Thursday, December 11, 2008

Suggestion #5: Avoid a crisis caused by "just-in-time"

Those who brought us the Financial Implosion have set the stage for another, worse calamity.  We must act to prevent a similar disaster in the Real Economy -- that could affect manufacturing, commerce and possibly survival -- exacerbated by a dangerous over-reliance on “just in time.”

To be clear, I am not talking about the plight of the U.S. auto-makers, who are near collapse because of inefficiency, bad management, and failures of imagination by both owners and workers. Nor do I refer to the painful-but-expected retrenchment and layoffs we’re seeing in many industries.  These are worrisome, but there is another problem looming on the horizon -- one that may hammer even competent enterprises, that imagine they are healthy enough to weather the storm.

I refer to a brittle weakness in our economy, courtesy of the same smartaleck caste of MBAs who brought us derivatives and hyper-leveraged finance.  A frailty that could, potentially, turn some short-term crisis into full-scale disaster -- and all because of a good theory that’s been taken way too far.


Another trick of “optimization”

For decades, we’ve been told -- by the same fellows who brought us “efficient finance” -- that manufacturing and commerce should be fine-tuned to squeeze every penny of profit, by trimming away all “fat.” Industries that hew close to the teachings of W. Edwards Deming and Toyota’s Taichi Ohno require that their suppliers deliver parts and raw materials at the precise moment when they are needed.  Under this principle, any reserves that are kept on-premises will only encourage sloppy management and incur unnecessary storage costs -- a calculation that has long been exacerbated by shortsighted tax policies that punish warehousing and inventory-keeping.

This approach, called “Just-In-Time,” is based upon the very same postulate that led the business-major types to bet our economic farm on arcane financial instruments, leading to catastrophic failure, in 2007 and 2008.  A wholly unjustified wager that the economy and its supporting systems will always remain stable and never experience disruption.

Of course, this was just another expression of the basic concept underlying America’s vaunted way-of-life -- thinking only for today, spending all we had, with the lowest savings rate in the industrial world, while assuming tomorrow will always be kind. The freighters will keep bustling into our ports while foreigners continue to buy our debts.  Trucks and trains will keep delivering everything where it needs to go, with perfect, computer-controlled precision.

Our ancestors’ age-old wisdom of putting a little aside for just-in-case robustness has been replaced by a delusion of just-in-time efficiency, based on a belief in perfectly reliable global interdependence.

But, in real life, animals - even efficient ones - carry fat reserves. Surprises and disruptions happen. And when they do, we worry less about tweaking widgets-per second, and more about survival.


For the record, and lest anyone misconstrue, I actually have a great deal of respect for W. Edwards Deming and the first- order effects that his teachings had on industrial practice, not only in 1950s Japan, where he was pivotal in rebuilding that shattered nation into a prosperous economic powerhouse, but everywhere that Just in Time and related notions had positive effects, introducing tight discipline into management. For example, when Toyota eliminated on-site parts overstocks, they exposed countless inefficiencies in production that would have been overlooked, if employees could just reach over and grab replacements from a big pile. I'll be the last to deny the benefits that spartan efficiency have brought to industry and commerce. Nor does my praise of the warehouse overlook the fact that inventories do incur an inherent cost.

Nevertheless, any good thing can be taken obsessively too far, especially when a conflict arises between corporate profit and the common good. It is the job of the state to ensure that robustness and resiliency are desiderata that are included in the mix. Rules and market forces can be tweaked that encourage a balance. Above all, Deming-style practices, when followed obsessively, constitute what's called "success dependent planning" - or fine tuning everything to be based upon an assumption that all eventualities have already been taken into account. And as Nobelist F. Hayek clearly demonstrated, that notion can be carried over a cliff into pure (and self-destructive) delusion.

NASA got into a lot of trouble trying to use success-dependent planning. And when it is done by armed forces (and today's U.S. military show many signs of falling into this trap), there is a sequence of events that will follow, as inevitable as sunset. Surprise, dismay and then defeat.

Let’s go back to the example of tax laws that punish warehousing and inventory keeping. These are nothing less than inducements for fragility. If such disincentives were removed, even reversed, good managers might feel more ready to stockpile a little. Perhaps even enough supplies to keep their enterprises going for a while, say, during an unexpected ice storm, or if terrorists toppled a few of our vulnerable, chokepoint bridges.  Or if any of ten thousand other things broke down, in our humming national and international grid.

Moreover -- and here’s a little bonus that’s highly relevant during a recession -- in the short term, just filling such expanded inventories could add to national demand, providing a little more economic stimulus at a critical time. By rewarding companies for investing in inventory, we might also foster a self-fulfilling expectation that good times will return. This, alone, might justify innovating ways to ease back from just in-time thinking.

Even more convincing is the robustness argument. A modicum of warehoused supplies could someday make a crucial difference. And this applies not only to industrial supplies, but to the food and other vital consumables that the people of a city might need, if they found themselves cut off for a while. The ancient Egyptians knew this. The puritans did. Anybody with a gram of wisdom does, especially if they accept that we still live in dangerous times.

Above all, this is the opposite of what the MBA caste wants us to do.  And, nowadays, what better sign could there be, that an idea is worth looking-at?

==========

And that completes my short list of unconventional ideas about the current economic crisis.  Next: foreign policy and national security.unconventional ideas about the current economic crisis.  Next: foreign policy and national security.

Continue to Suggestion #6: Repair the U.S. Civil Service...


24 comments:

Anonymous said...

An elaboration on that:

Fund the creation of a specific set of sustainable and survivable manufacturing complexes: Fertilizer plants, road and rail repair equipment and supplies, electrical grid equipment and repair supplies.

These wouldn't be cost efficient. They wouldn't be intended to supply the whole demand for these industries. But they'd be able to supply enough to get by in the case of hideous natural disaster, war, embargoes, etc.

The plants would be powered by little nuke plants, have large supplies of stores, and as they come on line be equipped with "fabs" to supply repair parts. They'd be designed to use recyclables and waste material (e.g., sewage and compost for the fertilizer plants; "bioplastic" and old tires and such for plastics.


prefecox: Prescription med, introduced in 2027, for controlling the sebaceous cysts resulting from latent Shenzen Pox.

Matt DeBlass said...

Having a stockpile of parts would mean that planned obsolescence is slightly less likely.

"hangsh"- what you do with your clothesh when you get home after having a few too many drinksh

Tony Fisk said...

JIT is fine as a technique.

I do think that a tax on warehousing is silly. I would think that the free market play between cost of storage vs cost of transportation would work out the optimum solution on a case by case basis (I remember seeing this posed as a problem as part of the economics unit of my *ahem* diploma. They even derived a bona fide equation for the relation! Then proceeded to solve it by trial and error rather than a bit of high school calculus!!
... accountants play with numbers without the benefit of mathematics!)

(I note the car bail-out has been defeated... I would have naively thought that the repubs would have been for and the dems against..What's the strategy going on here? Is it the political map colour near Detroit?)

ljessn: a new-age Scandinavian model of economic development.

Anonymous said...

Hi,
This is an area I have some experience in.
I agree that some "fat" is needed as reserves and that if the MBA's had eliminated the fat then we would be in trouble.
However I believe that we are nowhere near that position.
As far as implementing JIT the Japanese Auto companies lead the world
(the chip manufacturer's may be up there too)
The US auto companies are ahead of general industry - but they are a long way behind the Japanese.
The MBA culture is lousy at implementing JIT - somebody with an MBA in an office has no idea where the production guys can hide stuff.

The MBA guy does some posturing sets out a stock reduction, the stock still exists it is simply changed from "our stock" to "supplier stock" (and we pay extra for it) the MBA gets a bonus and moves on to his next job.
He claims major savings BUT nobody ever audits the claims - and if they did he has probably moved on to a more senior position.

We actually need more JIT not less.

Example
Toyota engine plant in Kentucky
In order for a locally produced part to be accepted it had to pass the "Torture Test"

The engine was run until hot then the oil and coolant was blown out and replaced with minus 40C oil and coolant.
Run the engine till hot and repeat - for 24 hours!

Parts that
Passed that test
Made with capable processes
Had had their processes examined looking for week spots

Then they could do JIT!


The "Planning for Success" culture is a bigger problem I have been sharply criticized for planning for failure - Liberal Arts people think that it is important to be positive at all times and have difficulty understanding that the metal doesn't care

Anonymous said...

Actually, one of the reasons that investors push companies towards JIT is transparency - it is easier for investors to work out what is going on in the supply chain from the accounts of the various companies if there is less lag due to inventory.

Also, companies do have fat. They just keep it in cash, usually. The companies which don't have cash are the ones that are going bust now that the credit markets have seized up.

dmon said...

Way OT but this is the collection of minds I'd like to hear an opinion from. I see that the hydrino squad (BlackLight Power) has a customer... can their generator possibly... actually... work?

My extremely limited knowledge of the subject matter, plus conversations with people who did grapple with Schrödinger's equation, leads me to agree with the belief that the hydrino notion is utter crackpot fantasy. But if the BLP generator works... energy must be coming from somewhere. Is it an exothermic reaction with the doping agent in the Raney nickel? I read a recent ramble that seems to assert that Mills is pulling energy out of another dimension via the Casimir effect.

All way out of my league, but I'm guessing this forum has some informed opinions.

What the...?

yamahaeleven said...

I'm with Duncan. JIT is a constant process, always looking for possible failure points. Excess inventory simply hides these points. Your extra inventory may keep other things going longer, but when the real problem is finally noticed, it may be much more disruptive to correct at that point.

To me, an over inventoried, fat system is much more dangerous than a lean and fit system. Firms like Toyota are far more able to recover from seemingly insurmountable disasters than any other system we know.

Anonymous said...

I don't see the connection between JIT inventory and investors using financial instruments they don't understand. Can you elaborate on this?

Anonymous said...

Unless we fear utter collapse, it's probably sufficient to have surplus of key things. Food. Oil, natural gas, fuel oil and gasoline. Electricity generation and alternative distribution paths. Key medicines.

An increased stockpile of fossil fuels would probably be a smart place to start. Given recent and likely near future pricing thrashes I think people would support a small gas tax to build more storage and buy up oil and gas, to squelch oil speculation and reduce producer nation's leverage.

That could be extended to other "securing the national welfare" stockpiles with the same tax, once sufficient fuel has been stored up.

Obama could probably win over some of the neo-con base by using the story of Joseph and the Pharaoh.

Anonymous said...

Oh - and a stock of electrical and electronics, protected from EMP burst. And a similarly protected yearly backup of other property records and of our vast and poorly secured base of scientific knowledge.

Acacia H. said...

Small reserves and inventories are better than none. For instance, let us assume for a moment that a critical component for an automobile is found to be flawed. The entire batch that just came in is suspect and most of it proves defective. Without a small inventory of supplies, manufacturing stops until that critical component can be found elsewhere, and tested to ensure it works properly.

Smart businesses try to ensure multiple sources for supplies. However, this doesn't always work. The free market system encourages larger businesses to devour their smaller brethren and to force competitors out of business. Thus what might have been a supply chain with several sources ends up a supply chain from a monopolistic source. If that source says "screw you" or ends up with a shoddy product, what can you do?

With inventories of parts, you can continue building while either a new batch can be created or while a start-up subsidiary for that supplier, created by your own company, starts to provide a more reliable source of parts.

In addition, certain things should not be under a JIT system. Food stores, for instance, work better with an inventory. Drought, famine, diseases, these can destroy entire fields of crops and result in shortages. As a result, food prices climb, people go hungry, and crime increases (and sometimes rioting) when the have-nots turn to one of the oldest professions in order to survive.

JIT works better to prevent large inventories of finished products. However, smaller inventories of parts are healthy for unforeseen events.

Rob H.

Bruce Cordell said...

Boing's current "Dreamliner" issues are can lay a big dollop of blame on the fact they tried to leverage this just-in-time philosophy to the bleeding edge.

Ed said...

This is not a question of private profit vs the common good. As we're discovering, you can't make profits if your enterprise gets wiped out in a crisis.

The longer things hum along fine, the lower the perceived probability that people assign to a crisis. Which actually increases the probability of crisis, because you get increasing clusters of short-sighted decisions.

Low savings, high debt, high leverage, high P/E ratios, low dividend yields on equity, bigger houses, bigger cars, fatter bodies. All symptoms of the same cultural phenomenon: high time preference.

Tom Craver said...

Robert: "The free market system encourages larger businesses to devour their smaller brethren and to force competitors out of business."

I disagree - the free market encourages growth to the point where economies of scale hit the point of diminishing returns and competition makes further expansion non-cost-effective.

But corporate tax policy - deferring personal taxes on long term capital gains - means that investors want corporations to re-invest profits and make them grow further, to raise the stock price (which the investors don't have to pay taxes on until they sell).

So we get absurd battles over a few points of market share by engaging in hugely expensive advertising that - to the extent that the target audience don't simply ignore it - is largely countered by equally huge advertising expenditures by the competition that is also trying to grow.

Corporations buy out other companies in hopes that it will let them break into a new market, only to shut most of those down a few years later, writing off billion dollar losses. Or they buy up companies for their market share - but in the process destroy much of what gave that other company it's market share.

End corporate income taxes, and increase short term capital gains taxes to make up the short-fall. Corporations will then be driven to return more net profits to investors, and to limit excessive efforts at growth.

Fake_William_Shatner said...

Tony Fisk said...
JIT is fine as a technique.

I do think that a tax on warehousing is silly. I would think that the free market play between cost of storage vs cost of transportation would work out the optimum solution ...


I think I've read that the tax on inventories was to help workers in manufacturing plants. With large inventories, a business could max out their production for a short period of time, and then close the plant -- it's cheaper than keeping it open all the time. But this would mean that workers had no revenue and had boom and bust pay cycles.

JIT has run its course. Efficiency is already integral to our supply chain -- but there really are places where you want inventories. Manufacturing is more efficient, such that they scale better to the demand of production. With larger inventories, you could have a factory spend part of the year making wrenches, and then change over and set all the equipment to make hammers.


(I note the car bail-out has been defeated... I would have naively thought that the repubs would have been for and the dems against..What's the strategy going on here? Is it the political map colour near Detroit?)


>> It kind of comes down to Southern "right to work" (fire) States that have foreign automakers allied with the Republicans who can't stand Unions. Their constituency is 1% who make most of the money, and the other people they can make angry enough about some nonsense. The bailout of American Auto, comes out to $5,000 per worker. Alabama spent $250,000 per worker, getting Mercedes to build a plant in their state. So, this has ZERO to do with fiscal responsibility, or even a remote interest on what is in the best interest of the nation.

Anonymous said...

David wrote: "Perhaps even enough supplies to keep their enterprises going for a while, say, during an unexpected ice storm, ..."

When you wrote that, did you know about the ice storm in New England that has cut off power to 1.25 million? Some states don't expect power to be restored until after Monday.

Joshua O'Madadhain said...

Do you have any evidence (citations, etc.) that JIT--by which you appear to mean primarily the reducing stockpiles of physical items to zero or near-zero--has actually "exacerbated", or will exacerbate, our current economic crisis?

It seems uncontroversial that having zero _cash_ on hand when credit dries up--when your business depends on easy credit to function--is a big problem. But the argument that there are currently any problems due to hypothesized fragility in the supply pipeline seems less obviously true.

I'd also point out that in order to store inventory (which you recommend as an amelioration of sorts to the current crisis, as a means of temporarily increasing demand, at least at one level), you have to either spend money or put it on credit. People and companies are currently reluctant to do the former, and the latter has been made very difficult.

No offense, but it seems to me that what's going on here is that first, you find distasteful and counterproductive the "success-based planning" philosophy. I share this position.

Second, however, you appear to be asserting that JIT must inherently be an outgrowth of this philosophy, and therefore a bad idea.

I agree that assuming that everything will work should be a last-ditch strategy, not a basis for all planning. But JIT, like any aspect of a complex process, exists in a context. You can certainly implement JIT, and potentially even zero inventory, while still accounting for risks and planning on that basis.

By all means, let's come down hard on success-based planning, and in favor of plans which actually use utility theory (in which you consider not just probabilities, but the costs and benefits to various outcomes). But I think that JIT is an inappropriate target, and your argument is weakened by focusing on it.

Tony Fisk said...

David has stated that he considers JIT to be a good idea, but taken to extremes.

Interestingly, Goldratt's 'theory of constraints' comes to a similar conclusion on how to make production efficient, but via slightly different assumptions.

David points out that even the most efficient predator has reserves of fat.
It makes sense to compare practices with those that have evolved over a few billion years. You can see JIT in the natural world as well: predators only eat when hungry. Thinking about it, continual browsers probably make a better comparison. How fat do they usually get?

When I commented on a competition between cost of transport vs cost of storage, I was simply pointing out that, if transport gets chronically disrupted, eg through sabotages, it will become more expensive, and push things in favour of storage.

'menaties' - convenient units of labour

C. Keith Ray said...

Software developers have lately been trying to learn from "lean manufacturing", "toyota product development system" and "theory of constraints" (each of which has some "just-in-team" components).

The goal isn't to reduce inventory just to reduce inventory, the goal is to reduce the time from accepting an order to shipping a product, or increasing "throughput".

This just became more clear to me, in Jason Yips blog, where he says "Pull systems allow you to stop if the car doesn't sell." http://jchyip.blogspot.com/2008/12/japan-lean-study-mission-day-4-toyota.html

The joint GM and Toyota plant, NUMMI, doesn't build any car unless they have an order for it (some dealer has committed to buy it). I've taken the tour, and a single line [with 100's of small self-organized teams] is assembling half dozen different kinds of cars. They appear to be able to shift emphasis on a dime (though to build Priuses, they were constructing a new building a year ago when I took the tour.)

Anonymous said...

Someone asked:
I don't see the connection between JIT inventory and investors using financial instruments they don't understand. Can you elaborate on this?

Both generate non-linear and unreasonably fragile large systems. In both cases, tiny disruptions in apparently trivial parts of the system can lead to huge systemic breakdowns.

In the case of over-complex financial instruments, when the complexity skyrockets to the point when no one really knows how the derivatives interact, then when some apparently non-vital part of the market undergoes stress (such as subprime mortgages, which, we were assured over and over again ad nauseum only represent 4% of the total mortgage market and were therefore trivial), the entire system can blow apart. This occurs because of a non-linear chain of progressive failures in which feed-forward loops grow on themselves and exponentiate...all due to the excessive complexity of the system. If there were no such things as CDSs or CDOs, then when 4% of the mortgage market blew up and went into foreclosure, it wouldn't have been nearly as big a problem.

I think David Brin is drawing an analog to the increased hypercomplexity of the supply chains required by JIT. You get backups that can potentially exopnentiate the same way the absurdly complex financial instrumetns did. One vital component of a Ford electronic ignition, for example, might not arrive in time due to disruptions in global shipping caused by the recent problems with letters of credit. Lack of that component makes it impossible to build enough electronic ignitions, so that model of car has to have its assembly line shut down, but that assembly line proves crucial for the profit of a plant which is necessary to keep the Ford Georgia opration running...and so on. "For want of a nail, the shoe was lost..."

JIT also adds large amounts of computerized bureaucracy to an organization to track the damn parts. If there's anything we don't need in today's top-heavy over-managed executive-overburdened bloated senile corporations, it's more layers of management.

reason said...

I think the fragility of the financial system was mainly due to excessive leverage. Certainly using derivatives as a form of distributed insurance exerbated the problems, because the counterparty risk was not transparent.

MaysonicWrites said...

See http://en.wikipedia.org/wiki/1997_Aisin_fire
for an example of the "fragility" of the JIT system. How long would GM have taken to solve the same problem?

Anna G said...

You are right – “just in time” is the wrong principle for developing a global economy. We need to re-educated society so that “what’s good for the whole” is the main principle.

Here is a 5-minute video I’ve put together with others that promotes this active movement of individuals trying to bring about positive world change. If this video inspires you toward bringing about positive change, then please forward it to your friends. The state of the world is intensifying, and all we want is for people to spread the message that we need to wake up. http://www.youtube.com/watch?v=g94iWT2SIwI

MikeNelson said...

Another brilliant, essay, David. I particularly like the meme: "just-in-case robustness has been replaced by a delusion of just-in-time efficiency."

Another impact of "cutting to the bone" and "just-in-time efficiency" is the toll it takes on workers, especially knowledge workers. When everyone's working 110% all the time and there's no "slack," when an employee does manage to take a badly-needed vacation there is no back-up. That leads to the phenomenon of 24x7 availability and e-mailing from the beach--and that leads to pervasive burn-out and sleep deprivation. Next time you are sitting at an airport waiting for a plane or riding on a subway, look around and see how many people look well-rested. You see a lot more eyes with dark circles beneath them than eyes with a sparkle in them. Is it possible that some of the stupid decisions made on Wall Street, in Washington, and elsewhere may be due to sleep deprivation, burnout, and the inability to lean back and ponder the "big picture?"