2008-09-29

Banks and libertarianism

I would argue the following.

If banks have negative economic externalities, such as causing boom and bust cycles, and such as causing large-scale economic disruption when failing; then you cannot consistently believe in libertarian principles, while also being in favor of banks.

I have described previously how banks are inherently volatile and prone to collapse based on rumors alone - unless their depositors are insured with a system, such as the Federal Reserve, which can print money freely.

On the other hand, if banks and bank-like institutions are protected like this, then they require regulation to avoid moral hazard leading to exploitation of the money printing facility.

The stability of banks thus inherently goes hand in hand, not only with big government, but big government with the ability to print money freely.

I suppose that, to libertarians, this should be anathema. There is no greater power in the world than to have hundreds of millions rely on your currency, while you have the ability to print it freely. Meanwhile, the marriage of government regulation and banking that is required for banks to be stable, limits competition and encourages the development of a few behemoths who monopolize the banking market - and, by extension, play a vastly (too) powerful economic role. Perhaps leading to an even greater chance of systematic failure.

Libertarian policy calls for a small and effective government that operates based on a few robust principles. These might be:
  • Rule of law.
  • Respect for property.
  • Using coercion only to prevent coercion.
Note that these principles are all reactive. A libertarian small government does not manage the economy. Such a small government is not proactive. It is simply a guardian of people's rights. Such a government only acts when called upon for: when an individual's rights are threatened, or when a dispute needs to be decided upon.

For reasons of principle, a libertarian government cannot get involved in regulating the financial sector, for instance. That would be 'managing' the economy. Instead, the following two states of the world are consistent:
  1. Banks are permitted, but not backed by the government. They frequently fail, often merely based on rumor alone. Depositors lose money, and might therefore consider things more thoroughly next time. But no one else is harmed. Wiser people spread advice that banking is a form of gambling, and that the interest you receive on your account is largely offset by the likelihood that the bank will fail. Unwise people continue to make bank deposits, and periodically lose money, but negative externalities for uninvolved parties are small.
  2. Or, perhaps, it is found that banks do indeed cause substantial negative externalities by causing boom and bust cycles and by spreading a systemic risk of failure. In this case, the libertarian government must react to prevent negative externalities on the uninvolved, and therefore prohibits fractional reserve banking. All loans are thus required to be credited with money explicitly provided for the full duration of the loan.
So: either fractional reserve banking doesn't truly have negative externalities, such as boom and bust cycles and spreading of systemic risks; or else, a libertarian government should prohibit fractional reserve banking.

Is there any third option?

2008-09-27

Monetary supply: The role of borrowing in economic growth

Note: This article is concerned with a less important issue than my previous post. The question of monetary supply becomes simple if we first recognize the fundamental issue of whether, in the first place, we need banks.

In my previous article, Without banks: A proposal for a prosperous and stable economy, I discussed the role of banks in the money supply. I described how an economy with institutions that lend long and borrow short is inherently volatile and exposed to risk. I proposed that this volatility and risk could be avoided by phasing out the institutions that lend long and borrow short - eventually banning this business model altogether; requiring that all loans be covered by credit explicitly provided for the duration of the loan.

Towards the end of that article, I added a monetary proposal: shifting to a fixed, or predictable, volume of currency.

Central banks, with their ability to print limitless amount of money, currently exist for two purposes. One purpose is reasonable, if we are to have banks at all: central banks create money in times of crisis so that they can bail out depositors at failing banks. The other purpose is the subject of this article: creating money in normal times to 'stimulate' the economy. That is, to poke people with the stick of inflation, 'stimulating' them to deposit their money with banks.

I discussed how banks inflate the supply of money by a factor of about 10, and how the varying moods of the banking sector can lead to economic booms as well as busts. A 'boom' occurs when all the borrow-short/lend-long institutions lend freely, to the maximum amount they can. This increases the effective amount of money in the system vastly, and there's a lot of economic activity going on - much of which later turns out to be waste. A 'bust' occurs when it dawns on people that much of the economic activity during the 'boom' was waste. This means that lots of the loans that were made were bad loans that aren't going to be repaid. This threatens the lenders' capital reserves, which are already low because they've been lending as much as they could. So the only thing that they can do is decrease lending. I showed in my previous article how, if lending is decreased as a proportion of incoming deposits only by 10 percentage points, say from 90% to 80%, this eventually reduces the effective amount of money in the system by 50%.

When the effective amount of money in the system is decreasing, how are even those borrowers, whose loans might have seemed sensible before, going to repay their debt? Many of them aren't; there's insufficient money to go around. So now, as banks lend less, borrowers are increasingly going bankrupt, producing what is called the bust.

I argued that, in an economy with no lending long and borrowing short, and with a fixed supply of currency, there would be no boom and bust cycles, while sturdier bridges between prospective lenders and prospective borrowers would arise.

However, if the supply of currency was fixed, the role of borrowing in such an economy would be less.

The reason is that, in an economy where the supply of currency is constantly being inflated, everyone who holds currency has to lend it in order to just stand still. If you keep your money buried in a jar, its value is eroded over time because the amount of money in the economy is steadily increasing, compared to the amount of services and products. Whatever the economic growth, which would normally cause prices to slowly decrease due to increasing production - the money supply grows more. Holding bare currency is thus not an effective store of value.

In an economy where the monetary supply is fixed, the reverse is true. Holding bare currency does become an effective store of value. On the face of it, this sounds good. It may very well be. Banks, however, would have you believe that it isn't. The reason they would have you believe that is, in an economy where currency is a good store of value, their business would do much worse. If money is losing strength, people are forced to put their money into banks. It's common sense: put the money you don't need in a savings account, or else you're losing value.

Since banks are institutions that lend and borrow on a massive scale - a magnitude larger than even the amount of issued currency in the system - it is important for the health of their business that everyone believes that lending and borrowing is crucial to economic prosperity. Since everyone understands that lending and borrowing is not crucial for their day-to-day life, banks would have everyone believe that lending and borrowing is crucial for economic growth. The classic argument is, entrepreneurs need to be able to borrow money to create new products and services that we all need. But do they?

[Edit: removed an unnecessary argument that was probably defective.]

Does Intel need to borrow money in order to build a factory that will manufacture the next generation of faster processors? They do not. They have cash. Does Microsoft need to borrow money in order to build a new software product? They do not. They have cash.

I do not believe that all businesses can be bootstrapped. However, I do believe that investing is best when done by people who already have money - who have proved that they have the capacity to earn it and keep it; who do not need to borrow it from a bank. And the best way for such investors to be repaid is through an equity stake, i.e. by earning a proportion of future profits. If their investment succeeded, i.e. if it has increased production of a resource that is valuable to the economy, then their dividends will pay it back. This is not a zero-sum game, because everyone can be continuously investing in their businesses to improve production, and if all such investments are smart, then everyone will be paid back - not necessarily in a larger total amount of money, but with larger total purchasing ability, than if they chose not to invest.

I do not believe that bank-like loans are as crucial to economic growth as banks and other borrow-short/lend-long institutions would have us think. There is a case to be made for borrowing on a reasonable scale. However, the rampant lending on which our current economy is based, fuels boom and bust cycles, encourages a gambling culture of investment, and introduces a systemic risk.

In a fixed monetary supply economy, people would still borrow and lend, but more investment would be made by people investing directly into things they think are worth investing, acquiring equity stakes. It would be the reasonable way. It in fact already is; however, we now have short-borrowers/long-lenders who make it easy to 'invest' money with them. But it is not 'safe'; and it is not healthy for the well-being of the economy, because the people who might otherwise invest money directly into plans they think are worthy, instead pass on their money onto agents whose activities have more to do with gambling and where genuine production improvement is just one of the possible side effects.

And if it turns out that people in a fixed monetary supply economy do not invest or lend enough, and that growth suffers as a result - we can always cause the monetary supply to start to steadily increase, thus poking people to consider investing their savings. The economy might benefit if people used their minds and spent more time thinking about where to invest; rather than abdicate responsibility and do the default thing, putting their money into equivalents of banks.

2008-09-26

Without banks: A proposal for a prosperous and stable economy

After reading Nouriel Roubini's The Shadow Banking System is Unravelling, I wonder if there's any real place, in an economy that wants to be robust, for institutions that lend long and borrow short. The most obvious such institutions are deposit-taking banks, but as Roubini explains, many other organizations work like that, including broker-dealers, hedge funds, private equity groups, structured investment vehicles and conduits, money market funds and non-bank mortgage lenders.

The problem with institutions that borrow short and lend long is that they are inherently volatile. They subject the economy to wild rides, while spreading a systemic risk of failure.

In normal operation, even if the amount of actual currency issued by a central bank is held constant, these short borrowers - long lenders cause a varying supply of money. When you deposit $100 at the bank, you think and act as though you still have $100, but in fact, your bank has gone right out and loaned $90 or more to someone else. The other guy is supposed to pay this back after a while, some day, somehow, but in the mean time, he deposits those $90 into his own bank account, so the bank takes that and lends another $81 to someone else again. This is normally repeated endlessly, and the net effect is an increase in the amount of money in the economy by a factor of 10! This produces a wild ride when the economy is alternately flooded with money when banks feel good and lend freely, then starved again when they feel vulnerable and lend less.

For example: if banks lend out the maximum amount permitted, say 90%, the amount of money in the system reaches 10x the amount of issued currency: a series of 1 plus 0.9 plus 0.9^2 plus 0.9^3 and so on, equals 1/(1-0.9), which equals 10. On the other hand, if the banks tone down their lending just a little bit - to, say, 80% - then the amount of money in the economy decreases by half: 1 plus 0.8 plus 0.8^2 plus 0.8^3 into infinity, equals 1/(1-0.8), which equals 5. (Now weren't those high school infinite series useful?)

The worse problem yet is that all these institutions are subject to 'runs on the bank'. You know that at any given time your bank has only 10% or so of everyone's money available. You know that, if this percentage ever truly reaches zero, you will be unable to use the money in your account; the bank's vault will be empty then. So if you hear the word that many others are going to start to withdraw money, your best option is to run to the bank and beat them to it. You tell your friends, and now everybody is running as fast as they can to the bank. A bank can collapse on a rumour, even if it's business was entirely solid - if 'solid' is, of course, a word that can even be used for a bank.

The bankers saw that this is no good, so they banded together at some time around 1910, joined heads with government, and the Federal Reserve was born. The Federal Reserve is basically a facility that allows unlimited amounts of money to be invented out of thin air if circumstances require. If there's a bank run, no problem. As the bank's cash reaches zero, it goes to the Federal Reserve, borrows an awful lot of money, invented right then and there, and gives it to everyone who wants to withdraw. The bank doesn't fail, and soon enough, people bring their money back. The invented money is returned to the Federal Reserved with some interest, where it's destroyed, and everything is back to normal. Assuming the bank run was only due to a rumour.

Ah yes. But this creates a moral hazard. With money that can be created out of thin air, banks must now be regulated. If they are not, any schmuck can create a "bank" which "lends" to friends with no expectation of repayment. People hear word, everyone withdraws, the Federal Reserve provides the money, and someone just got rich at the expense of everyone else.

So deposit-taking banks are regulated. But broker-dealers, hedge funds, money markets and non-bank mortgage lenders, so far, were not. Which means they don't have access to the Federal Reserve, which means that they are subject to bank runs. And as Nouriel explains, many of these just occured. 51 hedge funds collapsed since 2006, compared to 14 in all years before.

Arguably, people who invest in businesses that lend long and borrow short, and which do not get Federal Reserve protection, and who thus expose themselves to bank runs that can be triggered by rumours alone - well, such people are stupid-ass investors, who in addition can afford to lose. You can't invest in vehicles like these unless you have at least $1 million. So, most people who lost money there, will not be homeless soon.

However, the systemic risk that these entities create - and more importantly, the economic wild ride that results from normal bank operation - leaves me considering that we might be better off without these entities at all.

What would a world look look like where borrowing short and lending long is banned?

First and foremost, banning such transactions would require all loans to be covered by credit explicitly provided for the duration of the loan. In other words: a 30-year mortgage would require someone to provide the money and not expect it back for 30 years. "Hogwash," some people will say. "This will not work! Whoever will loan you money for a 30-year duration?"

Ah... but there are loads of money praying to be invested for the long run. All our pension savings are such funds.

And by the way: in case most people cannot get a 30-year mortgage, what happens to the price of housing?

It goes down! It goes down until it reaches a point where the market clears. Did you notice - it's hard to notice unless you look at historic charts, but: have you noticed that real house prices in the U.S. doubled over about the same time as women entered the work force, now working side by side with men? It used to take one person's work to buy a place for the family to live. It now takes the income of two people, because twice the number of workers are now competing for the same number of houses.

So, the economy adapts. Pension funds are an obvious large source of credit perfect for long term lending commitments. Without banks - or any other institutions to lend long and borrow short - the volatility of the economy drops to near zero. Any remaining boom and bust cycles would become much less pronounced. The economy would hum along at a steady pace. Growth would still occur; businesses would still borrow to invest. Big businesses could sell bonds. Meanwhile, the void produced by banning banks would be filled by investors who know exactly where they put money. New and sturdier bridges between investors and borrowers would arise. Web sites would match interested lenders with loan seekers fitting all desired profiles.

Best of all, the role of the central bank would be reduced to trivial. No more bailing out financial institutions; there would be no organizations to lend long and borrow short. No more setting fractional reserves; there would be no fractional reserve lending. No more target interest rates; these would be determined entirely by the economy, which would comfortably hum along.

Instead, the entire role of the central bank would be replaced by a single task: maintain a fixed amount of issued currency. No more inflation; the amount of money relative to the number of products and services in the economy would never increase. No chance of catastrophic deflation either; all money would be out there, it couldn't simply disappear - as 50% of the money would today, if banks merely reduce lending from 90% to 80%.

The only thing that would happen to prices, over time, is that they would slowly fall. As technological progress makes it possible for services and products to be provided more efficiently, more things would gradually be available, while the same volume of currency remains. Prices could be expected to fall a few percent per year in the long run, while wages and salaries remain constant.

Having explained that, I hereby declare that I am running for world president. ;) If I was one, I would:
  • Phase out lending long and borrowing short, thereby removing boom and bust cycles, with an economy that goes on growing.
  • Introduce a world currency with an issuing authority dedicated solely to maintaining a predictable amount of issued currency, thereby ensuring robust prices and a stable growing economy, subject to no more volatility.
With these changes to the economy, we can then only fear a natural event.

Tragedy and roller coaster rides are not an inherent design component of a well-functioning economy. We would all benefit to remove the components of the economic system that cause major and, in the large scheme of things, unnecessary volatility and risk.

UPDATE: In the above post, I proposed a fixed money supply. It is possible, though I don't find it certain, that growth would benefit more with a predictable money supply - one that would be steadily increasing. I discuss this in my next post.

The Silver Rule

I recently realized, while reading up on Confucius, that all my life I've been led by the Silver Rule:

Adept Kung asked: "Is there any one word that could guide a person throughout life?"

The Master replied: "How about 'shu' [reciprocity]: never impose on others what you would not choose for yourself?"

[Analects XV.24, tr. David Hinton]

The Wikipedia article states that this is equal to the Golden Rule, just stated in negative form. I find this not to be the case.

I rejected the Golden Rule when I was little. My father mentioned it to me several times in an attempt to teach me about life. I just found it dumb, but in a confusing way. I didn't know exactly why.

Here's what the Golden Rule says:

Matthew 7:12: "So in everything, do to others what you would have them do to you [...]"

Luke 6:31: "Just as you want others to do for you, do the same for them."

Note that this form of the statement arises primarily out of the Book of Suffering the Bible. Ancient Greek philosophers all use the Silver rule again.

How are these rules fundamentally different?

Here's how.


The Golden Rule compels you to do things which you think are good, even if you're wrong about them.

The Golden Rule further does not mention neutral things, which can lead one to believe that one must focus on "doing good" - no matter how misled one is about what's good.

The Silver Rule, in contrast, compels you not to do things which are bad, even if you're wrong about them being bad. The Silver Rule leaves you free to do things which are either neutral or good. Whatever you prefer; just try not to do damage.

The crucial difference is that people following the Silver Rule do less damage. They do less damage because they are compelled to avoid doing damage, rather than driven to do good.

The problem with doing good is that the amount of good any single person can do is very limited compared to the amount of damage any one of us can cause. Only few people can contribute significantly to the scale of human achievement. But every single bumblehead can undermine the efforts of hundreds.

The potential downside is so much larger than the upside, it is much more important for human beings to avoid doing bad, than to try doing good.

In addition, following the Silver Rule leads to a happier life. If you merely avoid doing bad things, you are free to pursue all the other ones which are good or are neutral. On the other hand, if you are driven to do good things, you feel guilty when you're not doing them, even if you're hurting no one. So you're driven to do needless things you think are good, that may in fact be harmful.

So please: follow the Silver Rule. Dump the "Golden" Rule. It is not for nothing that it comes from the Book of Suffering.

2008-09-25

Qualifications of writers at The Economist

The Economist is pretty much the only magazine I find worth reading - all the other magazines I've picked up many times, from Business Week to Newsweek to TIME, are utter crap.

However, I have now realized why contributors to The Economist are never signed.

I used to naively think that it was because of some professional or integrity related principle. Perhaps that editors have as much input in any article as the authors do. Perhaps that the magazine is more important than the authors.

Nope.

The reason could have something to do with that The Economist attempts to be a serious, considered publication that has weight. Apparently, though, the individual authors don't have such weight.

Don't get me wrong. Megan McArdle seems like an intelligent, above-average and intellectually well-versed person with lots of opinions I agree with - except, perhaps, on what best to do with the hard-core homeless people. She wants to take good care of them.

But her bio basically states her formal qualifications as... English major, and an MBA from UC.

That might be one MBA more than the requirements for most media outlets - which explains why so much stuff in magazines is crap.

But if an English major and an MBA is all it takes to write for The Economist, then... heck; I should be editor. ;)

The cost of supporting the hard-core homeless

This is just outrageous.

According to this, people who use homeless shelters are of three kinds:
  • One or two nights only. These people get on their feet and never come back. About 80% of people.
  • People who come periodically and stay for several weeks at a time. Mostly druggies.
  • Hard-core homeless, keep coming back all the time. About 2,500 of them.
Well, it turns out that the cost of that humble cot those hard-core homeless sleep in every night is about $66 dollars per night. So each one of those 2,500 hard-core homeless causes New York City damage to the tune of $24,000 a year - just for the cot alone.

But it gets worse.
Boston Health Care for the Homeless Program, a leading service group for the homeless in Boston, recently tracked the medical expenses of a hundred and nineteen chronically homeless people. In the course of five years, thirty-three people died and seven more were sent to nursing homes, and the group still accounted for 18,834 emergency-room visits--at a minimum cost of a thousand dollars a visit.
That's 30 visits per chronically homeless person per year on average, for a total of $30,000 each in medical expenses. There's more:
The University of California, San Diego Medical Center followed fifteen chronically homeless inebriates and found that over eighteen months those fifteen people were treated at the hospital's emergency room four hundred and seventeen times, and ran up bills that averaged a hundred thousand dollars each.
Megan McArdle supports a government program that attempts to address the social burden that these people cause by giving them free apartments. The hope is that this will help them get on their feet and, I guess, not come to hospitals so much.

This will not work. An important reason why most people avoid seeing the inside of a homeless shelter is because they know that homelessness is tough. If being homeless ceases to be tough - and in particular, if apartments are being given away for free! - this will increase the number of homeless people, as homelessness starts to be seen not only not-so-tough, but even advantageous.

But as the article describes, the tough sort of homelessness that we have now is hugely expensive when coupled with free shelter and medical help that is currently being given to these people free of charge.

The true solution is: Stop treating people as though everyone is equally valuable!

These people are not equally valuable! Those who are chronically homeless can never be expected to contribute anything of use, and they generally haven't done anything useful in the past to qualify them for anyone's largesse, either.

Anyone who becomes homeless on accident, gets off the streets fast.

The only fact that separates the chronically homeless from stray dogs is the bare fact that they are people.

In the calculus of "get what you deserve", merely being born as homo sapiens is not enough. You have to show effort, will, potential, creativity. These people are showing none.

Bail them out of homeless shelters after a few weeks. Is it freezing in the winter? Tough. Invite the guy into your apartment if you like. Not inclined to? Then why should the taxpayers?

Above all, don't give these people medical assistance. If they're dying from inhaling their own vomit in a drunken stupor, let them die. They don't care. They've already shown that. You shouldn't care either. But if you really can't stop caring - why don't you pay for their bills?!

It makes more sense to house 1,000 kittens in a pet shelter than to waste resources on these people.

Too big to exist

Senator Bernie Sanders recently remarked this in his comments on the crisis:
This country can no longer afford companies that are too big to fail. If a company is so large that its failure would cause systemic harm to our economy, if it is too big to fail, then it is too big to exist. If it is too big to fail, it is too big to exist. We need, as a Congress, to assess which companies fall in this category. Bank of America is certainly one of them. Those companies need to be broken apart. We cannot have companies so huge that if they go under they take the world economy with them.
I most certainly agree. There is a tendency these days for companies to grow beyond any economically substantiable limit. Corporations grow for the entirely wrong reasons.

First and foremost, it is necessary to understand that corporations don't grow because this makes them economically more efficient. Hugeness does not equal efficiency. A huge corporation is, in effect, a small centrally commanded economy. Being centrally commanded economies, the structure of large corporations has more to do with Soviet-era communism than it has to do with capitalist efficiency. (There is a small but important exception: corporations can choose their employees.)

I believe the following are two of the fundamental reasons why corporations grow and grow.
  • Double taxation of dividends. Any dividends that corporations pay to their shareholders are double-taxed. First, the government takes some 20-40% (depending on which government) on the amount of corporate profits. Then, those profits are distributed as dividends to shareholders, and the government takes an additional 20-50% of that, depending on the incomes of shareholders. Effectively, dividends are taxed at something like 50%.

    It is not only that it's double taxation. It's more than that - by paying dividends, a corporation first has to recognize that portion of revenue as profit in the first place. If the revenue is not recognized as profit, it cannot be used for dividends, but the corporation can use "creative accounting" to avoid paying corporate income tax.

    The consequence of this tax system is that shareholders are encouraged not to take dividends, but instead, to reinvest the money into corporate growth. This avoids the corporate and income tax, while shareholders get their due through an increased value of their shares.

    Most importantly, because dividend taxation is so huge, money is invested into corporate growth, even when it would be better used elsewhere. The investment opportunities elsewhere would have to be more than 100% better for dividend payments to make sense.

    This distortion should cause excess organic corporate growth.

  • Conflicts of interest between shareholders and corporate leadership. Directors and CEOs do not own the corporations they run; but they sure do run them. The bigger the company you run, the more power you wield, the more money you make, and the greater you are. Even when it isn't good for the shareholders, directors are motivated to pursue opportunities to buy other companies or merge with them, so that their company can grow even larger, so that they can earn and control even more.

    This distortion should cause excess growth through mergers and acquisitions.

So what is to be done?

Senator Sanders is certainly right that companies that are too large to fail are a liability, and are too large to exist in the first place. His proposal to solve this by threatening a breakup of corporations that grow too large, makes sense. Companies approaching a clearly defined breakup limit would be motivated to split themselves before the courts do. But the trend for companies to grow until they approach this limit would continue.

I think that the least that can be done to fix this trend is to at least remove the corporate income tax. Losses would be tolerable because corporations are avoiding paying this in the first place, and increasing inefficiencies and systemic risk through that.

However, it would be better still to remove both the corporate and personal income tax altogether, and replace them with a federal sales tax.

Among many other advantages, replacing income taxation with FairTax would remove the shareholder incentive to keep growing their companies when it makes no economic sense, and would leave only the agency problem of mergers & acquisitions. And to solve the problem of companies that grow "too large to fail" in that way, a threat of breakup as proposed by Senator Sanders would be a good way to avoid systemic risk.

Forest fires and risky interdependencies

I have recently proposed the idea (in a comment to this article) that government interventions in the economy are a bit like rangers stopping forest fires. In the short run, it seems as though the intervention works to protect the ecosystem, lives, and property. In the long run, the absence of regular small forest fires leads to a proliferation of undergrowth, which causes an eventual big fire that's so strong, it cannot be contained, and it proceeds to destroy everything.

Forest rangers have since learned that controlling undergrowth is key to long-term safety, so they will allow small fires to burn, and will even start fires on their own, so as to prevent a spread of undergrowth that might provide fuel for future fires.

In the economy, leveraging and other rewarding but risky interdependencies are similar to forest undergrowth. The more such risky interdependencies there are in the economy, the greater the risk of a failure becoming a catastrophe.

Such systemic risk cannot be avoided by simply outlawing complex interdependencies. If people want to put themselves at risk, they will find ways to do so that avoid breaking the law, while still creating risky interdependencies.

The way to avoid systemic risk is to discourage people from wanting to expose themselves in ways that inflate systemic risk. The way to do so is to take the people with most hubris - those whose confidence builds risky interdependencies - and to make them slightly more afraid of systemic risks. And the way to make such people afraid of the systemic risk is, there needs to be a fresh memory of real losses, incurred by people much like them, who ignored systemic risks.

In other words - the forest undergrowth needs to be regularly cleared through periodic small fires.

Is the current financial crisis a small fire? It is not. I'd say it is a medium-to-large fire. I am speculating that we may be having this fire now because past problems, which posed a threat to risky interdependencies, were quickly "fixed" through government intervention, limiting losses and encouraging the build-up of systemic risk.

We are now at a point where Deutsche Bank is leveraged 50 times, with total liabilities about 80% of German GDP. That means that, if Deutsche Bank's assets lose just 2% of their value, the bank lost all its capital. If that happens, Germany is too small to save Deutsche Bank; only the European Central Bank could, by printing large volumes of money - thus taking from everyone else in the Euro economy.

In the UK, Barclays is leveraged 60 times, with liabilities roughly equalling the GDP of the UK. This means that, if the assets of Barclays lose more than 1.6% of their value, the bank is done. No one can save it. The UK cannot print as many pounds as the ECB can print euros; the pounds are backed by a much smaller economy.

It could be that the opportunities to let the fires burn to clear the undergrowth have come and gone. Ten years ago, the Federal Reserve Bank of New York orchestrated a bailout of Long-Term Capital Management. In 2002, government passed onerous laws tightening regulation after the Enron and other bankruptcies. In 2007, U.S. government facilitated a Bear Stearns bailout.

What officials are saying, every time they intervene in the economy, is that when there's systemic risk, people can count on the government to "do something". The behavior this encourages is that people structure their business as if there is no systemic risk. This, in turn, creates systemic risk.

Had the government not intervened, losses in each instance might have been greater, but so would be the fear of systemic risk. An increase in such fear would have prevented the build up of risky interdependencies, which grow more threatening as each new crisis comes along. If they don't kill us this time, they will in the next, yet bigger crisis.

2008-09-24

American income growth over past 25 years

It has become an oft-repeated statement that American middle class household income stayed flat over the past quarter century, and that income gains have been accruing only to the wealthy.

That statement has turned out to be false:
New Evidence on Taxes and Income
(Arthur B. Laffer, Stephen Moore)

The new Census Bureau data on income and poverty reveal that many of the economic trends in this country are a lot more favorable than America's detractors seems to think. In 2007, overall real median family income increased to $50,233, up $600 from 2006. The real median income for intact families -- mother and father in the home -- rose to $78,000, an all-time high.

[...]

Income gains over the last 30 years have been systematically understated due to several factors. These include:
  • Fall in people per household. [...]
  • Earned income tax credit effect. [...]
  • Income mobility. [...]
Thanks to Rok Spruk for linking to this.

2008-09-18

Debt: let the cataclysm come

In the past few decades running up to the current financial crisis, the United States were increasingly a country of debt.

Debt, debt, debt. If you weren't taking on debt, you were being stupid. It was so easy to borrow money, it was so cheap, most businesses were considered stupid not to rely on debt for financing. It was cheap to borrow in the U.S. because the whole world was putting their savings into the United States, so there was too much money chasing too few investment opportunities.

Year after year after year, decade after decade, this opium of cheap money persisted, to the point that the largest financial companies in the U.S. became so dependent on borrowing, they could not survive without.

Then came a realization that ridiculously large amounts of money have been malinvested. This was caused by subprime mortgage.

Then, in a matter of several months, the ability to borrow disappeared.

Now, every single damn business whose viability has come to depend on borrowing is being threatened with collapse. Several of the largest already have, while others are being pounded.

It is a hurricane, and many people are living in wooden shacks.

Many people are living in palaces that are, as far as their construction goes, just wooden shacks.

Many more people are going to lose jobs working in those palaces, as they collapse.

Some buildings, however, are going to remain standing. Those are the buildings that were built to survive once-in-a-few-decades events. The businesses left standing are those that might have been looked down upon in the past. "Why are you not taking more advantage of debt," others said. "It is so cheap to borrow." "You are losing competitive advantage by not relying more on debt."

Well, it turns out, it is cheap to build a wooden shack. And it can be done much faster than building out of steel-reinforced concrete. The wooden shack is clearly economically superior... until the high winds come.

I see two ways of preventing a disaster like this in the future.
  • One. Let all the wooden shacks collapse, and let only the well-built buildings (and the few lucky shacks) remain standing.

    This causes huge trauma. But the people who were cautious are rewarded: not only are they left standing, but they have a unique opportunity to take a leading role in rebuilding the economy after the collapse. For all the rest, the hope is that their trauma will lead to learning, and that new houses will be built more carefully so as to be able to withstand high winds.

    People do learn from past trauma if they had to suffer it. A real danger is, in fact, that they will over-learn. But people over-reacting individually is better than the government over-reacting universally for all.

  • Two. Do what can be done to prevent most of the wooden shacks collapsing. Help the people whose houses collapsed, at the expense of those whose houses are still standing.

    Note that this is vastly unfair to everyone who put their efforts into building solidly in the past. Not only do they have to pay now to help everyone whose buildings were shabby; their opportunity to take a leading role in the rebuilding is also diminished, as leading economic roles will remain in the hands of those who led the shabby building trend to begin with.

    If people are helped, incentives to solve the causes of catastrophe are removed. Therefore, in order to prevent another disaster, the government must mandate that everyone has to build solid houses from now on. The government has to define what a "solid building" is, which it may or may not do accurately. The government may end up repressing all sorts of useful economic activity by enforcing the "solid building" requirement too much, putting the economy in shackles.
You might not be surprised to learn that I'm in favor of the first proposal. Let all the shabbily built things go down. If they do not, the people who built them are going to remain in power, and things are not going to change that much. That opens up the economy for future crises.

2008-09-13

Saudi Arabian judge okays killing satellite broadcasters

This is how insane Saudi Arabia is:
The most senior judge in Saudi Arabia has said it is permissible to kill the owners of satellite TV channels which broadcast immoral programmes.

[...]

The judge made the comments on a state radio programme.

He was speaking in response to a listener who asked his opinion on the airing of programmes featuring scantily-dressed women during the Muslim holy month of Ramadan.
Consider that, in their country, such attitudes can be considered normal.

This is a country where pretty much all citizens live not from productive work, but from oil. These days, a barrel brings over $100. It takes $2 to extract it out of the ground. Most of the difference goes to Saudis and their likes.

As a result of the whole country being financed by oil, they have the luxury of maintaining a "school" system which consists entirely of learning Qur'an, indoctrinating citizens with all sorts of dysfunctional, extremist religious abominations.

The only reason their people get jobs is because of government quotas, which mandate foreign companies to employ a certain proportion of citizens, who are then "employed" in positions where they can't do too much damage.

Check out also Carmen bin Laden's The Veiled Kingdom, in which she relates her first-person experience of what horrendous destinies await the women born into this place.

The ridiculous plot of Stargate Atlantis

My wife likes to watch Stargate Atlantis, and she also used to like Stargate SG-1.

If you didn't watch the original Stargate, the movie, the fundamental concept is interesting. Stargates are devices built by ancient aliens that permit almost instantaneous travel between planets. This much makes sense, and is a good speculative basis that provided a solid foundation for the film, and could be extended into a good series of stories.

The Stargate SG-1 series bastardized this idea in a number of ways.

To an extent, one can excuse issues such as that most planets have Earth-like atmosphere and gravity, and that inhabitants are or resemble humans. Such planets make sense if the ancient race chose to place stargates on planets where they could live, and the story has a reasonable explanation why humans are spread across the galaxy.

Also to an extent, one can excuse the fact that half the planets look like Canada. It's a TV show. They have to make 20 or so episodes every year. They don't have the budget for exotic locations, or to make everything with CGI.

It is a bit harder to explain why people on all planets speak English. Or even if we imagine it's not English - whatever language that first-time visitors from Earth always understand.

Then came Stargate Atlantis.

Stargate Atlantis puts forth the proposition that, in some galaxy far far away, there are thousands of Earth-like planets, inhabited by humans, who lack any relation to Earth, but nevertheless all speak English.

These humans, living mostly as farmers, but never developed more than 1940s Earth, are hunted by a physiologically and technologically superior civilization, called the Wraith.

The Wraith feed through a mysterious process by which they extract "life" out of people through slits in their palms, making the victim age and decay in the process.

The Wraith are intelligent and travel between planets through stargates, as well as on large interstellar ships. They are humanoid, awfully pale, all have long hair, and of course, they speak English.

Being an accomplished space-faring civilization, the Wraith are at once exceedingly smart, and amazingly stupid.

For example, millions of years ago, the Wraith are supposed to have fought a long war against a supremely advanced human race, which the Wraith won through their numbers as well as their cleverness.

And yet, at the time the Stargate Atlantis story takes place, the Wraith are not smart enough to solve a simple agricultural problem.

You see, the Wraith feed on people, but the number of people in their galaxy is not enough. So the Wraith go into hibernation for centuries, waiting for humans to multiply. But the expedition from Earth has awakened them prematurely, when humans have not yet sufficiently multiplied. So the Wraith go to war against each other, basically over who will eat.

Here's how the Wraith could have easily solved their sustenance problem - something they could do for millions of years before the expedition from Earth arrived to poke at them.

First, they would satisfy their hunger by eating most men, and most women over 35. Those are not very useful for reproduction, so they get eaten.

Then, the Wraith get to work. They use their technological prowess to start mass-producing food for humans, possibly on suitable planets. Meanwhile, they install their representatives among humans to establish a religion and keep the humans docile. They start bringing the humans food, and teaching them that the highest religious imperative is to eat and reproduce. The vast majority of humans are either girls or women of childbearing age, so reproduction is fast. The Wraith crank up their agricultural production as the human population grows.

In optimal conditions, if every woman of age 12-35 keeps reproducing and gives birth to only girls - it should not be difficult for the Wraith to achieve that - then annual population growth approaches 16% after 15 years. If the Wraith continue to feed humans for some time while eating few, the population multiplies 100-fold over a span of 30 years. Add some genetic engineering to make girls grow up faster and produce babies sooner, and you have a genuine agricultural boom.

The Wraith could even live in a much better kind of harmony with humans, than humans do with cows.

The humans would believe that the Wraith are benevolent gods who bring them food, medicine, and shelter. If they were wise, the Wraith would never eat a human in front of others, but would first take them to a different planet, under the guise of taking them to paradise - something that each human would be led to long await. A small minority of humans would, in fact, be taken to a kind of paradise, so that they could return back home and spread the word of how great it is.

As far as the show goes, no Wraith ethics committee requires that humans used for food must be wild-caught. So there's no reason for them not to farm humans, much the way that we farm cows.

Or, the Wraith could learn a few things about their biochemistry, and make some minimal adjustments so that they can feed on chicken or turkey instead.

2008-09-04

Evidence-Based Medicine

Robin Hanson and Free Exchange have recently focused attention on this amusing paper satirizing Evidence-Based Medicine - the über-strict drug licensing approach that has been a major factor in the exorbitant cost of bringing new drugs to market.
A couple of snippets from the paper:
Smith GCS, Pell JP. (2003). Parachute use to prevent death and major trauma related to gravitational challenge: systematic review of randomised controlled trials. BMJ, 327(7429), 1459-1461.

[...]

Results We were unable to identify any randomised controlled trials of parachute intervention.

Conclusions As with many interventions intended to prevent ill health, the effectiveness of parachutes has not been subjected to rigorous evaluation by using randomised controlled trials. Advocates of evidence based medicine have criticised the adoption of interventions evaluated by using only observational data. We think that everyone might benefit if the most radical protagonists of evidence based medicine organised and participated in a double blind, randomised, placebo controlled, crossover trial of the parachute.

[...]

One of the major weaknesses of observational data is the possibility of bias, including selection bias and reporting bias, which can be obviated largely by using randomised controlled trials. The relevance to parachute use is that individuals jumping from aircraft without the help of a parachute are likely to have a high prevalence of pre-existing psychiatric morbidity. Individuals who use parachutes are likely to have less psychiatric morbidity and may also differ in key demographic factors, such as income and cigarette use. It follows, therefore, that the apparent protective effect of parachutes may be merely an example of the "healthy cohort" effect. Observational studies typically use multivariate analytical approaches, using maximum likelihood based modelling methods to try to adjust estimates of relative risk for these biases. Distasteful as these statistical adjustments are for the cognoscenti of evidence based medicine, no such analyses exist for assessing the presumed effects of the parachute.

[...]

Only two options exist. The first is that we accept that, under exceptional circumstances, common sense might be applied when considering the potential risks and benefits of interventions. The second is that we continue our quest for the holy grail of exclusively evidence based interventions and preclude parachute use outside the context of a properly conducted trial. The dependency we have created in our population may make recruitment of the unenlightened masses to such a trial difficult. If so, we feel assured that those who advocate evidence based medicine and criticise use of interventions that lack an evidence base will not hesitate to demonstrate their commitment by volunteering for a double blind, randomised, placebo controlled, crossover trial.
A commenter (CPT4ICD9) explains:
EBM is something that on the surface makes absolute sense. Of course we want to see that the device or drug we are using has passed the highest level of scrutiny by the experts and authorities. The reality is that it is simply not ethical or practical to create or operate the trials necessary to meet the EBM requirements of the insurance companies and government. It is also prohibitively expensive to do for virtually every startup technology company and the reason why it now costs $1billion to develop a new drug. Just as we have computer simulation developing most of our movies these days, we need the industry to focus on the development of computer simulations and models that will allow the scientists and physicians to accurately assess the impact of the new technologies and drugs on humans WITHOUT having to wait ten years or longer to have the results prove out what we empirically already knew would happen. The challenge is to do so without missing a hidden element as was done with NSAIDS and many other devices and drugs beginning with DES. It took two complete generations before the problems inherent in DES were discovered. New drugs and technologies should not be denied to patients in need while we wait 20 years to see if anyone dies from them. The scientific community needs to develop models to assess their impact on patients with varying degrees of illness and injury. The healthcare system needs to eliminate the liability involved as long as someone signs a waiver when taking a drug or using a device. The insurance companies need to compensate physicians and facilities using the drugs or devices in a fair and consistent manner even if the evidence is being developed. Without such changes we will see a tremendous decline in medical advances outside of the university setting.

2008-09-02

Google strikes blow to Microsoft with new browser

Microsoft can now officially feel yet more threatened: Google has released a beta version of Chrome, its new browser. A harbinger of things to be is this text from their "Why?" page:
[...] We also built V8, a more powerful JavaScript engine, to power the next generation of web applications that aren't even possible in today's browsers.

This is just the beginning - Google Chrome is far from done. [...]
Microsoft has two vast revenue streams: Windows, and Office. Google already has web-based competition for Office, but online applications are still severely hampered by the capabilities of today's browsers. For now, this has tilted the playing field in favor of traditional applications that install on the user's machine, providing superior functionality.

But if Google's browser gets widespread adoption, it appears certain to lead the way into support for more advanced browser features which will allow online web applications to become more capable of competing with their desktop counterparts. And at some point in the next few years, you will find yourself at a computer that doesn't have Office installed, you will need to perform a spreadsheet computation, and you will find yourself considering:

Do I (A) go to the store and pay $400 for MS Office? Or do I (B) go to OpenOffice.org, and wait while I download a huge installer? Or do I (C) just surf to Google.com, and use an online spreadsheet that immediately works?

Considering also that, in option (C), your data is stored on the net, can be accessed any time from anywhere, and is automatically backed up?

More and more people will find themselves choosing (C) over the previous two options, until eventually, option (C) will make such obvious sense that Microsoft's Office revenue will drop right through the floor.

That's 50% of Microsoft's revenue being threatened right there. Can the desktop eventually become so unimportant that Microsoft's Windows revenue will also be threatened? Maybe.

Disclaimer: Since me and my company depend on developing SSH software for Windows, it is certainly in my best interest for Windows to remain very healthy for a very long time. That doesn't mean it's going to happen, though.

Planned public transportation in Santiago

Felix Salmon publishes a post relating to Santiago's failed attempt to convert a free-er, competition-based public transportation system into a more "planned" and highly regulated one. He relates an email exchange he had with Austan Goolsbee, now Barack Obama's economics advisor, where Mr. Goolsbee argued:
Are you under the impression that replacing competing companies with a few giant firms and then removing all incentives to the drivers is a good idea? It is a recipe for monopolization. Prices will rise and delays will get worse.
And, indeed:
Almost overnight, the new "planned" system cut mass transit ridership, increased congestion everywhere in the city, and tripled average commute times from forty minutes to two hours. As President Michelle Bachelet later said in a speech, "It is not common for a president to stand before the nation and say 'Things haven't gone well.... But that is exactly what I want to say in the case of Transantiago.... The inhabitants of Santiago, especially the poorest, deserve an apology."

The roll-out was not a total disaster, however. The new planned system did solve one of the major problems it had targeted: profits were eliminated overnight. Where the old system had made $60 million a year, the new planned system immediately began to lose, and has continued to lose, more than $600 million per year.
Michael Munger's article, which contains the above excerpt, is well worth a read, and discusses a few possible measures that would fix the previous competition-based system (which had problems such as aggressive drivers, polluting buses), while steering clear of the highly regulated, "planned" solution, which turned out to be much, much worse.

The way I see it, successful social interventions tend to be based on smartly reconfiguring the system of incentives and property rights so as to produce a favorable outcome. Systems generally work if they produce a favorable outcome when everyone acts out of their self-interest. On the other hand, systems whose design ignores people's self-interest, and the behavior stemming from the incentives they create, generally backfire disastrously.