He's right and he's right? They can't both be right!

Ryan Avent, perhaps prodded by my comment-thread heckling (yeah...in my dreams!), has started making a lot more sense in his arguments against the industrial-policy-ists:

And over the long-term, the needs in America are the same as they have been for decades. The country has been underinvesting in the basics, and the effects have become ever clearer. Its infrastructure—from transportation to water and power to broadband—is wholly inadequate. Educational attainment is plateauing and may begin to decline. Its labour market policies aren't up to handling sectoral shifts or millions of people in long-term unemployment. Its research funding is ample for any project with a potential military application and meagre for everything else. The weaknesses are clear.

This is absolutely right. The collapse of America's pubic-good provision is the single biggest problem with our economy. Trade issues, and industrial policy, are secondary to our crumbling infrastructure and dysfunctional education system.

And yet, when Tim Duy says things like this, it's hard to ignore the sneaking suspicion that maybe trade-imbalance and industrial-policy issues are a bigger part of the picture than we realize:

[I]n aggregate real wages and nonfarm payrolls have been stagnant for a decade. Where are these high wage paying jobs? Or even median wage paying jobs at this point? Silly me, I actually believe the unapologetic and unquestioning supporters of free trade need to answer this question. We are millions of jobs below trend, and we have lost millions of jobs in manufacturing - the manufacturing of goods that we still consume, no less. Moreover, these two trends occurred in the same decade, in concert with a third trend - the sharp rise in foreign official reserve accumulation. How can you not be even allowed to suggest that there just might be a connection?

The problem I see with Avent and his fellow-travelers (Yglesias) is not that they are wrong about the structural problems facing our economy - the lack of investment in public goods, the Republican commitment to ever-expanding deficits - but that they are so dogged in their belief that these are the only problems we face. Perhaps they are afraid that they'll labeled as economic know-nothings if they question the free-trade orthodoxy (a legitimate fear, given that most of the Grand Old Men of the economics professions are still ideological conservatives). Or perhaps a trade war with China is what they fear most.

But either way, if Tim Duy's question turns out to have an answer we don't want to hear - an answer that implies that a strong-dollar policy and a finance-favoring industrial policy have hurt our economy - then writers like Avent and Yglesias are preventing liberals from coming up with a workable answer to America's economic woes. Which, in the long run, is bad news for the liberal movement.

Trade models I don't teach - The "New Trade Theory"

I was going to have three posts on trade models - the two I teach, and the one I don't teach. But I realized that the middle model - the Heckscher-Ohlin model - is not different enough from the Ricardian model to merit its own post.

(Briefly, the Heckscher-Ohlin (HECK-sure o-LEAN) model tries to describe trade between rich and poor countries, by pointing out the fact that rich countries have a lot of equipment and machines and buildings and infrastructure (capital), while poor countries have a lot of cheap labor, so poor countries will make labor-intensive goods (e.g. clothing) while rich countries will make capital-intensive goods (e.g. cars). Fine. But the thing is, Heckscher-Ohlin treats each country like a single person, just like the basic Ricardian model, and so there can't be any negative externalities to trade...hence, the model assumes right off the bat that free trade is always and everywhere a good thing. Yawn.)

Instead, I want to discuss the "New Trade" model, invented by Krugman in 1979 and refined in 1991, for which (among other things) he won the Nobel prize a couple of years ago. This is the model that most economists actually use now to explain and predict trade. I am never asked to teach it in Econ 102, but that's only because it involves lots of math that most undergrads can't handle.

Krugman's model was developed in order to explain the fact that many trading partners produce very similar goods: Japan and America both produce cars, for example. The classic Ricardian theory of trade says that comparative advantage determines what gets imported and what gets exported, but the data seem to buck that. So Krugman brought in a different idea. People like variety, he said, so different countries will trade different varieties. Small initial differences between countries (example: Japan had no domestic oil reserves, and so focused more on smaller, more efficient cars) determine which variety gets produced where.

OK, so far, so good. Now, what about free trade? Unlike the Ricardian and Hecksher-Ohlin models, New Trade DOES have an externality. Specifically, since people like variety, the more variety I produce and sell, the more you can produce and sell as well. Each company becomes sort of like a monopolist, being the only producer of its own specialized variety of good. More trading countries means more global variety, making consumers better off everywhere.

Of course, this externality is purely a positive externality. Japan selling us cars is good because it allows us to focus on producing a different type of car. There is no negative side effect from increased trade, and so free trade, again, becomes by assumption the best possible policy, always and everywhere. The free trade orthodoxy is still baked into the cake.

In order to get a model that says that free trade might not be the best policy, we'd need to write down a model that has a negative externality to trade. Perhaps opening up trade allows companies to put their dirty factories in the country with the most lax pollution regulations, hurting the environment. Perhaps opening oneself up to trade creates financial instabilities because of international investment flows. Or perhaps countries can engage in "predatory" trade practices, to which the best response is a temporary trade war.

These models all exist (see links), but no one has yet paid them much attention. Instead, whether in the classroom or in our actual modeling, we continue to unquestioningly use models that assume that no trade restriction can ever be a good idea.

Now, I'm not saying these are bad models! And I am not saying that free trade is not, in fact, the optimal policy. What I am saying is that in every intro macro course, one of the first (and last) things we say is: "Macroeconomists agree on very few things, but they generally agree that free trade is good." And this is true, they do. But the reason that they agree is not that they've proven it. It's that they assumed it, and decided to use models that assumed it. When the inevitable chorus of economists turns out to cheerfully berate anyone who suggests "protectionism," they may be reveling in their rare opportunity to finally agree on something...but really what they are doing is drowning out dissent for the fun of it. Our profession-wide consensus comes at the expense of scientific honesty - and, if it turns out that the consensus is wrong, it has also come at the expense of American prosperity.

Michael Spence joins the heretics

Michael Spence, winner of the 2001 economics Nobel, joins the heretic movement that was given voice by Andy Grove:

The structural evolution of the US economy over the past 15 years has been driven by excess consumption, enabled by debt-fuelled asset inflation. The crisis put a stop to this, but structural deficiencies remain. America’s export sector is too small and underdeveloped. The financial sector became outsized, and is down-sizing.

A pattern of underinvestment in infrastructure has left the economy less competitive than it should be. Energy pricing issues have been ignored, causing underinvestment in urban infrastructure and transport. The education system has widespread problems with efficiency and effectiveness...

The real issue is employment: not just stubbornly high unemployment, but a bigger problem described recently in a thoughtful article by Andy Grove, the long-time chief executive of Intel. He argued that manufacturing is vanishing in the US, a trend that must be reversed. The question is how.

There is little doubt that America’s social contract is starting to break. It had on one side an open, flexible economy, and on the other the promise of employment and rising incomes for the motivated and diligent. It is the second part that is unravelling.

Incomes in the middle-income range for most Americans have stagnated for more than 20 years. Manufacturing jobs are moving offshore. Globally the set of goods and services that is tradable is expanding, but the US and other advanced countries are not competing successfully for an adequate share of the tradable sector.

The employment effects of these trends over the past 15 years have been masked by excess consumption and the overdevelopment of sectors such as finance and real estate. The latter are now set to shrink, as multinational companies grow where they have access to high-growth emerging markets in Asia and Latin America. Such companies will locate their operations where market and supply chain opportunities lie. In the tradable sector, in manufacturing and in a growing group of services, that means outside advanced countries.

The availability of low-cost, disciplined labour forces in developing countries reduces the incentive for these companies to invest in technologies that enhance labour productivity in the tradable sectors of the advanced economies. As a result, the evolving composition of advanced economies is increasingly weighted towards the non-tradable sector, combined with a set of high-end tradable services where both human capital and proximity matter. The rest of the tradable sector is shrinking.

The shrinkage creates problems. Over-specialisation could threaten independence and national security. Spillovers between R&D, product development and manufacturing will be lost if manufacturers leave. Employment will stagnate. Income distribution will move adversely and the social contract will erode further...

To avoid an outbreak of protectionism, there has to be an alternative. President Barack Obama’s new export council, announced on Wednesday, is a step in the right direction. But a bolder move is needed: a broad public-private partnership to invest in the development of technology in parts of the tradable sector where there are opportunities to make advanced countries competitive. The goal must be to create capital-intensive jobs that have labour productivity levels consistent with advanced country incomes.

Would this damage developing countries? Clearly not. The US (or even developed economies combined) does not have hundreds of millions to employ. A targeted programme would leave the vast majority of labour-intensive manufacturing right where it is now: in the developing world. With new credible growth strategies in America (and other advanced countries) developing countries may even be willing to play an important complementary role in restoring global demand through, for example, the reduction of excess savings.

I added the emphasis, in order to note something very interesting: Spence obviously believes that export competitiveness is heavily influenced by the kind of public goods that I am always ranting about: infrastructure, education, and most importantly research. It is precisely these public goods that conservatives have neglected over the past 20-30 years, in their zeal to eliminate all government intervention in the economy. But it may be precisely these public goods that enable a country to have good high-paying middle-class jobs.

Anyway, blogger Mark Thoma already endorses Spence's heresy. Paul Krugman hasn't addressed this particular controversy yet, but has made noises about China's currency policy. Could we have a movement on our hands?

(If so, one positive sign for the industrial-policy movement is the fact that the first people to criticize Grove's article have been editorialists from The Wall Street Journal and Forbes' Reihan Salam. As they say, "by their enemies shall you know them"...)

Trade models I teach - the Ricardian trade model

In my last post, I wrote:

I myself have participated in this, in the Introductory Macroeconomics course I TA for at the University of Michigan. We always include a lecture that pits pro-free-trade arguments against anti-free-trade arguments, and we make sure that the former always come out the "winner" at the end of the day. We use the very simplest models of trade, and we never mention industrial policy or the manufacturing sector or mercantilism. And another four hundred students tramps off to be America's next generation of businesspeople, lawyers, journalists, professors, and voters...

So to follow up on this, I thought I would explain the trade models I teach - and those I don't teach but which are in common use - and illustrate why these models say that free trade is always good. The first model I talk about will be - must be - the classical, or "Ricardian" trade model, which is the one I teach in introductory macro courses.

The Ricardian trade model deals with aggregate trade between countries - in other words, it treats each country as a single person, saying "Brazil can produce 100 cows or 50 peppers, Argentina can produce 75 cows or 25 peppers." You may already be able to see how this setup already assures that free trade is always optimal. The reason is: if Brazil is a single person, and Argentina is a single person, they wouldn't agree to trade if trade weren't good for both of them.

This is actually a very deep principle in economics. It's called the "first welfare theorem," and it says: two people will not engage in any economic transaction that doesn't benefit both of them. Thus, any economic transaction that happens - such as international trade, for example - must make both parties better off. the only way a transaction is bad is if it has some kind of effect on a third party not involved in the transaction - a side effect, or "externality." An example is pollution; when I sell you a piece of metal that I mined by dumping arsenic in a river, people get hurt who had no say in whether or not we did the deal.

But if there's only two people in the world ("Brazil" and "Argentina"), there can't be any externality, by definition! Thus, no matter what kind of forces drive international trade in the Ricardian setup, it must conclude that free trade is always and everywhere good. Game over.

But just in case you want to know, the Ricardian theory is all about something called "comparative advantage". This means that each country specializes in the thing it can produce relatively cheaply. To use an example from Mankiw's intro textbook, suppose I'm a lawyer who can make $2000/hr. lawyering or $100/hr. typing. And suppose my secretary types only half as fast as I do. It still makes sense for me to pay my secretary to do the typing, even though I'm twice as good at typing as he is, because I can make more money for an hour of lawyering than for an hour of typing. Even though I have an "absolute advantage" in typing, my secretary has a "comparative advantage" in typing, because he can't lawyer at all. It's impossible to have a comparative advantage in everything, so countries will always have an incentive to trade.

So that's how that works. But notice - "comparative advantage" is not necessary to conclude that free trade is always good. That conclusion was assured by the fact that the model treats each country as a single person. The mere fact that free trade allows more trade is what makes it automatically good in that sort of setup, comparative advantage or no; if trade is bad for one of the country-persons, that country-person will simply refuse to trade!

In fact, the only argument one can muster against free trade in the Ricardian setup is the "winners and losers" idea. This is the idea that, even if a country as a whole benefits from free trade, some people inside the country (for example, producers who lose their jobs when cheap imports put them out of business) might lose out. And this is the "argument against trade" that we always trot out in Econ 102 to explain why people aren't always happy about trade. But in the end, this argument always loses, because we just say "Instead of restricting free trade, we should find ways to have the 'winners' compensate the 'losers'" (you see those terms, "winners" and "losers," a lot in articles about trade). And voila, free trade is good.

So if you read a commentary by some smarty-pants columnist who says "Of course free trade is good, because comparative advantage exists, you dummy!", then now you realize that he's wrong. If externalities exist, then comparative advantage might not be enough to make free trade good. If you have a model that assumes no externalities, like the Ricardian model and many others, then free trade is good whether there's comparative advantage or not.

But where's the proof? If there are no externalities (and as long as a nation distributes its gains from trade properly among its citizens, which we assume it can), then free trade is always good. But how do we know there are no negative externalities to trade? The Ricardian model just assumes they don't exist! Of course, you can't prove a negative, so the burden of proof is on a free trade opponent to come up with an idea for a negative externality (a few have tried; so far, they have not been paid much attention). But merely having an assumption in place that no negative externality exists is not the same as evidence in favor of free trade.

If I were to teach trade models honestly, I would say: "Look, if these assumptions are right, then free trade is always good. But the fact is, we're not sure they're right, and we need to look into the matter a lot more before we bet our credibility on the statement that free trade is always good." But I don't teach trade models honestly, because scientific skepticism will not help my students pass their exams.

The sum of all heresies

Tim Duy weighs in on the industrial policy discussion touched off by Andy Grove's article in Business Week. His post is telling, because it gives voice, all at once, to all the nagging suspicions that are gnawing away at the guts of so many observers of current events. He writes:

Only one word describes the American labor market outcome of the last decade - abysmal. Not only is job growth well below trend, but the quality of jobs is in question. The jobs deficit is even more striking considering the supposed gains in productivity over the past 15 years. Job growth should not stagnate. Resources - including labor - released via higher productivity are supposed to be channeled into expanding sectors. Moreover, productivity growth is supposed to yield improved economic outcomes via higher real wages. Yet as spencer famously shows, labor's share of output has been steadily decreasing since the early 1980s...

Why has the American jobs machine failed so spectacularly? This should be the most pressing issue facing economists and policymakers. Are either up to the task?

Yves Smith directs us to an intriguing piece by Andy Grove, former CEO of Intel. Rajiv Sethi follows up and summarizes...I think the Yves-Sethi conversation is remarkably important, and should lead one to reexamine the importance of the manufacturing sector. I admit that in past years I tended to dismiss the manufacturing sector, seeing its relative decline as simply a transition to more productive knowledge-based work...

The loss of manufacturing capacity in the Groves scenario critically impacts the potential growth of the nation...

Note that a number of trends all begin in the 1980s. Absolute manufacturing declines, the rise of persistent trade deficits, the decline in labor's share of output, growing income inequality, and the Great Moderation. That the combination of these trends is coincidental seems unlikely...

If manufacturing is critically important to driving trends of national well being, an exploration of the decline of that sector is crucial. But that exploration almost always leads back to a very difficult place - international trade. And every right minded economist and policymaker knows unequivocally that free trade is good, and to even question that assumption makes one an ignorant heretic who has never heard of Smoot-Hawley. Therefore, the examination ends. Manufacturing's decline simply cannot be a problem if it is consequence of international trade because everyone knows international trade is good.

[The modern pro-free-trade orthodoxy] fails to acknowledge that while free trade produces net positive effects, that process can certainly be upset by the deliberate manipulation of currency values. And make no mistake, those values have been manipulated. There can be no other excuse for the massive buildup of official reserve assets in global central banks.

I don't think it is a coincidence that the absolute decline in manufacturing accelerated in the wake of the US strong Dollar policy, which provided the freedom for China to pursue an aggressively mercantilist economic strategy, perfecting what Japan's policymakers began in the 1980s. Thus I don't think the pernicious hollowing out of America's industrial base is simply the result of comparative advantage. I grow increasingly convinced that the disappointing economic outcomes of the last decade are the culmination of decades of industrial neglect...And I am increasingly convinced that these trends have been largely dismissed by the economics community because acknowledging them would cast doubt on value of free trade, failing to recognize that currency manipulation was turning free trade into a zero-sum game.

In short, I have become a heretic.

But what then is the appropriate policy response? Unfortunately, we are held captive by fears of a debilitating trade war...

Bottom Line: Something more than cyclical forces is weighing on the American jobs machine. Here I have tried to extend the Grove/Smith/Sethi discourse with additional focus on absolute declines in manufacturing jobs and distressing declines in capacity growth rates. These trends may be critically important in understanding the dismal performance of US labor markets. If they are in fact critical, they raise serious questions about US trade policy - questions that few in Washington want to address. Given the extent to which manufacturing capacity has already been offshored, those questions go far beyond the recently announced tiny shift in Chinese currency policy. Simply put, accepting the importance of manufacturing capacity and the possibility that offshoring has had a much more deleterious impact on the US economy than commonly accepted would requrie a significant paradigm shift in the thinking of US policymakers. If you scream "protectionist fool" in response, then you need to have a viable policy alternative that goes beyond the empty rhetoric of "we need to teach better creative thinking skills in schools." That answer is simply too little too late.

Tim Duy's post is not an economics paper, and it does not prove his case. But it tells a story (which is all most econ papers do anyway) that weaves together a number of disturbing threads - the decline of American manufacturing, Chinese mercantilism, rising inequality, stagnant real wages, and job insecurity. This is a story that is not supported by any economic model that I know of...but, given economists' ability to write down a model that supports any conclusion, and given the tendency of models that defy the free-trade orthodoxy to languish unread in third-rate journals, I'm inclined to give Duy a pass for now.

The fact is, many people instinctively believe that mercantilism can work if your trade partners don't fight back; that the export (or import-substituting) manufacturing sector is especially important for job growth and living standards; and that it is no coincidence that our inequality, stagnant living standards, and insecurity have worsened as our trade balance with China (and with the other Asian nations in China's supply chain) has worsened. These beliefs are considered heretical, but they just refuse to go away.

Now, many instinctive, heretical beliefs are actually totally false: for example, "young-Earth" theories, global warming denial, or disbelief in quantum mechanics. But in the case of the free trade orthodoxy, there is no hard evidence that free trade is always and everywhere the optimum policy. There is no fossil record, no climate record, no electron double-slit experiment that free-traders can hold up and say "Look, unbelievers! HERE'S THE PROOF YOU ARE WRONG!" Instead, economists have a bunch of models, all of which assume that free trade is good and proceed from there.

So instead or arguing from evidence, the free-trade orthodoxy relies on shouting down and insulting any who deny its precepts. I myself have participated in this, in the Introductory Macroeconomics course I TA for at the University of Michigan. We always include a lecture that pits pro-free-trade arguments against anti-free-trade arguments, and we make sure that the former always come out the "winner" at the end of the day. We use the very simplest models of trade, and we never mention industrial policy or the manufacturing sector or mercantilism. And another four hundred students tramps off to be America's next generation of businesspeople, lawyers, journalists, professors, and voters...

And yet the nagging little voice in the backs of our heads continues to whisper the things Tim Duy has dared to say in the light of day. It continues to tug and worry at the corners of our faith in free trade, crying "E pur, non e libero!"

Andy Grove comes out in favor of industrial policy

Andy Grove, the famous Intel boss who built the company into the world's leading semiconductor manufacturer, has come out in favor of old-fashioned industrial policy. His arguments are twofold: 1) that industrial policy creates more middle-class jobs than our current approach, and 2) that industrial policy creates "network effects" among industries that give advantages tomorrow's technology startups. Some excerpts from his article:

On job creation:

You could say, as many do, that shipping jobs overseas is no big deal because the high-value work—and much of the profits—remain in the U.S. That may well be so. But what kind of a society are we going to have if it consists of highly paid people doing high-value-added work—and masses of unemployed?

Since the early days of Silicon Valley, the money invested in companies has increased dramatically, only to produce fewer jobs. Simply put, the U.S. has become wildly inefficient at creating American tech jobs.

On industrial network effects:

There's more at stake than exported jobs. With some technologies, both scaling and innovation take place overseas.

Such is the case with advanced batteries. It has taken years and many false starts, but finally we are about to witness mass-produced electric cars and trucks. They all rely on lithium-ion batteries. What microprocessors are to computing, batteries are to electric vehicles. Unlike with microprocessors, the U.S. share of lithium-ion battery production is tiny (figure-E).

That's a problem. A new industry needs an effective ecosystem in which technology knowhow accumulates, experience builds on experience, and close relationships develop between supplier and customer. The U.S. lost its lead in batteries 30 years ago when it stopped making consumer electronics devices. Whoever made batteries then gained the exposure and relationships needed to learn to supply batteries for the more demanding laptop PC market, and after that, for the even more demanding automobile market. U.S. companies did not participate in the first phase and consequently were not in the running for all that followed. I doubt they will ever catch up...

[In the U.S. there is] a general undervaluing of manufacturing—the idea that as long as "knowledge work" stays in the U.S., it doesn't matter what happens to factory jobs. It's not just newspaper commentators who spread this idea. Consider this passage by Princeton University economist Alan S. Blinder: "The TV manufacturing industry really started here, and at one point employed many workers. But as TV sets became 'just a commodity,' their production moved offshore to locations with much lower wages. And nowadays the number of television sets manufactured in the U.S. is zero. A failure? No, a success."

I disagree. Not only did we lose an untold number of jobs, we broke the chain of experience that is so important in technological evolution. As happened with batteries, abandoning today's "commodity" manufacturing can lock you out of tomorrow's emerging industry.

And on what to do about it:

The first task is to rebuild our industrial commons. We should develop a system of financial incentives: Levy an extra tax on the product of offshored labor. (If the result is a trade war, treat it like other wars—fight to win.) Keep that money separate. Deposit it in the coffers of what we might call the Scaling Bank of the U.S. and make these sums available to companies that will scale their American operations. Such a system would be a daily reminder that while pursuing our company goals, all of us in business have a responsibility to maintain the industrial base on which we depend and the society whose adaptability—and stability—we may have taken for granted.

Now, in many intellectual circles in America it has come to be regarded as an article of faith that "protectionism" and "industrial policy" are the road to ruin (just as in Asia the exact opposite has become an article of faith). Just the other day, for example, a friend of mine - a very smart lawyer - suggested that Washington Post columnist Steven Pearlstein be "flayed" for arguing that only the threat of tariffs can get China to change its currency policy.

But after actually reading some trade theory and development theory at the highest academic level, I have to say that the only thing I'm more sure of is how little I ought to be sure of when it comes to this topic. Sure, trade in general is awesome. But to let the light of that basic truth blind us to the subtler questions of industrial policy, strategic trade interactions, etc. is as absurd as to let the fact that communism failed blind us to the efficiency of some kinds of government intervention in the economy.

The bottom line: I am not sure whether Andy Grove is right about industrial policy. But I am reasonably sure that no one else is sure whether he's right, especially the vast commentariat that shrieks "trade war" and "protectionism" whenever anybody suggests anything like what Grove is suggesting. And I am sure that Andy Grove, though old, is a very smart guy. And when very smart guys - like Grove, or like Paul Samuelson, the greatest economist of our age - start saying things that contradict our conventional wisdom, we should at the very least pay close attention.

Update: Yves Smith and Rajiv Sethi respond to Grove's article. Smith is broadly supportive. Sethi is skeptical, but, like me, recognize economists' fundamental ignorance in these sort of matters and are made uneasy by the possibility that Grove is onto something. Tyler Cowen, meanwhile, goes for an off-the-cuff defense of the conventional wisdom. Mark Thoma agrees with me and Sethi, and points out (correctly, in my opinion) that the social inequality case for industrial policy is stronger (so far) than the innovation-and-efficiency case.