Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
The Perils of Efficiency (newyorker.com)
16 points by danw on Nov 22, 2008 | hide | past | favorite | 53 comments


Interesting. I wonder if a bias or law against price gouging could be a problem, here. I mean, if you know there is, e.g. a 20% chance of a bad harvest that will let you sell food at twice the normal price, you probably have an incentive to store some yourself. But if you know that, in the event of a bad harvest, someone is just going to take your food -- or tell you that you can sell it, but only at an artificially low price -- you might not bother.

The usual question applies: can this complaint about the market be restated as a business plan? If so, why not do it? If not, is it really a solvable problem?


Insurance, perhaps? Treat price gouging as a natural disaster. Farmers pay for the peace of mind that they will be compensated in the event of a seizure.


I remember reading a similar article about the airline industry.

Despite great improvements in technology, fleet size and a greater diversity of airlines delays just get more frequent and more severe. Why? Because there is no longer any leeway in the system for error. Each flight is turned around and sent out in such minimal time that if it is late for any reason the problem cascades on to the next flight. And the whole system is so well coordinated that there is never any airplane just sitting around to make up for a problem. So the whole system, though more efficient is more fragile. And ironically, being more fragile, actually ends up making it less efficient.

Of course, as an engineer I simply say that you need to make the robustness part of your design goals to begin with and then design the system to meet that while being maximally efficient.


Because there is no longer any leeway in the system for error. Each flight is turned around and sent out in such minimal time that if it is late for any reason the problem cascades on to the next flight. And the whole system is so well coordinated that there is never any airplane just sitting around to make up for a problem.

American Airlines and all the airline companies wish that you were right. As is, the biggest problem with airlines is not the lack of robustness. As of now, most airline companies toss in generous leeways into their flights to accomodate late flights (now over 10 minutes, which does not seem like much until you realize that this is per every flight). The biggest problem is that the traffic control system is at capacity. Ie. It is too inefficient. To fix the delays, the FAA needs to modernize and increase efficiency in the airline traffic control system.


Good luck with that. It is technologically possible, but everyone who works with planes is scared shitless of the possibility that something can go wrong. It's bureaucracy at its worst. Nothing will happen with the existing system, the best we can hope for is that a new and better system will replace modern (obsolete) air travel.

Many major airports are clogged, however..it's more than just a problem of traffic control and coordination. There is a limit where you are operating so tightly that if anyting goes wrong, there will be a major accident. I think the direction we are going in is one of smaller aircraft and more decentralized terminals. With the level of stupidity in airline travel (et. al.) and the growing resentment among passengers, it is doubtful that the system can successfully develop into something even more centralized.

This whole industry is a prime target for disruptive change. The major problem is the same as for startups working in the financial industry: there is just too much regulation. (TipJoy, PayPal etc.) In addition, the capital requirements are pretty steep.


Flagged - you just can't post articles like this without the libertarians (or whoever is called out by the article du jour) frothing at the mouth and rendering the whole discussion nothing more than a repeat of a very old and tired internet discussion.


I think you're right, in hindsight I shouldn't have submitted this article


Perhaps this should be entitled "The Perils of Market Efficiency" instead, because this is more about market liberalization than improved productivity.

Although this article highlights some of the consequences of market liberalization, I think it's dangerously to extrapolate this case more broadly. First, structural readjustment programs took place in sub-Saharan countries that lacked necessary markets, transit infrastructure, and other prerequisites for private business growth. This differs from in the United States, where such infrastructure is substantially more mature. Second, many of these countries lack sufficient crop diversification, which when coupled with limited purchasing power, exacerbates famine during droughts.


I never understood what the reason was for the recent spike in food prices. Does anyone here know of a good study that explains this?


It was a combination of the spike in gasoline/diesel prices, a diversion of corn to make ethanol, which pushed up the price of everything that depended on corn (meat being a big one), and a shift to more corn production (which reduced the production of other things, increasing their price).

The US subsidizes domestic ethanol production and heavily taxes imported ethanol. Brazil produces ethanol far more efficiently than Iowa does. For those playing at home, that's a govt failure.


Ummm... does meat really depend on corn?


Yes. At least in the US, I believe, feed for a variety of meat animals is often composed heavily of corn. Ask your local rancher about this. Price of cow food goes up; price of tasty cow bits goes up.


Ahh, I hail from a grass fed country.


"... I never understood what the reason was for the recent spike in food prices ..."

Increases in population, the inability to feed it's own population and the drop in global rice production has caused increased rice prices in Philippines. The Philippines is now totally dependent on rice imports. Increased grain prices are certainly the result of complex combination of increased demand for wheat, the price of oil, introduction of competing bio-fuels ~ http://web.worldbank.org/WBSITE/EXTERNAL/COUNTRIES/AFRICAEXT...

"... Does anyone here know of a good study that explains this? ..."

The world bank has a good set of reports ~ http://econ.worldbank.org/WBSITE/EXTERNAL/EXTDEC/0,,print:Y~... there is also a wikipedia link ~ http://en.wikipedia.org/wiki/Food_crisis


I'm guessing it was inflation. There wasn't a food shortage, but everyone's money was becoming worthless. Now that the banks stopped lending, deflation is giving their money value again.


This is certainly a case where the loss is efficiency due to regulation is outweighed by the cost of a market failure. In all honesty, essentials like food, education, clean water, and so on should never be left up to the free market.


Must be nice to live in a world where govt fails less often than markets.


Actually, the government doesn't fail that much more than the market.

If you want to talk small scale, the government makes little mistakes every day. But the market fails every day too. Any transaction in which one party has incomplete information is a tiny market failure.

Large scale, neither fail that often. But when either does, it has drastic implications.

Honestly the idea that the market works the vast majority of the time is misguided.


Government and market are both human institutions. Market is really a convenient shorthand for "distributed, decentralized, uncoordinated economic transactions between consenting parties with an interest in the outcome", while government is shorthand for "centralized decision making made by disinterested economic actors". Please note that "disinterested" is a value-neutral term - it implies that the decision makers don't personally & directly benefit from the outcome.

Once you have the definitions right, you can see the benefits and pitfalls for both approaches. If you believe in maximizing human freedom, it is very clear that leaving people free to transact as they wish is the best path for all but those rare edge cases.


> government is shorthand for "centralized decision making made by disinterested economic actors". Please note that "disinterested" is a value-neutral term - it implies that the decision makers don't personally & directly benefit from the outcome.

Govt decision making is not done by disinterested actors. The interests may not be immediate or the benefits may be indirect, but they exist.


Agreed, the edge cases being the things that people absolute require: the essentials.


Why do you think that a govt decison maker is more likely to decide on the right mix of corn vs wheat (and where) than folks who directly benefit from making the right decision?


To ask this is missing the point; essentials are not about efficiency, they are about consistency. As I have said before, the article addresses this point. Did you read it?

"The problem is that, while this system is undeniably more efficient, it’s also much more fragile."


By what definition? And, by the way, why?


Because essentials always need to be available to people, even at the cost of efficiency. The article addresses why essentials are a special case.

"If the just-in-time system for producing cars runs into a hitch and the supply of cars shrinks for a while, people can easily adapt. When the same happens with food, people go hungry or even starve."


So "essentials" -- which, of course, are defined as things that are essential -- must be controlled by the horribly inefficient and incompetent government, instead of being provided by people with a vested interest in their provision. And this is true, because essentials "always need to be available". Yours is an argumentative wheel of Swiss cheese: completely circular, yet full of holes.


"If something goes bad with food, people could starve"

I was not aware of that fact. But in which way exactly does it call for government-controlled agriculture?


It doesn't call for government-controlled agriculture, and neither do I.

Regulation is far different than control.


Regulation is a form of control.


> Actually, the government doesn't fail that much more than the market.

If govt doesn't fail much more often than the market, why would one want the govt to be responsible for all essentials?

> Honestly the idea that the market works the vast majority of the time is misguided.

It's just better than govt.

I note that very few markets have set up death camps or waged significant wars. (The exceptions being folks dealing in illegal goods.) Markets don't try to throw me in jail for saying the wrong thing, markets not having jails and all.


> If govt doesn't fail much more oftenI note that very few markets have set up death camps or waged significant wars. than the market, why would one want the govt to be responsible for all essentials?

The article addresses this; because the government can be more consistent. Hence the need for them to be in charge of essentials.

> I note that very few markets have set up death camps or waged significant wars.

You're mixing apples and oranges here. How about a market that stops human rights violations?


It sounds like a great idea to let government regulate food. A very successful country did just that, the USSR. Where people stocked up on such nonessentials as butter when the factories finally could produce it.

Point being, arbitration and market forces will always ensure there are more resources available because the government is a monopoly with no challengers. In any markets with competition prices are lower, and more output is available. Besides, does anyone really think that governments can better judge and hedge the risk of various crop failures etc than the market?


> It sounds like a great idea to let government regulate food. A very successful country did just that, the USSR. Where people stocked up on such nonessentials as butter when the factories finally could produce it.

No one is proposing "regulation" on the level of the USSR.

> Does anyone really think that governments can better judge and hedge the risk of various crop failures etc than the market?

Does anyone really think the government can better judge and hedge the risk of mortgages than the market?


> No one is proposing "regulation" on the level of the USSR.

Why not? We're told that govts must be in charge of essentials because markets fail.

> Does anyone really think the government can better judge and hedge the risk of mortgages than the market?

The recent mortgage crisis pretty much proves that govts are worse at judging mortage risk than the market. After all, it was the subprimes, pushed by govt, that caused the problem.

And, it was govt that provided incentives to invest in Fannie and Freddie.

And, it was govt's backing that gave Fannie and Freddie the ability to borrow money that their assets couldn't pay back.


No, and the evidence backs this up. Guess who wanted the most to increase home ownership and pressed for subprime loans through stunts like the community re-investment act? Congress.

Guess whose legally protected rating agencies gave AA+ ratings to massive CDOs? Congress.

Guess who managed to figure out that subprime was going to blow and tried to guard against it? Goldman and assorted hedge funds.

Funny the double standards that we hold markets and government to.


Using Goldman as a representative of the market is insane--if it were the market wouldn't look anything like it does now.


Of course not. But thats still better than the government where nobody figured it out.


> The article addresses this; because the government can be more consistent.

Govts have no more inherent ability to be consistent than markets. And, if change is required, consistency isn't a good thing.

It's interesting that you keep comparing the platonic ideal of one alternative with the messy reality of another.

> How about a market that stops human rights violations?

Since markets create far fewer human rights violations than govts, markets are better here as well.


> Govts have no more inherent ability to be consistent than markets.

Not so. Being consistent requires a redundant system. A redundant system is not the most efficient one.

> Since markets create far fewer human rights violations than govts, markets are better here as well.

I will venture a guess and say that just as many human rights violations have been committed in the name of the free market than any government. Sweatshops anyone? But my point with that comment was not to argue semantics. It was to point out that you were arguing apples and oranges.


170 million murdered in genocides. Armenia, the Gulag, Auschwitz, the Great Leap Forward. Rwanda. Littler, still stark things like Varosha. World wars.

In the 20th century alone.

Can the market match that?


> Any transaction in which one party has incomplete information is a tiny market failure.

That's an absurd definition because no one has complete information.

However, if we're going to use it to criticize markets, it should be noted that it applies at least as much to govt actions.

It's an odd sort of comparison when one points to a problem with one alternative while ignoring the same problem with at least the same magnitude for the other alternative.


> That's an absurd definition because no one has complete information.

Not really. Markets depend on complete information. Else, how can someone accurately judge the value of something? If I buy a candy bar that's actually only worth $1.00 to me for $2.00, there is a market failure. I might think it was worth $2.00 because of the ingredients on the back. Little do I know that the candy bar is actually made with cheaper ingredients. There is a market failure. Now apply this analogy on a large scale and you can see how incomplete information causes market failures.


> > That's an absurd definition because no one has complete information. > Not really.

Feel free to cite anything about which you have complete information.

> Markets depend on complete information. Else, how can someone accurately judge the value of something?

There are no guarantees, so you act based on perceived value. When you're right, you're better off. When you're wrong, you're worse off.

Note that the lack of complete information applies to govt decison making as well, but govts lack diversity and apply force.

> If I buy a candy bar that's actually only worth $1.00 to me for $2.00, there is a market failure.

Huh? If that candy bar is only worth $1.00 to you, why did you pay $2?

Ah, because you think that ingredient cost matters. That's a failure of end-to-end thinking.

Note govts fail in analogous ways, and often with greater ill effect, yet you don't mention that possibility.


> Feel free to cite anything about which you have complete information.

You are proving my point valid; pure free market theory doesn't work because you never have complete information.


"pure free market theory" (whatever it is...) is just a model, and as such it can't perfectly describe what it is supposed to describe.


That's not absurd. It's basic economic theory: http://en.wikipedia.org/wiki/Market_failure


Yes, let's quote Wikipedia: "Thus, there is often a choice between imperfect outcomes, i.e. imperfect market outcomes and imperfect government outcomes."


> Large scale, neither fail that often. But when either does, it has drastic implications.

Large scale market failures are less common because there's almost always some diversity in a market.

Govts and regulation have no diversity. When the US govt's banking regulation fails, all regulated US banks are at risk.

That's why govts are especially prone to black swans.


The government can always stockpile food in case of shortages.


People can stockpile too.

However, they won't if they think that govt will seize their savings.


You mean, like speculators?


I first read this as "The Penis of Efficiency" and almost spit out my beer.


Time to reinstall Verdana.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: