Radical claims about recalculation

Via Interfluidity, Noah Smith writes:

In a typical microeconomic model, the market clears, because price adjusts to balance supply and demand. In a PSST world, this does not happen. The pattern of specialization and trade will not always be disturbed by small changes in prices, because the global pattern itself represents a stable equilibrium (i.e., is “sustainable”). How many computers I buy and sell will depend not only on the price of computers, my desire for computers, and my cost of producing computers; it will depend on the prices, desirabilities, and costs of a bunch of other goods throughout the whole economy. The economy will be riddled with network externalities, and the resultant weakening of the price mechanism means that any market may or may not tend toward efficiency on any given time scale. In other words, in a PSST world, there is no invisible hand.
This opens the door for a hugely expanded role for government (or other large, centralized actors) in the macroeconomy. If global patterns matter as much as local prices, then an actor large enough to perceive and affect the overall pattern might be capable of nudging the economy out of a bad equilibrium and into a better one. Dani Rodrik has been saying this for a long time in connection with newly developing economies, but the same may be true in rich countries when faced with disruptive technological change or globalization.
Believe it or not, that’s not exactly my view.  The most serious network externalities problem is most likely underinvestment in new innovation and its supporting infrastructure.  Subsidies to basic research can yield very high returns, as they have done for the computer, the internet, and through NIH.  Otherwise, when it comes to recalculating the resource allocation on top of the basic scheme of knowledge, I am skeptical that the public sector will do a very good job, for both information and public choice reasons.  Private sector rigidities and rules of thumb generally will mean that readjustment is too slow, not that it will fly off the rails into hyperspace, finance being one notable exception.  Education and confidence building can speed readjustment, as can nominal gdp stabilization, but if anything the public sector is especially sluggish itself.  Traders rang the alarm bell on the subprime crisis, and its later and broader offshoots, well before the regulators did.  Has the EU been ahead of the curve on Greece?  I don’t think so.

Comments

Network externalities also extend across geographic and nation state boundaries.

A hiccup in China means a sick stomach in Germany.

What does PSST mean?

RTFA

NSTABBUII

Never saw the acrynom before but understood it immediately.

NSTIBBUII for pedants

"If global patterns matter as much as local prices, then an actor large enough to perceive and affect the overall pattern might be capable of nudging the economy out of a bad equilibrium and into a better one."

Mr. Smith needs to read his Hayek. No actor -- regardless of "size" -- can perceive the details of the system-as-a-whole, much less alternate possible configurations.

On a more obvious note, claiming that PSST means "there is no invisible hand" is a gross overstatement. Local markets still function they just do so imperfectly, which is all that has ever been claimed in the first place.

The state doesn't need to perceive all the details; it just needs to perceive 'enough'. e.g., regardless of any other details, a sustainable pattern with significant amounts of ZMP workers is not desirable, so if the economy appears to be ending up in this state, shocking it so that it moves out of it is advantageous.

And how do you do that? No one knows. Note the scope of that claim: not just "the state doesn't know", "no one knows".

The knowledge problem is insurmountable by any individual who is not both omniscient and omnipotent. State action cannot "shock" "the economy" into a superior pattern except by accident.

And of course, the way you know that nobody knows is ... is ...
is that you know more than anybody can know?

Fortunately, the state isn’t one person.
Furthermore – have you ever heard about inventions such as the internet, email or the telephone. These technologies would enable state employees to, in addition to their own knowledge, gather data and opinions from other people.

Patterns of Sustainable Specialization and Trade

There is a brief description in Noah Smith's article that is linked.

That sounds exactly right - the invisible hand doesn't always get it right, but neither does governement. So we need to be very thoughtful about where government steps in and where it does not.

In general this probably leads to libertarian findings in some areas and social democratic findings in others, and (unfortunately) probably in ways that correspond with the commentor's personal political views.

E.g. I think the government needs in to health care and environment and out of intoxicating substances and urban land use, but I'm an urban liberal yuppie so of course I think that.

"The economy will be riddled with network externalities, and the resultant weakening of the price mechanism means that any market may or may not tend toward efficiency on any given time scale. In other words, in a PSST world, there is no invisible hand."

First off, price!=cost. If you're assuming that production ought to occur anywhere where the price of production is favorable, then you've likely never produced anything before. Paying for inputs is only one cost of business. Network externalities should be subsumed within the total cost of doing business, they need not be thought of separately.

And, re: "there is no invisible hand"...a bit of an exaggeration, don't you think? Suppose that you found an exception: does that mean that we should set fire to the old rule? Magnitude matters.

"This opens the door for a hugely expanded role for government (or other large, centralized actors) in the macroeconomy"

How did we get from "solution exists" to "we can find the solution", or even "government can find the solution"?

"This opens the door for a hugely expanded role for government"

You made the model more complicated and this makes it EASIER for central planners? I kinda doubt it.

This is not pure Austrian calculation; it's Klingian re-calculation - Kling permits the process of markets extracting information to go a lot worse.

Unsurprisingly, the central planner can also become a lot worse, before it comes worse than Klingian recalculation. Austrians claim that the process of price discovery is a positive at every transaction, but PSST is completely silent on the nature of the resulting equilibrium.

The economy will be riddled with network externalities, and the resultant weakening of the price mechanism means that any market may or may not tend toward efficiency on any given time scale. In other words, in a PSST world, there is no invisible hand.

Isn't this fallacy of composition.? It seems to me that it doesn't necesarilly follow.

I would contend that _some_ "traders rang the alarm bell...", i.e. the hand has five digits, not always pointing in the same direction at the same time.

I am skeptical that the public sector will do a very good job, for both information and public choice reasons."

Fair enough. But how does this compare with the inability of markets to account for facts that fall through the cracks of "imperfect knowledge"?

"Education and confidence building can speed readjustment, as can nominal gdp stabilization, but if anything the public sector is especially sluggish itself."

Or, education and confidence can divert adjustment to the interests of preeminent actors (who may have plenty of market shares already, if they have the means to invest in expansionary market practices). This reminds me of the mantra I hear from socialists all of the time - that we need to educate the masses to support socialism. But when a narrow group determines the character of that education...

"Traders rang the alarm bell on the subprime crisis, and its later and broader offshoots, well before the regulators did."

There are many more traders than regulators - and many, if not all "whistleblowing" regulators were small time. The fact that a narrow clique has the privilege of calling themselves "regulators" accounts for this far more than a regulator/trader conflict, since plenty of pro-regulatory individuals - notably Yves Smith and some her counterparts at NakedCapitalism - did indeed "ring the alarm bell." They were ignored, of course.

"In other words, in a PSST world, there is no invisible hand." Yeah, and that is why we are always in massive coordination-failure depressions.

"New innovation"? Well, I suppose there is--was--"old innovation." Still, . . . .

"My Masonomics line is that “Markets fail. Use markets.”
--Arnold Kling

The fatal conceit rears it's head once again. Again, size matters not in the perception of knowledge. This part is disastrously wrong:

then an actor large enough to perceive and affect the overall pattern

.

One way to look it: "the invisible hand doesn’t always get it right, but neither does governement."

Another way: the invisible hand (the market) is not wrong, it's players that are and government is but one of them.

"Fair enough. But how does this compare with the inability of markets to account for facts that fall through the cracks of “imperfect knowledge”?"

it doesn't say anything about markets tackling imperfect knowledge- it demonstrates that Government face a hurdle- that is public choice, in addition to the imperfect knowledge hurdle. This sets the bar higher for justification of government intervention which means we default to market activity without comprehensive evidence against it.

it demonstrates that Government face a hurdle- that is public choice, in addition to the imperfect knowledge hurdle.

This is an unfair comparison. Public choice interferes with the government's attempt to serve the public interest, but private actors aren't even trying to serve the public interest, they're trying to serve their private interests, which only coincide with the public interest by coincidence or when deliberately aligned that way by policy. The private-sector counterpart of public-choice problems is misalignment-of-interest problems; *both* also face the imperfect knowledge hurdle.

The public choice problems of the PSST models would be largely ameliorated by the solutions offered to public choice in our current system:
- Pay Congressmen/Regulators bonuses based on real GDP performance over a 10-20 year horizon following their service. (We would need a more honest and independent GDP deflator in this instance)
- Fix campaign finance so only individual voters can contribute, not "faux persons" like Exxon and the MPAA.
- Replace the current voting system (which creates powerful Party dynasties) with something that allows more entry and competition, like Approval Voting or Range Voting.
- Allow easier recall/replace petitions.

Etc.

You can't fix the knowledge problem, but knowledge isn't the main problem. Numerous improvements to the status quo are obvious even to someone like me (an interested observer and student of economics, but not more). There's a lot of low-hanging fruit here. The problem is incentives - only the status quo contributes to campaign slush funds.

---

Hey, here's an idea to let "the future" vote. Let parents votes on behalf of their children until they are age 18. If parents can vote multiple times it will weigh against the "current interests" in voting.

To what extent is the financial industry effectively a private version of the called for government intervention that Smith suggests to address these short term market failures? It is more likely that the efficient price discovery, multitude of players, and collective business experience in the financial markets will be more effective at redeploying capital than any government.

Is this a possible explanation for the increasing (until recently) role of the financial industry in the world economy? Many investors essentially try to individually determine where additional capital might be better employed by the economy, in exchange for returns on those investments (that's not the only investment strategy employed, obviously, but it is common). If the value of that redeployment service was increasing, it would explain the increasing importance of the financial services industry.

Of course, another explanation (which might simultaneously be true) is that various government policies have encouraged the growth of the financial sector, in what seems like a reasonably good case study against the value of government intervention to improve markets.

It probably extends beyond basic research. What would have been the state of the aerospace industry without the postal subsidy, or the development of radar without defense? The former would have been much slower to develop, and the latter would probably have not developed at all.

But what else would have developed if that money was available to be spent elsewhere?
I'm reasonably confident that the desperate need to defeat the Nazis drove the development of radar rapidly, but the UK achieved its defence spending during WWII by borrowing like mad and cutting back on other forms of consumption and investment. Economies have grown and innovation has happened with very little government spending.

Likely nothing. Growth, yes, innovation, no. Private innovation occurs, but it is smaller, more marginal, and less swift. The focus is on market extension rather than creation. Money is not the finite limited resource that the naive conceive it to be. Money spent somewhere is not money that cannot be spent elsewhere.

Okay Mr Pendant. Money is not a finite resource. It's the resources that money represents that is finite. The UK government was spending finite resources on fighting the Nazis by things such as developing radthoar, if it had not been for the war those resources could have been spent elsewhere. Happy now?

As for the innovation thing, I don't know what you think was not market creation about the invention of the electricity grid, the railway, the telephone or the novel.

Oops, Mr Pedant, not pendant. :) And radar, not whatever I spelt.

"Growth, yes, innovation, no. Private innovation occurs, but it is smaller, more marginal, and less swift. The focus is on market extension rather than creation."

Yes. Consider the analogy of bacterial evolution.

Say there's a new challenge (an antibiotic, or a new food source when a population that depends on the old food stays at the edge of starvation.

Mutation is risky, often there is no positive payoff at all. So usually the result is a change in regulation. If the antibiotic inactivates some particular enzyme, the easy response is to make more of that enzyme so that enough of it will function. Or if there's an enzyme that destroys the antibiotic at low efficiency, make lots of it rather than find a better destroyer. Similarly if there are existing enzymes that metabolize the new food source inefficiently -- make more of them.

You can find big important mutations by killing off everything else. Kill all but one in a billion and you can find mutant enzymes the antibiotic does not affect, that typically don't ever work as well as the usual enzyme did without the antibiotic. There was an extreme case with streptomycin -- mutant bacteria that survived high doses of streptomycin, that incorporated streptomycin into an essential function and could not survive without it.

But the usual result is to do the same things in different proportions. That needs less costly innovation and generally works better.

It's the same for business innovation. You must first invest in research that might not produce anything. Pay to develop results into functional products. Create demand for new products when customers start out with no concept that the product is available or even desirable. And then beat off competitors. Far easier to adjust the existing product mix to respond to variation in existing demand.

Government can over-produce innovation, though it could also under-produce. And government can be a large single buyer for particular innovations, giving particular innovators a leg up. Much easier to attempt the difficult project of creating demand when there is already a large continuing customer.

Innovation is a gamble. Rarely, it pays off big. So it's hard to estimate how much innovation is worth doing. Rational actors will tend to under-produce it because they will not depend on rare jackpots or miracles. Government by serving as a sort of random factor can result in increased innovation, with highly variable result.

I doubt if having a large, single, irrational buyer for particular innovations is necessarily a good idea. It may well give particular innovators a leg up, but the taxes necessary to pay for the purchase, plus the associated transaction costs of the taxation, will be giving other innovators a bit of downwards pressure. The more the government does the interventions, the more the downward pressure on the people paying the taxes.

(And your analogy is unconvincing, entrepreneurial innovation amongst humans doesn't come at the same risk of death as genetic innovation amongst bacteria. This is not to say that your conclusion is incorrect, just that you haven't supported it sufficiently).

The bacterial analogy is not exact. For one thing, bacteria tend to have a higher mutation rate when they are on the edge of starvation and have fewer resources to spare for DNA error-prevention. But businesses in recessions tend to choose R&D as one of the first things to cut.

But in both cases, the costs of innovation are paid up front, while the potential rewards come much later. When a bacterial cell gets a mutation, there's a chance it will become less competitive -- but the cost of that falls logarithmicly with time as its descendants lose market share. The fewer of them are left, the less they have to lose. Meanwhile the business can pay partly a known upfront cost.

After an innovation reaches market, the payoff increases exponentially for awhile, and then levels off and falls exponentially.

If it knew all the facts, the bacterial cell could compute an estimate of the average payoff per mutation. For the harmful ones it pays a logarithmicly decreasing cost. For the good ones it gets a benefit that increases exponentially for a time. It could compute the best mean time between mutations. If on average it loses by mutation, it might then determine how much metabolic cost it should accept to minimise the mutation rate. But it doesn't have that information, nor the needed computing power. Bacterial populations can however evolve toward an optimal mutation rate provided the optimum stays the same for many selective cycles.

How do you know how much the payoff for R&D will be? You don't. How do you know when the payoff will come? You don't. How much is the average payoff for R&D? You don't know. But to the extent that you can extimate it, your estimate will be low because most of that payoff comes from rare jackpot events. The rarer they are the less information you have about how big they are and how rare they are.

It's mostly inevitable that rational managers will underestimate the payoff from innovation. They look for dependable results and try not to depend on rare jackpots.

Markets do not give you the right answer for this question, because rational actors do not know how to find a good answer.

You may be right that depending on a single irrational entity to increase innovation is not the best solution. It is demonstrated to be better than mere free markets, though. Do you have a suggestion for a better solution?

Oh, wait. I'm responding to Tracy.

Again, I'm not convinced. Bacteria risk death, entrepreneurs risk bankruptcy. Not the same negative outcome. And a rational manager can't calculate the odds and payoffs of not innovating. Just because buggy whips had steady groeing demand for the last 500 years doesn't mean that some other company isn't out there introducing the car.
I missed your proof that a government does better than the free market, can you please point where you showed that out for me?

Life is too short to argue until Tracy is convinced.

Still, the cost of mutation and the cost of R&D are both somewhat calculable. The average bad result of mutation is known, and there are no big outliers -- the worst that happens is that one cell dies, one of a group of identical clones that may number in the hundreds. If you do R&D that does not result in a product, an accountant can count the number of dollars spent. It's rare that R&D results in a product you bet the company on.

You can't measure the odds for innovating, and you can't measure the odds for not innovating. The best value -- for you -- different for each business -- depends on rare events. How is supply-and-demand going to optimise this? It can't.

I missed your proof that a government does better than the free market, can you please point where you showed that out for me?

To the extent that the amount of R&D is determined by people who actually try to maximise profits, they will rduce R&D expense because R&D cannot pay off in the short run and usually does not pay off in the long run. It pays off incredibly -- rarely -- but you cannot plan for that. If investors want to see R&D you can buy a lab and pay people to stand around in white lab coats. If governments pays you to do R&D, likewise.

D&E -- Development and Exploitation -- pays reasonably well. You put some doodads on your product and then you advertise the hell out of them. Research does not. Every dollar actually spent on research represents a market failure. Efficient markets tend to eliminate research.

Therefore, government which actually does result in actual research, which eliminates some of the barriers to research (by paying for products like for example nuclear bombs and nuclear reactors which no rational person would want anything to do with) (and paying for research that nobody in his right mind would expect to pay off in any reasonable time), can hardly help but do more research. Whether this is a good thing or not is open to question, but it cannot be disputed that it's true. Or I should say, it cannot be disputed unless you're somebody like Tracy.

This story strikes me as too clever by half. Of course there are network rigidities in the private sector. Of course government intervention can shake these rigidities and move towards a more efficient solution. The question that is unaddressed is how rigid the new government interventions themselves are.

To me, this is exactly the story of totalitarian success. A lot of intellectuals in the west became infatuated with Stalin, not because they were secret Communists, but because his success was so stunning. The revolutionary government was able to evaluate technology from the West, and implement them in a manner that was state of the art in 1915--far more efficiently than our inertia-ridden markets. This led to an incredibly remarkable turnaround in the state of the Russian economy. However, what the west was good at is a slow process of continual adaption--even if not optimal, it was superior to the Marxist rigidities.

Is this not exactly where China is today? The question is how well will they adapt to the inevitable failures of some programs? Some governments are far more flexible than others, some have better technocrats, but it is not clear that even the best are more flexible than markets in all situations, that government efficiency is anything more than a one-time intervention per sector, or that public interventions are immune from their own inefficiencies.

I like the insight of PSST, but Noah's policy implications seem a touch regressive.

What technology did Stalinist Russia manage to implement in a way that was state of the art, and far more efficiently done than in the West? The space program's results were after Stalin.
And from memory of my high school history classes (we did a module on the Russian and Chinese Communist Revolutions), wasn't Czarist Russia growing at respectable catch-up rates before WWI?
(I assume that your date of 1915 was a typo, as that was during the Czarist regime).

Tracy,
The technology implemented under Lenin and Stalin wasn't capital-T Technology so much as a forced transition from an agricultural to a industrialized society, which was still pretty revolutionary at the time (remember, this was less than a decade after the Model T). This industrial base mostly borrowed ideas from elsewhere and implemented them in a manner that was, at the outset, almost entirely up to date.

And no, 1915 wasn't a typo, just shorthand for a snapshot in time shortly before the revolution got underway.

I don't follow. If I remember my history lessons right, the Czarist regime also saw very rapid industrial output growth, and the cities were expanding.
A forced transition from an agricultural to an industrialised society may have been revolutionary, but I think that depends on whether you'd describe the Japanese or German industrialisations as "forced". I've never explicitly studied either time period, but I understand that while the relevant governments could prompt plenty of legitimate complaints from a liberal point of view, they weren't as oppressive as the 1920s Communists in Russia.

Is this so radical?

Micro activities form the basis of macro economics. To recognize that there could be a consequence of microeconomics network effects on macroeconomics is to state the obvious, particularly when most micro econn and strategic business courses teach extensively on how network economics can be used in business planning, marketing strategy, etc, such that a number of businesses are modeled purely on network economic models.

I think what is radical (that is, radically conservative) is that macro took so long to recognize this in macro analysis.

I am also a bit bemused by some who argue that the complexity of network industries, at the heart of a technological economy, somehow means that government would have less of a role, because the system is more complex.

Just the opposite if you understand network economics.

Sometimes the cacophany is interrupted by a big drummer who can get the parts to match, or synch with, a steady beat. Network industries are very responsive to signals around which they can coalesce and organize. To say, for example, that we will focus on energy efficiency, serves as an organizing force for the network industries--the component, assembly, and installation elements--to act on the basis of that governmental signal.

Otherwise, its all chicken and egg economics. Or, Alfonse, no you Gascone, go first.

This isn't the complaint I am reading here, and certainly isn't the one I would make, Bill- that the government can't make changes.

Space program. Efficiency standards for appliances. Cobol and standardized computer languages to meet DOD specs. Efficiency standards for autos.

Do you want any other examples of how industries organized or were built around government standards or policies. This is how network industries develop.

When you say "the government can’t make changes." You are confusing the standard setter with the actor--in my examples above, business was the actor, and the government was the standard setter or customer.

But, if it makes you happy, you can deny history.

Were any of the government programs you've just listed responsible for lifting the economy out of a "re-calculation" style recession? However positive these interventions may have been, keep in mind what the claim is.

Jesus Christ, Bill- you still can't fucking read.

And once again, rather than admit he made a mistake, Bill scurries away.

I don't follow your argument. Cobol - who uses that commerically, who isn't working for the DOD? From this site, the most popular language is C, followed by Java, C++, and PHP, all non-DOD languages.

Appliances and car manufacturing aren't, as far as I know, network industries. That the government can drive engineers in those areas to focus more on energy efficiency than other factors is plausible, but I'm confident that if the government wasn't setting energy efficiency standards the desire for profits would be sending innovation on in those areas of the economy. Component, assembly and installation elements developed long before the government started setting efficiency standards - these happened with the first development of cars, for example, or steam engines to support mining.

Space program - this is the program that got man to the moon, then we stopped going? Not very impressive. Or are you talking about commercial space activity, which focuses on launching satellites for things like communication signals and navigation systems? And if the second, on what basis do you conclude that government provided the organisation point?

"I am skeptical that the public sector will do a very good job, for both information and public choice reasons."

Understatement of the year?

Hey, how about we let the government fix the health care system, where they are already substantially involved. If they get that right, THEN maybe we can talk about letting them control the rest of the economy.

This shows which they last very much lengthier and thus saving you income which could otherwise are actually utilized to purchase new ones.556

This shows which they last very much lengthier and thus saving you income which could otherwise are actually utilized to purchase new ones.32560

Comments for this post are closed