Maastricht Chronicles

Wednesday, March 24, 2004

The best neoclassically-inspired rebuttal I've found yet 

I'm in trouble now. I've spent the last few days reading New Growth Theory papers - ie the literature that arose in the 80s and 90s to take account of non-convergence (i.e. no evident trend towards rich and poor countries converging in their living standards) and to try to solve Neoclassical theory's "innovation problem." We're talking Lucas and Romer mostly, but also Grilliches, Grossman and Helpman, Laura Tyson, Mankiw, Stiglitz, even a bit of Krugman.

Lucas and Romer are definitely the most perceptive analysts in this bunch, and I have to admit my infatuation with Krugman has abated significantly, because he doesn't really seem to understand the nature of the innovation problem. Romer, on the other hand, takes a very peculiar and piercing look at the problem founded on a clear eyed understanding of the nature of the innovation problem. His conclusions straddle the border between neoschumpeterianism and neoclassical thinking - either an arroz con mango or a brilliant synthesis, depending on your mood.

His Oct. 1990 paper in the Journal of Political Economy called "Endogenous Technical Change", is particularly fascinating. He starts - as everyone in the field does - with a simultaneous reverence and critique towards the famous Growth Accounting Model proposed by Robert Solow in the 1950s. It's the model that started the debate, and controversies in the field tend to be structured in terms of what each theorist thinks Solow got right and what they think he got wrong.

Solow's model is disarmingly simple. Economic goods, as everyone knows, are a mixture of labor and capital. Economic growth, then, will be a reflection of the growth rate in labor and the growth rate in capital. If today you have two workers with two shovels and they produce 20 holes per day, all you need to do to double your hole-per-day output is double the level of inputs: 4 workers with 4 shovels should be able to dig 40 holes per day.

The technology problem suggests itself right away: what if instead of doubling the number of inputs, you improved the quality of your capital. Say instead of shovels you gave them bulldozers. Then the productivity of both the capital and the labor you have will raise faster than the growth in inputs. A bulldozer may cost 1000 times as much as a shovel, but if it permits one worker to dig 2000 as many holes as he could with a shovel, the introduction of the new bulldozer techhnique will increase both the productivity of capital and the productivity of labor. Technology raises what Solow called Total Factor Productivity - the overall productivity of the factors involved.

The resulting equation is a Cobb-Douglas production function, where

O=A*(K^b + L^1-b)

Where O is output, A is the level of technology, K is the stock of capital, L the stock of labor, and b is a number between 0 and 1 that determines the percentage of remuneration of capital (whereas 1-b is the remuneration to labor.)

So already in Solow's original formulation, technology had this semi-magical property. In this formulation, A is exogenous to the model: it stands outside the theory, and is assumed to merely grow continuously and diffuse costlessly throughout the economy.

Now, New Growth Theory is mostly about trying to include this process of technological change and difussion into the theory, to endogenize it. Various routes have been proposed - the most usual rely on using the concept of Human Capital to explain changes in technology, others on postulating "spillover effects" where knowledge produced by one company seeps into the rest of the economy. Romer criticizes both of these approaches, though he doesn't discard them.

For Romer, Human Capital cannot be a complete stand-in for technological growth because unlike knowledge, human capital is a rival good. It is costly to train people to do things. Knowledge is non-rival. Once the design for a bulldozer has been created, it can be replicated as many times as you want at the same time. But if you teach me to use a bulldozer, I can only use that one bulldozer at the specific place where I'm working - I cannot be everywhere at the same time, the way a bulldozer design can. Romer suggests that the Human Capital fix is merely a convenient way to run-around the Innovation Twilight Zone, to re-cast the strange properties of innovation into a conceptual model that's easier to deal with because it's been reduced to a kind of capital that can be expected to behave like other standard private goods (like normal capital.)

Romer then takes a detour into the two characteristics that define an economic goods. In standard economic theory, goods can be private or public according to two criteria: rivalry and excludability. Clean air, the typical textbook example of a public good, is non-rival (because you can breathe as much of it as you want without undermining my ability to breathe), and non-excludable (because you can't stop me from breathing clean air, and you can't profit from it.) A banana, by contrast, is your typical private good. But what about knowledge?

Romer follows Arrow in characterizing the weirdness of knowledge. Like a public good, knowledge is non-rival - many people can have it at once without reducing the total stock available for others - but unlike a public good, knowledge is only sort-of, kind-of non-excludable. Plenty of mechanisms exist - from patents to TRIPs to trade secret laws - to help companies exclude others from using the knowledge they have developed.

So Romer's intriguing proposal is to break down the "A" in Solow's original formulation into two bits "AE" (for the part of knowledge that is excludable and "AN" for the bit of knowledge that is non-excludable. The non-rivalry of knowledge, together with the AN term, suggest pervasive positive externalities to knowledge-production, and suggest why innovative activity can generate society-wide rents that are much higher than the economic effort made in generating them.

Romer also criticizes Solow for simply assuming constant returns to scale (since b + (1-b) = 1 by definition) - an assumption that's wildly out of step with the real world. In many industries, and in especially in relatively technologically advanced industries, it pays to be big. There's a reason you don't see mom-and-pop car manufacturers or banks...in these industries b+(1-b) equals way more than one - they show increasing returns to scale.

Romer notes something that's been well-understood by generations of economists: in the presence of increasing returns to scale, a few companies will grow real big and drive out all the others, and at the end you'll have either oligopoly or monopoly, but certainly not a perfectly competitive market in any kind of Marshallian or Pareto-optimal sense. The pervasiveness of increasing returns situations in modern industries puts at risk, one would think, the entire edifice of general equilibrium. Romer goes as far as to -gasp- cite Schumpeter admiringly. At this point I'm thinking, shit, he's come around...he's an evolutionary economist.

Alas, no. Romer accepts that increasing returns and the incomplete excludibility of knowledge puts a component of monopolistic rent-seeking at the center of capitalist development. But instead of extrapolating from this insight into an overall system where firms try to outdo one another by innovating better than their rivals - like Schump does - he proposes a model where all capital goods are differentiated and each is produced by a separate, permanent monopolist that NEVER EXITS but instead extracts monopoly rents from their designs in perpetuity. These perpetual monopolists limit themselves to producing new designs and selling them through arms length transactions to "normal" neoclassical firms that remain price-takers! Though Romer understands that this is an analytical simplification (since firms generally internalize R&D efforts instead of buying their technology from third firms) - he seems to think this is an appropriate way of conceptualizing the co-existence of market dynamics and monopolistic behavior in the economy.

As a result, the representative firm is preserved as a concept outside the monopolistic R&D sector of the economy, and therefore general equilibrium can remain. Moreover, the R&D producing firms do not compete with one another in any meaningful sense since each produces just one monopolistic innovation and then exploits its quasi-rents in perpetuity.

In Venezuela we'd say Romer killed the tiger and got scared of its hide. By ghettoizing the R&D sector and keeping it well away from the rest of the economy, Romer misses the entire point of Schumpeterian analysis - that oligopolistic markets are pervasive, price-taking the exception, and the representative firm a myth.

Romer's model remains fascinating, though. It's the most comprehensive attempt I've seen yet to include the weird aspects of knowledge into an overall theory of general equilibrium growth. I'll have to read the paper 3 more times to quite understand it.

Why it is so hard for people as bright as Romer to understand the way General Equilibrium is incompatible with competition as businesspeople know it I don't quite get. General Equilibrium seems to be such an ingrained concept that neoclassically trained economists can't conceive of discarding it as a concept- it's as though they only think it's proper economics if there's a general equilibrium formulation at its center.

Still, Romer faces up to the fact that innovation is produced by profit-seeking companies that do not act as price-takers. This is a key insight - and from there to the Schumpeterian plunge there's but a short step. It's not a plunge many seem willing to take.

Saturday, March 13, 2004

Letter to Jason 

Dear Jason,

The debate on capitalism is a long and involved one. My question back to you would be which capitalism? US capitalism is NOT like German capitalism which is NOT like Japanese capitalism which is NOT like Singaporean capitalism which is NOT like Danish capitalism. Capitalism is not any one thing, it's a menu of options. Me? I want to see Danish-style capitalism in Venezuela. Lots of people will say I'm crazy to even think that's possible, but I think it is.

Especially for those who care first about the poor, the case for capitalism is way too strong to ignore in good faith. Only one economic arrangement has ever shown that it can bring the vast bulk of a large society from mass scarcity to mass comfort, an that's capitalism. There were over 50 famines in just 100 years in Florence in the 18th century, if you believe Braudel. Mass scarcity of basic items like food has been the common scourge of humanity for literally thousands of years, since agriculture was first invented. Only capitalism has shown, again and again, that it can take the bulk (in Northern Europe, virtually all) of society wih it from generalized hardship to a real security for just about everybody. No other system has ever proven itself able to do so.

Capitalism has shown itself as a devastatingly powerful ally against mass poverty not only in Europe, Japan and North America but also in large if localized parts of the poor world. Done in a certain way, capitalist development has taken a country like South Korea go from the current-day situation of Haiti to the current-day situation of Spain or Greece within a single generation. In other parts of the world, the same change is taking place. Largely ignored by capitalism's critics due to its nasty, awful government (which I don't endorse at all) China has shown that, if well managed, a process of rapid capitalist development can take literally hundreds of millions of Chinese peasants (campesinos) from living on the edge of starvation to living lives that start to resemble the life you take for granted in the US, where you may still be poor, but you can see your children's generation will have a better life than you because of how fast the economy is growing, and how much opportunity that capitalist growth brings them - opportunities they would not have under any other system.

Now, I think there are other debates that are far more important and interestingly than a broad philosophical debate about capitalism as such, and I'm pretty sure you would be bored with the long battery of statistics showing the plainly evident: that as economies grow, large sections of the population go from being desperately poor to being more or less economically secure. The impact of growth on real earnings over time is not really, I think, a contentious issue in economics because the studies of the link are so clear and unequivocal about it.

Now, if you're going to offer me an alternative system that you have reason to believe could work without using markets to allocate resources I'd like to see it - and it better not be old-style Marxism, which is so discredited even the Communist Party of Vietnam, the people who used to shoot at John Kerry in the jungle, has more or less chucked it out the window as a relic and are now trying to emulate the evidently more effective alternative Korea set out. If you have an alternative proposal, if you know of any that have a minimum of scientific plausibility - lets see them!

Otherwise, lets move on from the "capitalism or not" question, which is actually the wrong question, not to mention a much less interesting question.

The more interesting question is which capitalism, the details of the arrangement. Because those "details" are the difference between the particularly brutal and socially disruptive form of capitalism you have in the US, and for instance the Japanese model where companies continue to fight job cutbacks even in a 10 year depression because that is simply not done, if you're an honorable Japanese manager, except as the very last resort. Both are capitalist, but in each country capitalism comes cloaked in a set of societal values that determine what is possible and desireable for businesspeople or government officials to do and not do. Some places, like Denmark, seem to have found a system that satisfied everyone: a fully free economy with government support of companies and open trading links with Europe and the world but, at the same time, very high marginal taxes - in some cases amounting to a virtual "maximum wage" - and aggressive state redistribution of the proceeds to ensure a maximum of social cohesion and minimize income inequality. The result is a rich European economy full of world-class mid-sized firms in biotechnology and agriculture as well as furniture and industrial design fused with a quasi-socialist state that goes out of its way to help business at the same time it taxes them up the wazoo. A society with virtually no poverty, very little violence, high educational standards and economic security for everybody. It's not a theory, it works if you go to Denmark. Holland, to a certain extent, is quite similar, I should add (though not as aggressively redistributionist.)

But this all started with a debate about Venezuela, so I want to be quite clear about what I'm saying. (I wrote a background essay about this you might like to read.) If Hugo Chavez was proposing any sort of plausible, systematically thought out, pathbreaking new solution to the age-old problem of the state's relationship to the economy, I might give him the benefit of the doubt, or at least listen to him. If Chavez was a Marxist, like Allende, and had a clear plan like Allende to switch control of the economy entirely into state hands and institute a soviet-style dictatorship, well, I might agree or disagree with that plan. I can debate it. There is a theory there, an idea about how it is the New Society is to function. For Allende, it was simple, the state would simply seize control of all productive assets and operate them according to a central plan. Well, at least he had a well-developed vision of where he was going, like it or hate it.

But as Manuel Caballero always says, Chavez is not a marxist, he's not a stalinist, he's not a fascist, he's not a bolivarian, he's not a castrista, Chavez is a chavista.

President Chavez plainly could not understand this essay if he tried to read it - he has simply no grasp of any of these questions or concerns. He's an army man, a military man, his understanding of reality is simple: command and control. Every institution he can't control, every institution that won't bend to his will he destroys. As for the economy, he doesn't even pretend to understand anymore, which is nice: his explanations of economic affairs during some Cadenas in 2001-2002 were truly abominable - like a seventh grader trying to fake his way through a pop quiz he hasn't prepared for.

What I'm saying is that there is no such thing as a chavista economic doctrine for me to critique. I would love to engage such an animal if any such thing showed itself. But chavez doesn't need a theory to govern. He has power, and for him, that is enough.

Chavez now personally controls the bulk of state funds and his chief aim is to dispense the money through populist goodies and handouts to try to stay in power. Any question of the economic consequences of such behavior is pushed to the far bottom of the pile. Most of the time, critical reporting on the government's many, many, many economic blunders has simply been dismissed by officials. With the National Comptroller's (our General Accounting Office) office in the hand of a chavista hardliner who never investigates anything, corruption propagates in this environment virtually unchecked. None of the government's new social programs are auditable, which makes little difference, since they weren't going to be audited anyway. Where the money comes from? where it goes? only Chavez and a few palace and Finance Ministry insiders really knows.

The scale of the shambles the economy has become is so vast, I think it would be hard for someone from the first world to understand. I tell my Dutch friends here about it and they can barely believe it. In five years, the country's lost 60% of its manufacturing firms - kaput, bankrupt, gone. Venezuela has shed hundreds of thousands of jobs, the informal economy has for the first time ever grown to employ more people than the proper, legal, tax-paying economy and 90% of those in the informal sector earn less than the minimum wage, which is now below $80/month, parallel rate. With inflation running at 25-30% and wages raising much more slowly - especially in the informal sector - most people have seen large cuts in their already often very low wages. 4 million Venezuelans go to bed without dinner every night, and a three-meal day is beyond the reach of at least half of the population. In a country like Venezuela, where wages are already low, any outside shock really ends up having an effect on how much people eat. A sobering thought considering the 18% fall in GDP Venezuela has experienced over the last two years.

In short, it's just false that Chavez is helping the poor, it's another regime propaganda lie. For the vast majority of Venezuela's poor people, the last five years under Chavez have been an economic calamity. People have to work more and more hours for less and less purchasing power. Things like going to the movies or meals at McDonald's come to seem like extravagant, unaffordable luxuries. The latest statistics show that even coffee consumption has fallen sharply, and you know a Venezuelan is *really* hard up when he stops drinking coffee. It's evident to me that the set of ideological dinosaurs and bumbling youngsters who've been running the economy all this time (from Giordani to Nobrega) bear primary responsibility for this.

The problem is not that they're anti-capitalism, Jason, the problem is that they're not pro-anything beyond very short-term crisis management. Until now, they've simply not produced any kind of plausible vision of an alternative to capitalism - they've been too busy running the country into the ground to bother with trifles like that.

So Jason, please, don't believe the hype.


Thursday, March 11, 2004

What Schumpeter doesn't say... 

...is that static equilibrium is meaningless, it's not. The insight of static equilibrium does hold some very important lessons. For instance, it's easy to see that in those markets that most closely ressemble truly competitive markets (agricultural commodities, for instance, or entry level electronics like floppy disks) then something like static equilibrium does seem to take over. With marginal production costs at just a few cents per unit on technology available to many, many producers, profits will always be vanishingly slim. That's the reality of static equilibrium.

These markets are not attractive to the big companies of global capitalism. The large players seek to derive advantage overwhelmingly by specialized, specialist knowledge. Whether you're Hitachi or Vodaphone or Halliburton, what you're really selling is specialist know-how only you have. Through innovation, you've created a technical assymetry in given markets where you can perform a service or build a product better than anyone else's, and that market asymetry by definition leads to a loss of static welfare. It's by shocking the system through innovation and by profiting from the shocks that global capitalism can and does grow.

Wednesday, March 10, 2004

A typical World Bank extract 

I'm reading Rethinking the East Asian Miracle, edited by Stiglitz and Shahid Yusuf - World Bank big beasts. The book is a re-think of the bank's controversial 1993 oeuvre, "The East Asian Miracle", which sought to argue that the East-Asians were really neoliberals in desguise, that they succeeded by "getting prices right" and getting out of the way of the market. "The East Asian Miracle" served as a manifesto in imposing neoliberal policies on people from Nairobi to Hanoi to Barquisimeto.

The 1997 crash left the world bank with a bit of egg on its face, but when strong growth resumed in 1999, they felt a need for a critical reappraisal of the entire story. The book tries to explain what happened, but its conclusions are fairly muddled.

Missing in action from the analysis is any kind of settled microeconomic understanding of the role of learning interactions in shaping the dynamic effects of trade on development. To show you what I mean, I'll just quote a typical World Bank extract on the relationship between trade and development:

The starting point for understanding the link between trade and growth is the realization that trade can have both static and dynamic effects. Traditional arguments about why countries gain from trade are typically static in nature (Ricardo.) If a country moves from autarky to trade, theory tells us that production and consumption will change in such a way as to raise overall gross national product. These gains are static in the sense that once a country has opened to trade, all of the benefit from trade will be obtained on liberalization. Although traditional theory provides strong arguments for reducing trade barriers, these are essentially seen as one-time gains. Once these gains have been achieved, this theory has little to tell us about future performance.

Other considerations point to dynamic effects that could operate through their impact on competition and profitability. However, it is not obvious whether these effects will be positive or negative. Opening an economy up to trade will increase competition, and this could affect innovation, but economists are divided on the relationship between innovation and competition. On the one hand, there are those like Hicks who believe that competition is good for innovation because monopoly leads to lethargy and a search for "the quiet life"; on the other hand, there are those like Schumpeter who believe that some degree of monopoly is required to stimulate innovation. In fact, it is likely that neither perfect competition nor monopoly is particularly conducive to innovation and that intermediat market structures that combine rents to innovation with competitive pressures will be more stimulative. (p. 381)

They then go through a series of studies showing contradictory results on the link between competition and innovation, before concluding:
Overall, theory is quite ambiguous on the dynamic effects of trade. There are reasons to expect that increased international competition could accelerate productivity growth, but also reasons to expect the reverse. "Sometimes", as the saying goes "a kick in the pants gets you going, and sometimes it just hurts." Given this ambiguity, it is perhaps not surprising that views on the likely impact of trade on growth remain widely divided. (Lawrence and Weinstein, p 384.)

The extract raises a number of concerns. In the first place, if neoclassical theory does not support the view that trade is beneficial to growth, why do WTO accords prevent all countries from using trade barriers when circumstances suggest competition would be harmful? If even neoclassical economists cannot agree on the dynamic effects of trade, why should the poorest people in the world pay the price?

But beyond that I cite that last paragraph - which is not literally speaking typical, insofar as only rarely do World Bank research paragraphs contain folksy sayings - because it shows the neoclassical muddle clearly. Lawrence and Weinstein take the epistemological system of neoliberal economics and use it to try to infer a relationship between trade and growth from that aparatus. They look at various possible hypotheses that throw contradictory results and throw up their hands, saying, in effect, "hell, we can't work this out at all."

The reason is that neoclassical thought is basically aggregative, and has no clear link with any given view of microeconomic behavior beyond broad platitudes of utility maximization. Various patches have been proposed, but the fact is that the logic of neoclassical economics flows from the aggregative downward, not the other way around. There's no unifying microeconomic principle to help sort out the moddle.

For a Schumpeterian, a passage like that is a complete hash. Neoschumpeterianism does have a set of microeconomic foundations, a very definite view about the ways firms actually behave in the real world, and how that micro behavior can be aggregated to produce predictions that fit the historical data at least as well as standard growth accounting models (Nelson and Winter 1974 being the founding study of neoschumpetarianism in this regard.)

By borrowing mathematical models from evolutionary biology, they set out to model the behaviors of firms as though they were engaged in a strategic game with the other firms in the sector. Facing competitive pressures, firms have a choice of how much they will spend on innovative activity, and what strategy they will pursue (passive, imitator, or innovator). Some firms will thrive in a given competitive environment, others will wither and fail. Some will do well in one period in badly in the next. Some will get lucky today and have an R&D dry-spell tomorrow.

The point is that evolutionary theory endogenizes technological competition from the begining, making companies' technological choices - together with an element of chance - as the central engines of change within the dynamic system.

To their credit - and encouragingly, for me - Lawrence and Weinstein do at least cite Nelson and Winter as part of the discussion, even if they clearly treat their heterodoxy with some cautious disdain:

There are paradigms that are different from those of traditional profit maximization in which managers may be stimulated to innovate when international competition threatens their rents (for example, Nelson and Winter 1974). This involves the existence of managers who satisfice rather than maximize and behave under conditions of what is sometimes termed bounded rationality. Basically, they do not innovate continuously; rather they only innovate when subject to an unusual stimulus. In this world, import competition could spur competition, while the greater profitability of exports could do the reverse. (p. 383)

Well, at least they got a mention, even if one feels the authors kind of missed the point of the evolutionary framework altogether. The point about evolutionary models is that, unlike neoclassical models, they are based on behavioral assumptions about firms that are actually believable. They are based on microeconomic foundations you could describe to a businessman without having him furrow his brow and look at you like you're a lunatic. The technological/competitive decisions at the center of evolutionary models are the types of decisions business leaders spend their time making - rather than the optimizing abstractions of the standard model.

(Perhaps for this reason, Innovation Economics is far more prevalent in business schools that in Economics' Departments. It's not surprising - unless you're in finance or banking, if you're a businessman neoclassical theory is about as much use to you as an ashtray on a motorcycle.)

A neoschumpeterian, therefore, can only shake his head at the World Bank extract above. They didn't get it at all - they start from production and proceed to try to infer something about knowledge interactions from that. But we know that that's backwards. You need to start from knowledge interactions and infer production dynamics from that. You'd get something like:

Different types of host country conditions in different kinds of industries affect the potential for learning interactions as a result of trade. In those markets and industries where conditions are conducive to rapid learning, international trade will contribute to pro-poor growth. But when technical gaps are too large, institutional structures are not present, human capital is too scarce or other structural factors undermine the potential for meaningful learning interactions, trade will inhibit the innovative capacity of host country firms. The resulting configuration, led by Multinational companies, may be more technically efficient, but if possibilities for knowledge spillovers to local industry are thwarted, they may do little for the innovative capacities of host countries and will contribute little beyond wages to the economy.

Bengt-Ake Lundvall, the other big-beast of neoschumpeterianism, once put it devastatingly clearly saying that the basic mistake in neoclassical models is that neoclassical theory sees the three fundamental economic processes as production, consumption and accumulation. For neoschumpeterians, the three fundamental economic processes are learning, remembering and forgeting. Products and transactions are simply the embodiments of given states of technical knowledge. They're epiphenomenal. Knowledge interactions operate at a much more fundamental level, and ultimately explain the given production, consumption and accumulation decisions made by firms and households - not, as neoclassical theory would have it, the other way around.

Sunday, March 07, 2004

Thomas Friedman regains a few marbles... 

now the guy writes like a schumpeterian...

The Secret of our Sauce
Bangalore, India-
Yamini Narayanan is an Indian-born 35-year-old with a Ph.D. in economics from the University of Oklahoma. After graduation, she worked for a U.S. computer company in Virginia and recently moved back to Bangalore with her husband to be closer to family. When I asked her how she felt about the outsourcing of jobs from her adopted country, America, to her native country, India, she responded with a revealing story:

"I just read about a guy in America who lost his job to India and he made a T-shirt that said, `I lost my job to India and all I got was this [lousy] T-shirt.' And he made all kinds of money." Only in America, she said, shaking her head, would someone figure out how to profit from his own unemployment. And that, she insisted, was the reason America need not fear outsourcing to India: America is so much more innovative a place than any other country.

There is a reason the "next big thing" almost always comes out of America, said Mrs. Narayanan. When she and her husband came back to live in Bangalore and enrolled their son in a good private school, he found himself totally stifled because of the emphasis on rote learning — rather than the independent thinking he was exposed to in his U.S. school. They had to take him out and look for another, more avant-garde private school. "America allows you to explore your mind," she said. The whole concept of outsourcing was actually invented in America, added her husband, Sean, because no one else figured it out.

The Narayanans are worth listening to at this time of rising insecurity over white-collar job losses to India. America is the greatest engine of innovation that has ever existed, and it can't be duplicated anytime soon, because it is the product of a multitude of factors: extreme freedom of thought, an emphasis on independent thinking, a steady immigration of new minds, a risk-taking culture with no stigma attached to trying and failing, a noncorrupt bureaucracy, and financial markets and a venture capital system that are unrivaled at taking new ideas and turning them into global products.

"You have this whole ecosystem [that constitutes] a unique crucible for innovation," said Nandan Nilekani, the C.E.O. of Infosys, India's I.B.M. "I was in Europe the other day and they were commiserating about the 400,000 [European] knowledge workers who have gone to live in the U.S. because of the innovative environment there. The whole process where people get an idea and put together a team, raise the capital, create a product and mainstream it — that can only be done in the U.S. It can't be done sitting in India. The Indian part of the equation [is to help] these innovative [U.S.] companies bring their products to the market quicker, cheaper and better, which increases the innovative cycle there. It is a complimentarity we need to enhance."

That is so right. As Robert Hof, a tech writer for Business Week, noted, U.S. tech workers "must keep creating leading edge technologies that make their companies more productive — especially innovations that spark entirely new markets." The same tech innovations that produced outsourcing, he noted, also produced eBay, Amazon.com, Google and thousands of new jobs along with them.

This is America's real edge. Sure Bangalore has a lot of engineering schools, but the local government is rife with corruption; half the city has no sidewalks; there are constant electricity blackouts; the rivers are choked with pollution; the public school system is dysfunctional; beggars dart in and out of the traffic, which is in constant gridlock; and the whole infrastructure is falling apart. The big high-tech firms here reside on beautiful, walled campuses, because they maintain their own water, electricity and communications systems. They thrive by defying their political-economic environment, not by emerging from it.

What would Indian techies give for just one day of America's rule of law; its dependable, regulated financial markets; its efficient, noncorrupt bureaucracy; and its best public schools and universities? They'd give a lot.

These institutions, which nurture innovation, are our real crown jewels that must be protected — not the 1 percent of jobs that might be outsourced. But it is precisely these crown jewels that can be squandered if we become lazy, or engage in mindless protectionism, or persist in radical tax cutting that can only erode the strength and quality of our government and educational institutions.

Our competitors know the secret of our sauce. But do we?   

Schumpeter's departure starts from... 

...his appraisal of what is needed to make a statically optimal world into something more life-like and realistic. You'll have guessed his answer is innovation, but what's interesting is how he treats innovation and why he thinks it makes not only change but growth both possible and inevitable.

What changes when you add innovation into the system?

The first thing to notice is that when an innovation is made, it does not magically and instantenously extend to all of the firms in the economy equally, as neoclassical models would have you believe. The notion that all firms of necessity optimize and therefore all tend to produce at the same technology level are some of the more counter-intuitive in standard growth models.

In reality, innovations normally develop within individual firms, and those firms go to great lengths to protect them through patents, trademarks, trade secret rules, TRIPs, and a whole host of other institutional mechanisms. These institutions allow - in fact, are designed to - allow individual firms to maintain temporary monopoly rights over the things they invent. The innovation economy is an economy of many firms each seeking technological advantage through intensive, on-going, well-funded research and development efforts, and then seeking to protect their right to make money off of those innovations monopolistically for as long as posible.

Notice what has happened here: you've upset the system of perfect efficiency by introducing a market imperfection. You've granted the right to temporarily make additional profits to a single company in the system by granting them a patent over something they've invented. Due to those temporary monopoly pricing rights, that company will be able to extract a portion of the utility that would normally correspond to the consumer - the consumer surplus - because that is what monopolies do.

Now, for the duration of the monopoly (now 20 years by WTO agreement worldwide) that market will not operate as an efficient market. It will have one dominant producer that can extract consumer surplus at will due to patent protection, and other firms who will be forced, if they want to stay in the market, to themselves innovate, either by copying the original innovator as much as they can without infringing their patents, or by developing wholly new techniques that will leave the original innovator in the mix.

Schumpeter's point is that innovation is pervasive. Tens of thousands of patents are issued in the US each year for every conceivable technology in every conceivable market. The temporary market imperfections caused by innovation and accentuated by intellectual property rights laws are a pervasive destabilizing force, keeping the economy from settling into any one equilibrium position where all producers have equal access to technology and there is no tendency to change.

The somewhat counterintuitive result is that capitalism works due to market imperfections not despite them. And that there is a central tension within the system between the allocative tendencies of the market towards steady equilibrium and the pervasive efforts of business to top one another by trying to beat each other out technologically.

I should add quickly that I am not referring just to so-called High-Tech industries here but to the whole of the economy. The technology that allows a fish farmer to produce more fish more efficiently is just as important here as the battles between Microsoft and Sun.

Notice, first off, that once innovation is thrown into the mix, we're faced with a much more life-like, much more intuitively plausible world of corporate behavior than we could conceive in a steady-state equilibrium world. Once innovation is introduced and a competitive dynamic is set off between companies, we see the kinds of corporate behavior we've come to expect from contemporary corporations: aggressive cost-cutting and marketing, an emphasis on new product development and brand image, the works. The Schumpeterian firm is a firm that lives and breathes, a firm that's easy to see as a stylized representation of a real-world firm than the squadrons of mathematical optimizers that neoclassical theory would have us believe runs the business world.

The only problem is that if it's a certain set of market failures that unleash the competitive dynamic that gives capitalism its recognizable face, then most of the theoretical basis for economic policy making are wrong. Correcting market failure comes to be seen as a wrong-headed and simplistic slogan. The problem becomes a much more intricate one of consciously fostering certain kinds of beneficial market failures, of intervening at certain times and in certain places to force the market to fail in order to allow a company to take advantage of a technical breakthrough and thereby set off a competitive dynamic within the sector. The blind adoration for the allocative powers of market mechanisms will not help developing country policy makers to adopt policies that can work in those circumstances.

Selective intervention, therefore, becomes a hot button item. From a neoclassical perspective, there's really no way to explain how a government intervention into a well-functioning market could imaginably improve the performance of that market. But that kind of policy, is precisely Japan and Korea adopted in the 50s and 60s, and look where it's gotten them in just a couple of generations.

The post on the Asian Tiger's as posterchildren for schumpeterian theorizing will have to await until I've read more, but my preliminary point is simply that if Schumpeter is right, then for sure much - but not all - of what the WTO and the IMF are doing is totally backwards.

Saturday, March 06, 2004

Schumpeter starts... 

...with a piercing critique of the neoclassical consensus. He does it by taking us into an imaginary world, a world we would not recognize as our world but that, he thinks, corresponds to the reality suggested by neoclassical theorizing.

He calls it the "circular flow" economy - a neoclassical utopia of steady state general equilibrium where all markets clear fully, all the time.

To Schumpeter, such a world would ressemble nothing more than that Bill Murray vehicle, Groundhog's Day. In general equilibrium, all consumers are fully rational and consistent, and so are all producers. Consumers face a demand function - a mathematically derived "consumption basket" that scrupulously maximizes their enjoyment - while producers face a well-defined supply function that dictates their optimal production strategy rigidly.

Because tastes are assumed to be consistent, the demand function doesn't change. Because innovation is not part of general equilibrium theory, the production functions facing producers also remain constant over time.

Nothing changes. Each days producers get up and make the same optimal quantity of their product as they made the day before using the same technology in the same way. Consumers, similarly, wake up every morning and buy the same mix of products at the same prices as the day before, and every day they optimize the utility they get given their spending power. There is nothing in the system to push it to change in any way, at any time. That's what general equilibrium means.

Now, Schumpeter does have a purpose in writing this somewhat arcane reductio ad absurdum. He wants to show, conclusively, how weird general equilibrium is as a goal, and how the postulates of neoclassical theory, if followed to their logical extreme, yield up a society very different from what is commonly refered to as capitalism. Steady state equilibrium theory has a certain mathematical parsimony and elegance that Schumpeter could not help but admire - he waxes dithyrambic on Leon Walras, who he regarded as first to have worked out the system of differential equations implicit in Adam Smith. But as a description of the actual world we live in, Schumpeter argued, it fails badly.

This page is powered by Blogger. Isn't yours?

Weblog Commenting by HaloScan.com