.
An export requirement is one example of a “stick.” This improves the nation’s balance of payments and forces domestic producers to meet global standards for quality and cost. This policy can be implemented in a wide variety of ways, some not immediately obvious as such, like giving companies import quotas for raw materials based on their export performance. Another method is a so-called “rolling” local content requirement, where a company importing goods is required to produce a gradually increasing percentage of the final value of the product domestically. This creates a pressure to produce locally without getting so far ahead of market outcomes as to be hopelessly inefficient.
Other patterns of successful industrial policy emerge. It has tended to maintain domestic rivalry within industries, rather than concentrating resources on a single superficially-strong national champion.692 It has tended to involve local ownership and understanding of core technologies, rather than the “Lego brick” manipulation of sophisticated inputs in an unsophisticated way. It has tended to combine investment in education with investment in industries that can actually absorb educated workers. It has tended to use access to the national market as leverage to get foreign corporations to locate a share of production there, not merely as a shield for domestic producers or as a source of tariff revenue to be wasted on political pork. (Pulling in state-of-the-art foreign producers also keeps domestic producers on their competitive toes without subjecting the economy to an uncontrolled flood of imports.)
What did Latin America do wrong? It allowed domestic competition to wane. It permanently protected mature industries that should have been able to survive on their own by that point. It lacked an interest in exporting, so its industries were not disciplined to reach world standards. Lack of exports then caused a lack of the foreign currency needed to import state-of-the-art production technology. Its industrial know-how therefore lagged behind the rest of the world, as it never developed comparable domestic sources of technology either. And Latin American nations either failed to emphasize education, or failed to create industries that could absorb educated workers, the latter causing investments in education to dissipate in brain drain abroad rather than accumulate as human capital at home.
THE WORST AND BEST INDUSTRIAL POLICY
In the developing world, the very worst industrial policy has tried to break into new industries merely on the basis of having cheaper factors of production, which mostly comes down to cheap labor. (Number two is probably cheap raw materials, followed by cheap land.) Unfortunately, industries based on cheap labor continually attract new entrants because cheap labor is an undifferentiated commodity, available all over the world. But incumbents are blocked from exit by costs they have already incurred, trapping them in these industries. Today’s cheapest labor source is always vulnerable to being undercut by an even cheaper one tomorrow, and rival governments will subsidize even where there is no preexisting cost advantage.693
A nearby example of this misguided strategy is Mexico’s string of maquiladora plants along the U.S. border. These 3,000 American-owned factories employ over a million workers. Though they often contain the latest production technology and have the highest productivity of any industry in Mexico, they have spawned no industrial revolution there. Although these plants often consume fairly sophisticated technology, in the form of imported capital equipment, what they do with this technology is not especially sophisticated. So the Mexican economy accumulates neither human nor any other kind of capital; the products produced there have no scale economies at the assembly stage.694 For example, according to Rick Goings, CEO of Tupperware, which has a major plant in Toluca, Mexico failed to grasp the opportunity handed to it by NAFTA:
When all of a sudden the borders opened and all these [jobs] were created for assembly and sending [products] back to the United States, they didn’t invest what they needed to in building the skill base of Mexican workers. So you go down there now and what are they complaining about? Losing their jobs to China. All you have to do is follow Nike’s pattern over the last 25 years: Korea, China, Vietnam. You just keep following that low labor cost—you just keep following that dragon. Unless you build in these countries an infrastructure and a skill base, they may have a short-term advantage, but it won’t last.695
Such industry is a technological and economic dead end. For all that anybody will ever learn or develop by working in it or even owning it, they might as well be picking coffee beans by hand—or owning a plantation. The question a nation should always be asking about its industries is, “Is there anything left to learn here?” If there isn’t, it’s time to let another nation further down the ladder of industrial development take over that industry and move on. And if it isn’t feasible to move on, then something is wrong with the nation’s industrial strategy, because it has gotten stuck, and growth requires that it continually be able to upgrade.
What does the most successful industrial policy look like? As economies try to make the jump from the Third World to Newly Industrialized Country status and finally to the First World, the real key to growth turns out to be proactively anti-Ricardian, namely getting away from their immediate comparative advantage. This key is therefore profoundly contrary to free trade. Above all, this means getting away from advantage based merely on given factors of production and transitioning towards advantage based on created factors of production. Ultimately, it means transitioning from so-called lower-order sources of advantage to higher-order sources. As Michael Porter explains it:
Lower-order advantages, such as low-cost labor or cheap raw materials, are relatively easy to imitate. Competitors can often readily duplicate such advantages by finding another low-cost location or source of supply, or nullify them by producing or sourcing in the same place…Also at the lower end of the hierarchy of advantage are cost advantages due solely to economies of scale using technology, equipment, or methods sourced from or also available to competitors….
Higher-order advantages, such as proprietary process technology, product differentiation based on cumulative marketing efforts, and customer relationships protected by high customer costs of switching vendors, are more durable. Higher-order advantages are marked by a number of characteristics. The first is that achieving them requires more advanced skills and capabilities such as specialized and highly trained personnel, internal technical capability, and, often, close relationships with leading customers. Second, higher-order advantages usually depend on a history of sustained and cumulative investment in physical facilities and specialized and often risky learning, research and development, or marketing.696 (Emphasis in the original.)
INDUSTRIAL POLICY, AMERICAN-STYLE
For contemporary Americans, one common roadblock to understanding industrial policy and protectionism is the myth that our most successful industries have made it on their own, without government help. We tend to see industrial policy (if we accept it at all) as perhaps suitable for up-and-coming nations, but not for nations like ourselves that have already arrived. But in reality, the fingerprints of industrial policy are easy to find in our own economy, even in the post-WWII era of increasingly free trade (and increasingly strident laissez faire rhetoric after about 1980).697 Let’s look at two of America’s most touted industries, semiconductors and aircraft, to see how they really became so strong—and thus why the purely free market model of economic growth is so wrong.