An Answer I'd Like to Hear

In the past couple of weeks, I've heard innumerable questions that go something like this:

How can you justify a system where the little guy has to play by one set of rules and the rich get even richer playing by another set?

The typical answer (usually drowned out by applause) involves mumbling about job creation or possibly agreeing that the rules need reform, but the rich are playing by the rules as they're currently set up. Which leads to the inevitable comeback question (more of a semi-triumphant statement, actually) that the rich can afford lobbyists to bend the rules in their favor.

Wrong answer. Here's the right one:

All legislation and regulation attracts big players that seek to influence the rules to their own advantage. This is inevitable. But putting additional rules in place to prevent this is only pouring gasoline on the fire. The more complex the rule structure, and the more overbearing the government becomes in selectively enforcing the rules, the worse the problem becomes.

If you want to help the little guy--especially the small businessperson--then reduce the power of government to compel compliance. Compliance is a fixed cost; it costs close to the same amount for a company with $5 million in sales to comply with the regulations as it does for a company with $50 billion. So the cost per revenue dollar of the regulations is 10,000 times higher for the small company as it is for the large company. It's no wonder that the big guys love the regulations; it's a virtually free way to stifle any upcoming competition.

Of course, this requires arithmetic, and therefore no politician will ever use this argument.

Democracies and Industrial Policy

Fareed Zakaria had an interesting comment about industrial policy on his show yesterday:

Andrew Liveris, the CEO of Dow Chemical, has been arguing for a national policy aimed at reviving manufacturing. In his book, Make It In America, Liveris argues that not only would a manufacturing policy produce good, long-term jobs, it would upgrade the job skills that are crucial to keeping innovation alive.

"Innovation doesn't just happen in laboratories by researchers," he told me. "It happens on the factory floor. The process of making stuff helps you experiment and produce new products. If everything is made in China, people there will gain the skills, knowledge and experience to innovate. And we will be left behind." He worries that with tablets like the iPad and Kindle being made mostly in Asia, the next generation of these products could well be imagined there.

Companies can't compete with countries, he says.

Take solar energy, an industry largely invented in America, in which the manufacturing has largely moved to China thanks to massive state subsidies. Or consider wind turbines: China's biggest windmill makers have received more than $15.5 billion in credits from state-owned banks.

As a result, despite many concerns about quality, they won their first major foreign orders in the past year.

In theory, I am deeply skeptical of government industrial policy. Government doesn't know how to pick winners and losers, it makes mistakes, the process gets politicized.

And yet, when I look around the world, particularly in Asia, I see governments playing a crucial role. They do make mistakes – their versions of Solyndra – but they seem to view it like venture capitalists.

Their role is to seed many companies, only a few of which will succeed. Once successful, the government helps these companies to compete against big American multinationals.

There used to be a joke about Marxist economists who would say of a deviation from pure Communist economics: "It might work in practice, comrade, but it doesn't work in theory."

That's what industrial policy looks like these days.

The theory doesn't make much sense but it's hard to argue with the results.

I'm more than deeply skeptical about government industrial policy, but I think the key difference between Fareed and me is that I'm mostly skeptical of such policies in a democracy.  China and most of the former Asian tigers aren't democracies; they're authoritarian capitalist economies.  When you've got only marginally accountable leaders, they can indeed act like venture capitalists, because they can make nimble decisions about where to invest.  (The fact that a certain number of those nimble decisions would be considered corrupt by a democracy is almost irrelevant to the leaders of such states.  An authoritarian system exists to perpetuate the power of the leadership and its friends.)

The Chinese government in particular is its own venture capital firm, because the government has no compunctions about being a shareholder in whatever companies it views as interesting.  Tian and Estrin estimate that the Chinese government holds a majority shareholder position in  31% of all Chinese publicly listed companies.  The government is perfectly content to reap a big chunk of the profits from these companies, something that would be unheard of in the US.

If a democracy tries to do the same thing, it is immediately besieged by rent-seekers whom it is not allowed to shoot.  Every decision will be lobbied, questioned, and criticized, with no risk of acquiring unsightly holes in the back of one's head.  This has two effects:  First, it makes all investment decisions incredibly bland, and bland investing tends to have a lousy return.  But more important, it makes every investment decision permanent.  Once the rent-seekers get involved, the machinery becomes impossibly ponderous.

So the real question is, can democracies have a government-funded venture capital system that makes intelligent investment decisions, but is relatively immune to rent seeking?

As I understand it, most venture firms raise capital for their investments as private equity funds.  Large investors form limited partnerships and commit capital for the fund, which is then called down by the fund managers as investments are made in the startups in the portfolio.  This is very different from a typical public LLC, which would just sell stock and raise capital from the sale.  I'm sure there are lots of good reasons to use the private equity model, but it does necessarily tend to keep the amounts of capital bounded, and there is less incentive to scale the investments up, or across more startup companies.


I keep wondering how this would change if there was some kind of GSE that could securitize debt for venture capital firms, giving them a much larger access to cheap capital than they'd have from private equity partners.  If this all sounds a little iffy after September, 2008, you're right--it is.  But you could presumably issue the debt instruments so that you couldn't write credit default swaps against them.  It would then be fairly easy for the government to select sectors to invest in, and direct more or less money to each sector by changing the securitization commitments, without ever getting involved in picking the individual companies.   The government could even buy the bonds themselves if the private sector didn't have an appetite for them.

This is hardly immune to rent-seeking, but at least the lobbying is occurring at the sector level, rather than at the level of individual companies.

I'm sure when I wake up in the morning this will sound like a terrible idea.  Still...

Could This Be Made Into a Business?

Reports of Romney opening up a bit of wiggle room on mortgage modification got me to thinking about why somebody isn't making serious, organized money off of short sales--and helping homeowners at the same time.  The market for this would the set of people who should strategically default, that is, homeowners whose houses are more than 10% underwater but who can still pay their mortgages.  These people haven't defaulted yet for one of the following reasons:

1. They're emotionally attached to their homes and will do almost anything to avoid moving.
2. They're afraid of their credit rating taking a hit, thereby preventing them from buying another house.
3. They think it's morally wrong to default on your debts when you can pay.

So here's my question:  why wouldn't a mortgage originator purchase short sales for these people with the (contractual) understanding that they would immediately sell back the house to the homeowner for a fixed percentage price greater than the short sale price, and give them a 100% mortgage at the prevailing price?

In a little more detail, here's how it would work:

1. Shorts 'R' Us approaches Joe Blow, the homeowner:  "We understand from the tax rolls in your area that your house is 20% underwater.  We'll broker a short sale with your lender, buy the house ourselves, sell it back to you at a profit to us, and give you a new mortgage.
2. Joe Blow signs a contract with Short 'R' Us.  Joe Blow can get out of the contract if Shorts 'R' Us can't broker the short sale with the lender.
3. Shorts 'R' Us goes to the lender offering to buy the property in a short sale, of course disclosing the deal with Joe Blow, as required by RESPA.
4. Will the lender agree to this?  If it's convinced that the alternative is strategic default, I don't see why not.  They have to mark the mortgage as a loss, but if they carry the foreclosed property, they're writing down the same loss when they auction it off.  Of course, if there are secondary lienholders, things get a lot more complicated.
5. The short sale goes through.  Shorts 'R' Us buys the house and pays off the lender.
6. Shorts 'R' Us then immediately sells the house back to the homeowner for 2-3% more than the short sale value, and originates a new loan for the homeowner.
7. Shorts 'R' Us then sells the loan to the secondary market (or as much of it as is now allowed by Dodd-Frank). 
8. Lather, rinse, repeat.

What's in it for the original lender?  Well, he doesn't have to deal with the foreclosure, and he's got the crappy loan off of his books.  He does have to recognize the loss, but, duh, is that really an issue at this point?

What's in it for the homeowner?  He's got an almost-sustainable loan, with substantially reduced principal, although at a somewhat higher interest rate.   (More on this in a moment.)  He's got a fairly bad ding to his credit rating, but he probably doesn't care that much because he's got a new mortgage.  As long as he pays as agreed for a couple of years, he'll be fine, even if he trades properties.  Besides, a quarter of the country has the same ding on their credit rating.  Who cares?

Let's now look at Shorts 'R' Us in a bit more detail.  First, they're instantly making their markup on the buy-resell transaction, virtually risk-free.  It's possible that the original mortgage lender will demand that they kick in some commission, which would either reduce their profitability or force them to demand a higher markup.  Note that as the markup increases, you lose some chunk of your addressable market, because the homeowner's going to make the decision to do this based on whether his payments go down or at least stay the same.  It's likely that the new mortgage has a higher interest rate, because it's instantly underwater, although not by much.  What kinds of secondary markets are available may be somewhat limited, and the amount by which they're going to discount the loan is going to be a key variable.  On the other hand, the credit quality of the loans is higher than it looks on paper, because the homeowners have been paying as agreed and they're motivated to stay in the homes, or they wouldn't be doing this.

To sum up:

Credits to Shorts 'R' Us:

• The markup on the re-sell

Debits:

• Commission to the original mortgage lender.
• Discount on the sale of the loan to the secondary market.

What am I missing?  This sounds like a slam-dunk business.  Is there something that makes it illegal?

Follow the Web of Competencies Toward Personal and Professional Fulfillment!

There's an excellent piece at Forbes by Daniel Jelski on re-architecting the higher education system.  As I was commenting on it, I realized that there's something implied in the Open Education Manifesto that I should have spent a bit more time on.

We tend to think of education in terms of classes or courses:  "Third grade Math". "Western Civilization", "High School Chemistry".  "Nineteenth Century Romantic Poetry".  But that's mostly a convenience for an outdated form of educational record keeping.  When you peel off the cover of these courses, you find that they're a shorthand for a particular bag of competencies.  Some of the competencies are obvious.  For instance, in third grade math, you learn to multiply multi-digit numbers.  In Western Civ, you learn a set of facts about classical civilizations, the Middle Ages, the Renaissance, etc.  But there are other things you learn that aren't as easily quantified but may be more valuable.  For instance, in order to multiply two four-digit numbers together, you have to have a slightly larger coding span than you did to learn your arithmetic facts.  To understand Western Civ, you have to infer the role of demographic migration in shaping societies.  Neither of these skills will ever show up as the title of a unit in a syllabus for the course, but they are more broadly applicable than the knowledge imparted in the course itself.

Your education teaches you not only static knowledge but a set of behaviors.  The knowledge is often essential for correctly executing the behaviors, but the behaviors are what your eventual employer is looking for.  In short, the employer doesn't want to know what classes you took.  He wants to know a list of things you're competent to do.

A couple months back, there were a spate of posts (here, here, and here) that pointed back to a paper by Lauren Rivera, who researched hiring practices at super-elite organizations (high-end financial firms, white-shoe law firms, etc.)  The key takeaway was that elite firms tended to hire only from super-elite universities, often from as few as four universities, not because they thought that the curriculum or educational quality was significantly better, but because acceptance into those universities was interpreted as a signal that the applicant possessed exceptional qualities.  In other words, the firms were outsourcing their applicant screening to the admissions departments of the super-elite schools.

At the bottom of all of this is a very simple fact:  Hiring excellent candidates out of school is a crapshoot, because there is no objective way to determine the candidate's actual competency in the tasks for which he's being hired.  There are two huge problems here.  First, the actual performance of companies is significantly reduced because a fair percentage of new hires don't work out.  Second, and more important from a societal standpoint, access to elite jobs is highly constrained by acceptance to elite universities, which in turn is a consequence of the applicants fitting an extremely narrow set of criteria.  And those criteria seem to be inculcated best when your parents had the same educational background.  This has profound negative effects on social mobility.  Per Scott Winship:

...evidence indicates that American children born since the 1950s have had lower educational mobility than children in Sweden and other Western nations. And recent research indicates that the link between parental income and educational advantages on one hand and child academic outcomes on the other is stronger in the United States than in other Western countries. So it may be that higher pay for better slots and narrower opportunities to occupy the best slots both now contribute to lower earnings mobility in the United States.

Megan McArdle concludes from this:

You can argue about why this is--are the upper middle class transmitting real skills, or pull?  But does it matter?  As an editor at The Economist once noted to me, it's actually rather more worrying if what they're giving their children is a strong education and an absolutely ferocious work ethic.  An aristocracy that simply bequeaths money and social position to its children will eventually fall.  [An] aristocracy that bequeaths the actual skills required to earn more money than everyone else is self perpetuating.  

And self-legitimating.  The old aristocracy was, I think, at least dimly aware that it wasn't quite fair for them to have what they had by mere virtue of being born to the right parents. But in the new aristocracy, it is rarely enough to just get born to the right parents; you also have to work very hard.  (Higher earning men are now more likely to work more than 50 hours a week than are men in lower earnings quintiles.)  Whatever the systemic injustices, it's also quite clear to everyone . . . even parasitic leeches of investment bankers . . . that their salaries only come as the result of frantic effort.  

The ability of one's parents to confer such enduring advantages is obviously unfair.  And while I don't want to say that a society cannot last that way--obviously, many have, for hundreds of years--I don't think it's healthy for society.  It is hard to get civic engagement, or respect for the law, when the bottom 40% or so feels that the game is rigged.

This is a hell of a social consequence for what is essentially a data collection deficiency.  The hiring managers at these companies are not idiots; if they have access to data that allows them to get a higher hit rate from new hires than they would by looking for good schools on the resumes, they'll use it.  And the really sad fact is that they know exactly what they're looking for.  With only a little reflection, I'm sure that they could reel off a set of attributes that their perfect new-hire would have.  If only they had the data!

In an open education environment, there's absolutely no need to lump multiple competencies into a single piece of courseware.  And even if they are lumped together, the testing and record-keeping for them is easy to separate.  As a result, hiring managers can get exactly what they're looking for.

There's another potential advantage to fine-grained competency records:  They're a pedagogical goldmine.  Imagine a web of competencies, ordered by prerequisites, starting in early childhood and extending through the life of each indvidual.  It starts with walking and talking and proceeds up through a set of motor skills, also through shapes, colors, numbers, and letters, then on into the set of academic skills learned in elementary school.  But you don't have to just include academic skills.  Physcial education and social skills could also be mapped and tested.

The result of such fine-grained, exhaustive record-keeping will almost certainly show that children that have difficulty with more advanced competencies have patterns of difficulty with more elementary skills.  This will allow earlier invervention and remediation.  It will also almost certainly allow teachers to classify children into various learning styles and allow optimal teaching methods and courseware to be applied to each child.

As children enter high school a good guidance counselor today uses experience, intuition, and testing to begin to feed students into educational paths that will ultimately result in a rewarding job for which the individual is well-suited.  To say that the success of modern guidance counselors is less than stellar would be an understatement. But full competency records would take most of the guesswork out of this process, allowing students to enhance their strengths with sets of more advanced competencies that depend on those strengths.  This process can continue throughout secondary education and on into higher education.

Except in this model, there is no higher education.  There is merely a process of continuous enhancement, leading ultimately to a career.  When the individual transitions from student to worker is now merely a function of where the chain of competencies for which he's suited--and best enjoys--finally qualifies him as a valuable employee.

Let's now think about an individual in an entry level job.  He has that job because he's attained good scores on the competencies required by the job.  But this individual is ambitious, and wants to move the next step up the ladder to a job with more responsibility.  Because that job also has the competency requirements listed, he know exactly what additional education--or on-the-job experience--he needs to acquire.  Investments in continuing education therefore become highly focused and require a minimum of additional expense.  No more taking course that lead nowhere.

Even more important, the same process applies for displaced or dissatisfied workers.  If it turns out you hate your job, or your career has just evaporated because a robot can do it better than you can, you're going to want to acquire the minimum new set of competencies that will qualify you for a job that meets your temperamental and remunerative requirements.  Furthermore, your past history can provide a good indication of how likely you are to be successful if you choose retraining toward a particular career.  Competency-based education therefore retrains displaced workers in the minimum amount of time, at a minimum expense, while simultaneously maximizing the chances of success.

Human beings are complex animals, exhibiting a complex set of interacting behaviors.  But the number of behaviors associated with each individual is measured in the high hundreds or low thousands, not in the tens or hundreds of thousands.  It is well within our current technology to enumerate those behaviors, organize them into logical progressions of new skills, and guide children and adults through those skills to attain employment that makes the best use of each individual's talents.  The world is turning into a scary place, with technology steamrolling workers faster than the workers can keep up.  Competency-based education and record-keeping provides a way to remove some of the fear from life in world growing ever more complex.

What I'd Like to Hear From Romney About the Free Market

From RedState:

I don’t like seeing pro-free-market Republicans attacking the concept of what Bain does, any more than I liked seeing Romney attack Rick Perry from the left on entitlements. But just because the role of red-in-tooth-and-claw capitalists is a crucial and necessary one does not mean that they are likely to be popular candidates in today’s general election environment. Criminal defense lawyers, for example may be crucially necessary to our system of justice, but if they have represented a lot of unpopular clients, they are not likely to be politically viable. I continue to think that Romney’s business record is an under-explored political vulnerability (one Ted Kennedy used against Romney in 1994, but didn’t even use all the ads he cut) that the Democrats will exploit ruthlessly. And Romney’s existing defenses of that record are fairly weak. We should not be caught unawares by this in the summer and fall when it’s too late to pick another candidate. In many ways, it’s like the swift boat story. You’ll recall that the centerpiece of John Kerry’s electability argument in 2004 was his military record – not any policy proposal on national security, mind you, but the simple fact of his biography as a war hero. Given that Kerry had decades-old enemies from his activties as an anti-war protestor, it was unwise for Democrats to assume that this biographical narrative alone would go unchallenged in the general election. But that’s exactly what they did, and the Swift Boat Veterans’ ads (especially the ads using Kerry’s own Senate testimony from 1970) did terrible damage to Kerry.

Romney’s story is much the same. There’s no serious argument that Romney’s record of supporting free enterprise and job growth in his single term as Massachusetts governor is better than the records of Perry, Gingrich, Santorum and Huntsman; his claim to be a job creation specialist is grounded in his record at Bain, and just like Kerry’s war hero biography, this claim is bound to attract scrutiny. It would be foolishness in the extreme for Republicans to demand that nobody talk about this during the time when we’re choosing a candidate. The harder question, for free-market Republicans, is how to have a serious debate on this point without compromising our integrity and our principles.

At this point, I'd say that Romney is the only viable candidate to shake out from the Republican Bag o' Nuts, and that I'm more closely ideologically aligned with him than any other candidate--Republican or Democratic--that's on the field right now.  And I think his public speaking and persuasiveness has improved by leaps and bounds over the last four years.  You could call me a supporter and I wouldn't protest.

But he's still sooooooo cautious.

Mind you, caution is essential when every word out of your mouth will be carefully graded and selected for how good a cudgel it makes for your enemies.  But the problem is that, per above, Romney's key advantages are not easily condensed to sound bites.  Modern business--both big and small--is incredibly complex, because business systems evolve the same way that biological and social and economic systems evolve, and evolution is all about complexity.  Unfortunately, evolution is also about survival and competiltion for resources, so modern business is incredibly ruthless.

So Romney needs a simple explanation for what it is that he did, and what it is that business in a free market does, and why, ruthless though this may be, this ultimately contributes to human progress better and faster than any planned "humane" alternative.  That's hard to do without acknowledging that the last five years have scored very high on "ruthless" and very, very low on "humane".

Perhaps a way forward is suggested by David Brooks in his recent missive in faint praise of Rick Santorum:

Mr. Santorum understands that we have to fuse economics talk and values talk. But he hasn't appreciated that the biggest challenge to stable families, healthy communities and the other seedbeds of virtue is not coastal elites. It's technological change; it's globalization; it's personal mobility and expanded opportunity; it's an information-age economy built on self-transformation and perpetual rebranding instead of fixed inner character. It is the very forces that give us the dynamism and opportunities in the first place.

Mr. Santorum doesn't yet see that once you start thinking about how to foster an economic system that would nurture our virtues, you wind up with an agenda far more drastic and transformational.

If you believe in the dignity of labor, it makes sense to support an infrastructure program that allows more people to practice the habits of industry.

If you believe in personal responsibility, you have to force Americans to receive only as much government as they are willing to pay for.

If you believe in the centrality of family, you have to have a government that both encourages marriage and also supplies wage subsidies to men to make them marriageable.

If you believe social trust is the precondition for a healthy society, you have to have a simplified tax code that inspires trust instead of degrading it.

If you believe that firm attachments and stable relationships build human capital, you had better offer early education for children in disorganized neighborhoods.

If you want capitalists thinking for the long term and getting the most out of their workers, you have to encourage companies to be more deeply rooted in local communities rather than just free-floating instruments of capital markets.

I generally take a dim view of all forms of social conservatism, because it boils down to the government imposing a set of social solutions instead of letting the social solutions evolve to fit the conditions. But Brooks is on to the most palatable marriage of libertarianism and social conservatism that you're likely to find. True or not (and it's mostly true), the free market going to be perceived as a vast, howling wasteland, littered with the hulks of gutted businesses and the corpses of formerly decent jobs. The only way to survive in that wasteland is to take shelter in your family and your community.

Romney has to make the case that, Darwinian though the free market may be, it is superior to all of the alternatives. This is a hard case to make, because the natural response to this is, "Well, yes, I'm mostly in favor of a free market, but couldn't we cushion the blow if we just did this... and this... and this..." In fact, this has been exactly Obama's set of talking points. He's a great believer in the free market, as long as it can be constrained so that there's no down-side.

It can't be done.

The free market is made up of a large number of agents. Those agents compete for resources, just like any biological system. And, just like any biological system, the agents that can't acquire resources have to die. There's a slightly subtle point here: if the unsuccessful agents didn't die, the resource distribution would have to be fair, and a fair distribution guarantees that the successful agents can't grow. In effect success for some requires death for others.

Fortunately, we're talking about companies and jobs here, not people. When the job dies, the person goes and finds a new one, because the successful agents grow and proliferate. The only alternative--the "fair" alternative--must inevitably result in less growth, and therefore less opportunity for each worker to grow as well.

Obama will further argue that there's a middle ground, where businesses fail, but where the successful businesses are constrained to protect the common weal. This is partially true; you can identify certain pathologies of the free market (monopolies, undercapitalized banking, naked derivatives trading, to name three) outlaw them, and still have a robust system. But growth is closely tied to freedom of action. The more constraints you place on economic agents, the less mobile they become in the free market space, the less likely they are to seize on truly productive opportunities, and the less they'll grow.

Obama believes that there are a nearly unlmited number of constraints that would improve the system, and that the sooner they are discovered and implemented, the more the system is improved overall. He doesn't understand that the harm that the constraints do escalates exponentially in proportion to their number.

This last point is key to Romney's success. He needs to say as loudly as possible that the term "regulated free market" is almost--not quite, but almost--an oxymoron. Crafting this message is incredibly difficult, because it's non-intuitive. But he simply can't be successful without getting this across.

Which leaves us with the short-term social consequences of unfettered capitalism. Per Brooks, this can be made into an opportunity to unite the social conservatives with the libertarians if it's done sensitively and adroitly. The message has to be that families can be trusted to protect themselves as long as jobs are available, and as long as the public space is a level playing field for the best social ideas. The basic compact needs to be that government won't impede job growth, and it will admit that it doesn't know anything useful about social engineering. Then it's up to each family to turn off the TV, monitor their children's internet usage, attend their school board meetings, and talk to their neighbors.

Disruption. Adaptation. Evolution.

I recently ran across Charles Stross's post on how to build a near-future science fictional world.  Stross thinks that the future is merely new layers on top of old stuff.  The only way an old layer of stuff gets discarded is if the new layer of technology/culture/whathaveyou can completely replace the old one, and even then you have to wait for the old stuff to decay.

That's close, but it's not quite right.  Having recently read Tim Harford's Adapt: Why Success Always Starts with Failure, I've been thinking a lot about disruptive events as a framework for looking at how cultures change.

First, a few words on disruptions.  We tend to think of a disruptive innovation as one that suddenly invalidates an existing market, but that's not really what happens.  Most disruptions start out as an innovation in a market that's adjacent to, but not congruent with, the market that's eventually going to get killed.  The innovation is usually crudely featured and not even close to competitive with the state of the art of the disruptee, but it offers a specific capability that a small segment of users desperately need.  Harford's example of this was web-based mail, particularly gmail.  Web mail started out life with bad distribution list management, poor threading, no calendaring, trivial formatting capabilities.  In short, it was utterly unsuited for the enterprise market, which was dominated by Outlook.  But web mail offered one feature that was vital to a certain class of user:  the ability to log on from any browser, anywhere, and interact with mail.  For the user that roamed from computer to computer, device to device, and couldn't afford to install Outlook everywhere he went, the advantage of location-independence offset the crudity of web mail system.

Of course, Microsoft was an early entrant into web mail with Hotmail.  But it never made sense for MS to invest in Hotmail, because they always got a better return investing in knocking out the next 150 features in Outlook to satisfy the Fortune 1000 customers than it did to divert resources to Hotmail.  Google, on the other hand, didn't have a cash cow to protect, so they slowly enriched the Gmail feature set until it's now on the cusp of being able to "suddenly" disrupt the enterprise email market.  I'm confident that 5 years from now, Outlook sales will be down sharply.  10 years from now, Outlook will be quaint.  15 years from now, it will have vanished.  So disruption always sneaks up on you.

The email example is instructive, but it's a fairly modest disruption when we're dealing with the difference between web mail and client-server mail.  But disruptions can impact the culture in huge, unanticipated ways when use of the new, disruptive innovation changes one of our core cultural processes.

Human beings are a lot simpler than you might imagine.  We like to think of our lives as complex, ever-changing tapestries of experiences and interactions with our environment.  But the brain's great talent is to categorize varied experiences into as small a set of conditions as possible, then react to those conditions using as small a set of behaviors as possible.  Yes, those behaviors are executed with minor variations for the unique portions of each condition, but that's the brain's other great talent:  it can adjust a learned behavior in real-time without having to invent a completely new behavior.

This set of learned behaviors is actually pretty small.  You can create a complex culture with a small number of learned behaviors:  Driving a car.  Eating at a table.  Meeting a new person.  Using a toilet.  Taking a shower.  Tucking your kids into bed at night.  Playing catch.  Going on a date.  Buying something.  Engaging in ritual.  Believing--or not believing--in God.  Reading.  Making an argument.  Solving a problem.  Dealing with your emotions.

This is obviously only a fraction of the things that people do, but I'm willing to bet that it's closer to 1% or 2% of the total list of behaviors than it is to .001% or .002%.  Of course, these learned behaviors are assisted by a few doozies that are instinctive or semi-instinctive, including such things as language, sex, locomotion, eating, etc.  But, even though those behaviors are incredibly complicated, they're still fairly finite, and they act as primitives upon which the culturally instilled behaviors rely.

Right about now you'll be asking, "What the hell does this have to do with disruption?"  The answer is that the goods and services we consume have deeply-ingrained cultural behaviors underlying them.  There are tens of thousands of products that dovetail perfectly with how we've learned to prepare and communally eat food.  Other tens of thousands play to the ways we were taught to socialize and attract mates.  Still other products assist our ability to identify and solve complex problems.

In general, products that don't dovetail well with our set of cultural behaviors don't do very well in the market.  But the same processes that cause disruptive products to succeed as they evolve adjacent to more established products can sometimes cause a new, disruptive behavior to be learned adjacent to an established, culturally pervasive behavior.  Cultural disruptions are much rarer than market disruptions.  Returning to the web mail example, the cultural behaviors of using written language to express yourself to another isn't much different between snail mail or client-based email or web mail.  But when you marry the convenience of communicating in text from mobile devices, even though the modes of written expression that you can engage in are vastly more limited than in a letter or email, you begin to see the beginnings of a major cultural disruption, radiating out along several different behaviors.  People who wouldn't dream of composing a letter or email in public are now perfectly content to spend 30 seconds texting someone, with all the attendant rudeness associated with it.  Formal written language gets mutilated because texting is so awkward that linguistic shorthand has to be developed.  And, possibly most significantly, linguistic-based reasoning itself gets changed from an expository form, where large thoughts are formed, expressed, and responded to, to a more interactive form, where simple thoughts bounce back and forth between two or more people to form a sort of gestalt or groupthink.

Just as with product disruptions, the insertion of a cultural disruption doesn't immediately destroy the old forms of behavior, because the old forms are better suited for the majority of the culture that's still oriented around the old environment.  But, even while letters and books and email still make up the majority of the words being written, and while the state of the art of these behaviors becomes increasingly refined, the texters and tweeters and, to a lesser extent, the bloggers and even the PowerPointers are rapidly evolving a new, rich set of cultural behaviors that will ultimately become superior to expository prose and will therefore sweep it away.  We sure as hell aren't there yet, but it'll happen.

So Stross is right about the layers of new stuff, but only up to a point.  Disruptive layers ultimately digest the old layers next to which they evolved.

It's worth looking a little more at today's culture through the lens of cultural disruptions.  They're harder than you might think to spot, because all disruptions appear to be innocuous at first, and the amount of time required for them to affect the culture significantly can be tens or even hundreds of years.  At some point, I'd like to make a list of all of these, but for now, I'd like to concentrate of the most serious cultural disruption of the last thousand years:  the rise of electronic media.

We're used to media now.  It pervades our culture.  But it's important to realize that media has not only disrupted the culture by increasing the flow of information to the individual by at least four orders of magnitude; it's also changed our habits of thought.

Human beings are evolved to process information in real time.  See the sabertooth tiger, run away.  Spot the tasty berry, eat it.  Find your offspring, protect it.  None of these things require complicated reasoning or abstract thought.  And yet abstract thought conferred such a survival advantage that it became highly conserved.  From the rise of language, to the neolithic revolution, up through the civilizations of the ancient and classical world, and even through the rise of the printing press, the renaissance, and the early industrial revolution, the culture conferred greater and greater rewards on those who could think abstractly and use language and writing to communicate those thoughts to others so that they could be acted upon.

So we've got 3 or 4 million years of evolution that has tailored human beings to use an "immediate paradigm":  see, hear, act, repeat.  And that was significantly disrupted by a new, culturally agile "expository paradigm":  talk, listen, read, imagine, think, communicate, act.  The vast majority of cultural progress is based on the expository paradigm.  Effectively, cultural progress was based on words.  And not just any words:  it's based on exposition and reasoned arguments.

This brings us up to 1909 and the first mass radio broadcast by Charles Herrold.

Suddenly, mass culture could be impacted not just by exposition but by conversation.  Drama, which previously had been available only to small audiences, became available to a mass audience.  Movies and television soon followed, with visual depictions of stories and the conversations that went along with them.

Note that we're still firmly rooted in Stross's "new things are layered on old things" paradigm at this point.  But the seeds of the disruption are now sown:  for the first time since the neolithic revolution, there are now cultural constructs that can appeal to humanity's evolved, immediate see/hear/act paradigm.  They're crude constructs.  They're one-way.  And the expository, carefully reasoned forms of communication aren't in much danger of being supplanted.  But a new paradigm, a "media paradigm", much more closely aligned with the evolved immediate paradigm, is in place.

Fast-forward to today.  The expository paradigm is holding on by its fingernails.  Give children a choice between a book and a video game and a majority will choose the latter.  The video game fits the more natural media paradigm, and is therefore much more easily assimilated by a child than the expository paradigm used by the book.  And it's not that the video game isn't educating and developing the child, both cognitively and culturally.  But it isn't training the child to function in the expository world.  We are, in essence, about at the same point in media's cultural disruption that web mail is in its market disruption.  The old paradigm is still dominant for getting real work done, but the smart money is already trying to figure out how to get the same work done in the new paradigm, because it's seen to be inevitable.

Unfortunately, in the case of media, figuring out how to get the same work done with media that is currently being done with exposition isn't obvious.  The world still runs on advances in science, technology, philosophy, ethics.  But the media paradigm doesn't have good replacements for most of the thought that goes on it these disciplines, the majority of whose advances are still firmly rooted in exposition.  The net result is likely to be a temporary slowdown in innovation as the culture develops the tools to accomplish the tasks that only the expository paradigm can still do, in the new media paradigm.

There are glints of light at the end of the tunnel.  Real science can now be accomplished through simulation, which is fundamentally a media tool, albeit one grounded on a simpler, more immediate kind of expository reasoning.  The aforementioned revolution in texting and social media is slowly evolving a new way to interact with complex problems.  Web resources like Khan Academy on other computer learning systems are slowly enriching children's training in the expository paradigm by driving it through media channels.

But the gap is still huge and, unlike Stross's new-layered-on-old hypothesis, there are likely to be real deficits in education for the foreseeable future.  In education, in effect, the new is cannibalizing the old, because we don't know how to develop a child's cognition in both paradigms simultaneously.  Until we figure that out, the disruption will continue.

Too Much Money in Politics? On the Outside or the Inside?

I'm largely unsympathetic to campaign finance reform because it's a form of price control and, as we know (rule #1), price controls cause shortages.  But it occurred to me that it's odd that the finance reform crowd only whines about how lobbyists use campaign money to scuttle policies that are good for the public at large in favor of those that are good for the lobbyists' clients.  Because this is merely money on the outside of the government that's used to get to the real money that's on the inside.

Lobbying is cost effective because, when successful, every dollar spent in lobbying is returned a hundred- or thousand-fold.  Once a client has burrowed through the thin outer skin of the government, it can feed in the ocean of tax dollars sloshing around on the inside.

So if you really want to do something about campaign money on the outside of the government, the first thing to do is to reduce the money inside the government.  If you want more government in the public interest, you have to stop making the government so attractive to private interests.

The Second Most Important Economic Rule of Thumb

Rule #1 is, "Price controls cause shortages." But for some reason Rule #2 has only recently crystallized for me:

Economic Rule #2: Whenever the government subsidizes something, its price goes up.

Let's look at some examples from recent history:

• Companies were subsidized through tax policy to provide health insurance; medical costs went up.
• Healthcare was further subsidized by Medicare and Meidcaid; medical costs went up a lot.
• Student loans were subsidized through Federal loan guarantees; the cost of higher education skyrocketed.
• Ethanol production was subsidized; the price of corn, and all of its dependent foodstuffs, went up.
  
  
• Home ownership was subsidized through federal loan guarantees and cheap bundling of mortgages through Fannie and Freddie; we had a housing bubble.

I'm sure I could add another fifteen or twenty examples of this if I wanted to do some research, but you get the idea. Why is this true? The answer is simple: Free money. It may come out of our tax dollars, but the ability to get not only your own tax dollars but a hunk of everybody else's is at worst a good deal and at best the basis for a lucrative business model.

So: Suppose I'm a doctor, and I charge my patients $10 a visit (hah!), because I know that it's what they can afford if they're going to buy my service on a regular basis. Then I read in the paper that Medicaid will pay for $10 an office visit, and I know that 25% of my patients are on Medicaid. To me, I now know that my patients have, on average, an extra $2.50 that they'll happily give me without changing their buying habits at all. So I raise my price to $12.50 and office visit.

The reason for the price increase is obvious, but there's a deeper, more corrosive dynamic at work under the surface. In the example above, there are winners and losers. For the 25% of patients on Medicaid, they now have an extra $7.50 that they wouldn't otherwise have had. But for the 75% of patients not on Medicaid, they've each lost $2.50 in spending power. The doctor's doing great, the Medicaid patients are doing better, but the other 75% of the doctor's patients are getting screwed. We've destroyed real purchasing power by picking a few winners (the Medicaid patients and the doctors) and disadvantaging a vastly larger number of losers (the other 75%). And in the end, because we've favored the few at the expense of the many, more people need Medicaid, which drives prices even higher, which further impoverishes the many, and the beat goes on.

But it's even worse than that: The doctor increased his prices because he knew how much money was out there to be had. So the net result is that the economic value of the subsidy is completely destroyed by the price increase.

So, what does this tell us about a philosophy of government? Are subsidies ever justified? We clearly want to have some form of social safety net. That implies that we subsidize the poor, but we have to understand that we need to be extremely careful about how we allow the subsidies to flow back into the economy, or they simply become worthless.

But lots of subsidies are instituted for policy reasons, to encourage some sort of behavior that the government finds desirable. These are effectively worthless, because the price increases will always wipe out the value of the subsidy, causing the economic activity that was being encouraged to decrease back to its equilibrium level. In short, policy subsidies are worthless in the long run.

There are still lots of collective action problems that only government spending can address. But behavior modification is never going to be the government's strong suit.

A Couple of Comments on the New Malthusian Limit

I've been thinking a bit more about whether productivity growth rates can permanently outstrip economic growth rates, and what that means if it happens. Comments from me here and here.

I'm starting to wonder if we're not about to bump up against a new kind of Malthusian limit. The old one had increased output leading to birth rates that exceeded food supply, which caused collapse. Is the new version of this improved technology leading to productivity rates that exceed growth rates, causing the number of available jobs to collapse? And does the collapse of jobs presage the collapse of population? Of civilization?

I'm also wondering: what other Malthusian-ish transitions might we have gone through? Clearly there was a point near the Neolithic revolution where we stopped being one of the variables in a predator-prey equation, which seems like a pretty big transition. Are there others? Can we foresee others in the future?

Why Are We So Risk Averse?

From Neal Stephenson:

In his recent book Adapt: Why Success Always Starts with Failure, Tim Harford outlines Charles Darwin’s discovery of a vast array of distinct species in the Galapagos Islands—a state of affairs that contrasts with the picture seen on large continents, where evolutionary experiments tend to get pulled back toward a sort of ecological consensus by interbreeding. “Galapagan isolation” vs. the “nervous corporate hierarchy” is the contrast staked out by Harford in assessing the ability of an organization to innovate.
Most people who work in corporations or academia have witnessed something like the following: A number of engineers are sitting together in a room, bouncing ideas off each other. Out of the discussion emerges a new concept that seems promising. Then some laptop-wielding person in the corner, having performed a quick Google search, announces that this “new” idea is, in fact, an old one—or at least vaguely similar—and has already been tried. Either it failed, or it succeeded. If it failed, then no manager who wants to keep his or her job will approve spending money trying to revive it. If it succeeded, then it’s patented and entry to the market is presumed to be unattainable, since the first people who thought of it will have “first-mover advantage” and will have created “barriers to entry.” The number of seemingly promising ideas that have been crushed in this way must number in the millions.

What if that person in the corner hadn’t been able to do a Google search? It might have required weeks of library research to uncover evidence that the idea wasn’t entirely new—and after a long and toilsome slog through many books, tracking down many references, some relevant, some not. When the precedent was finally unearthed, it might not have seemed like such a direct precedent after all. There might be reasons why it would be worth taking a second crack at the idea, perhaps hybridizing it with innovations from other fields. Hence the virtues of Galapagan isolation.

The counterpart to Galapagan isolation is the struggle for survival on a large continent, where firmly established ecosystems tend to blur and swamp new adaptations. Jaron Lanier, a computer scientist, composer, visual artist, and author of the recent book You are Not a Gadget: A Manifesto, has some insights about the unintended consequences of the Internet—the informational equivalent of a large continent—on our ability to take risks. In the pre-net era, managers were forced to make decisions based on what they knew to be limited information. Today, by contrast, data flows to managers in real time from countless sources that could not even be imagined a couple of generations ago, and powerful computers process, organize, and display the data in ways that are as far beyond the hand-drawn graph-paper plots of my youth as modern video games are to tic-tac-toe. In a world where decision-makers are so close to being omniscient, it’s easy to see risk as a quaint artifact of a primitive and dangerous past.

The illusion of eliminating uncertainty from corporate decision-making is not merely a question of management style or personal preference. In the legal environment that has developed around publicly traded corporations, managers are strongly discouraged from shouldering any risks that they know about—or, in the opinion of some future jury, should have known about—even if they have a hunch that the gamble might pay off in the long run. There is no such thing as “long run” in industries driven by the next quarterly report. The possibility of some innovation making money is just that—a mere possibility that will not have time to materialize before the subpoenas from minority shareholder lawsuits begin to roll in.

Today’s belief in ineluctable certainty is the true innovation-killer of our age. In this environment, the best an audacious manager can do is to develop small improvements to existing systems—climbing the hill, as it were, toward a local maximum, trimming fat, eking out the occasional tiny innovation—like city planners painting bicycle lanes on the streets as a gesture toward solving our energy problems. Any strategy that involves crossing a valley—accepting short-term losses to reach a higher hill in the distance—will soon be brought to a halt by the demands of a system that celebrates short-term gains and tolerates stagnation, but condemns anything else as failure. In short, a world where big stuff can never get done.

I have long thought that a too-perfect flow of information might be socially (if not economically) deleterious. Exhibit A for this is the self-selection that occurs in the blogosphere, where the difference between what the community will tolerate and honest disagreement and what it considers trolldom is vanishingly small. Ditto for political discourse, where the cycle required to attack an opponent's statements, which used to be weeks, has shrunk to minutes. But I have to admit that I hadn't really thought about the internet as a huge idea buzzkill.

On the other hand, the near-pefect flow of information files the rough edges off of genuinely good ideas, so that they're brought to market more efficiently. But any idea with too many rough edges--which is pretty much any large project--is likely to be milled down to nothing.

In Which the Blindingly Obvious Becomes, Oddly Enough, Blindingly Obvious

Michael Lewis has a piece in Vanity Fair that kinda rocked me back on my heels.  It's nominally about debt in California--at all levels of government, but the interesting part comes from an interview with neuronscientist Peter Whybrow:

“Human beings are wandering around with brains that are fabulously limited,” he says cheerfully. “We’ve got the core of the average lizard.” Wrapped around this reptilian core, he explains, is a mammalian layer (associated with maternal concern and social interaction), and around that is wrapped a third layer, which enables feats of memory and the capacity for abstract thought. “The only problem,” he says, “is our passions are still driven by the lizard core. We are set up to acquire as much as we can of things we perceive as scarce, particularly sex, safety, and food.” Even a person on a diet who sensibly avoids coming face-to-face with a piece of chocolate cake will find it hard to control himself if the chocolate cake somehow finds him. Every pastry chef in America understands this, and now neuroscience does, too. “When faced with abundance, the brain’s ancient reward pathways are difficult to suppress,” says Whybrow. “In that moment the value of eating the chocolate cake exceeds the value of the diet. We cannot think down the road when we are faced with the chocolate cake.”

The richest society the world has ever seen has grown rich by devising better and better ways to give people what they want. The effect on the brain of lots of instant gratification is something like the effect on the right hand of cutting off the left: the more the lizard core is used the more dominant it becomes. “What we’re doing is minimizing the use of the part of the brain that lizards don’t have,” says Whybrow. “We’ve created physiological dysfunction. We have lost the ability to self-regulate, at all levels of the society. The $5 million you get paid at Goldman Sachs if you do whatever they ask you to do—that is the chocolate cake upgraded.”

The succession of financial bubbles, and the amassing of personal and public debt, Whybrow views as simply an expression of the lizard-brained way of life. A color-coded map of American personal indebtedness could be laid on top of the Centers for Disease Control’s color-coded map that illustrates the fantastic rise in rates of obesity across the United States since 1985 without disturbing the general pattern. The boom in trading activity in individual stock portfolios; the spread of legalized gambling; the rise of drug and alcohol addiction—it is all of a piece. Everywhere you turn you see Americans sacrifice their long-term interests for short-term rewards.

What happens when a society loses its ability to self-regulate, and insists on sacrificing its long-term interest for short-term rewards? How does the story end? “We could regulate ourselves if we chose to think about it,” Whybrow says. “But it does not appear that is what we are going to do.” Apart from that remote possibility, Whybrow imagines two outcomes. The first he illustrates with a true story, which might be called the parable of the pheasant. Last spring, on sabbatical from the University of Oxford, he was surprised to discover that he was able to rent an apartment inside Blenheim Palace, the Churchill family home. The previous winter at Blenheim had been harsh, and the pheasant hunters had been efficient; as a result, just a single pheasant had survived in the palace gardens. This bird had gained total control of a newly seeded field. Its intake of food, normally regulated by its environment, was now entirely unregulated: it could eat all it wanted, and it did. The pheasant grew so large that, when other birds challenged it for seed, it would simply frighten them away. The fat pheasant became a tourist attraction and even acquired a name: Henry. “Henry was the biggest pheasant anyone had ever seen,” says Whybrow. “Even after he got fat, he just ate and ate.” It didn’t take long before Henry was obese. He could still eat as much as he wanted, but he could no longer fly. Then one day he was gone: a fox ate him.

The other possible outcome was only slightly more hopeful: to hit bottom. To realize what has happened to us—because we have no other choice. “If we refuse to regulate ourselves, the only regulators are our environment,” says Whybrow, “and the way that environment deprives us.” For meaningful change to occur, in other words, we need the environment to administer the necessary level of pain.

Did our value system only work because it was designed for scarcity?  Or would the old-timey Christian values of hard work, community, cooperation, and charity still function, if only they hadn't been eroded by abundance?  Is there any market-based solution that will encourage long-term thinking?  Are we approaching another Malthusian limit where the wealth of our society no longer causes a self-limiting population explosion, but instead fuels a self-limiting explosion in consumption?  Or is this all nonsense, and all we need is the ability for everybody in the society to understand compound interest?

If this thesis is correct, it has vast implications for the future success not only of government in the developed world, but of market-driven economies in general

Why Medicare Cost Control Will Never Work

As part of a series of ongoing attempts to sharpen my argument for why Medicare as structured can't work, even if you institute cost controls, let's briefly return to our repeated attempt to graph the average health care cost for one extra year of life, at various ages.

But this time I have a picture!  This semi-log graph plots age against how much health care costs, on average, to live one extra year.  (Note that all of these numbers are going to be arm-waves, and they're going to be averages.  Look at the shape of the curves, not the actual numbers, and you'll get the idea.) 

Let's look at this.  First, note that all of these curves end abruptly at some value.  This value is the average life expectancy for someone born in the year specified for the individual curve, but there's a better way to think of this number:  It's the point at which medical science simply can't help you any more, no matter how much money you're willing to pay.  This number is important, because today it denotes the only consensus value for when public policy says it's OK to stop paying to extend people's lives.  In other words, by policy, we assume that everybody's life has infinite value, but we acknowledge that at some point we luck out, because everybody dies before they can consume so many medical resources that the whole system collapses.

If you were born in, say, 1820, before any semblance of modern health care, medical care cost virtually nothing, because it was pretty useless.  You were born, you lived, and you died at about age 35.  Medical science couldn't extend your life at all.

By the time people were born in 1920, we'd gotten over the first hump:  You paid considerably more for childhood and early adult health care, but that allowed you to live up until age 45 with routine maintenance.  After 45, though, health care got linearly more expensive, until you died around age 70.  In other words, we could do some stuff for you to make sure you lived to old age, but by the time old age set in, medicine was powerless to do much for you.

A baby boomer born in 1950 not only paid more for childhood and young adult health care, but, when they began to age, their health care became exponentially more expensive.  Now death could be staved off once old age was achieved--at least for a while, and given heroic amounts of money.  Life expectancy rose into your early 80s.  This is close to where we are today.

So what about the kids born in the next 10 or 15 years?  We might guess that intervening at a younger ages will be necessary to ensure a very long and healthy life, but still, as they age, we can guess that the cost of each additional year of life will increase.  However, we can also guess that the cost of living to the same age will cost less and less over time, so each extra year of life is cheaper for the 2025 kid than it was for the 1950 kid--at least to start with.  But we can also expect that the new, jazzy medical science that lets Mr. or Ms. 2025 live to (in this case) 110 costs more and more, eventually far surpassing the 1950 curve.

OK, all you calculus fans:  If you integrate under the curve for as far as its defined, you derive the total amount spent on health care for each individual over his entire life.

Now, imagine if you will, that we eventually extend life indefinitely.  Is there any reason to think that the trend of ever-increasing expense for ever-longer life will be interrupted?  We can certainly expect that the cost of living to the same age as some cohort born earlier will cost less as technology improves, but there are two things that completely trash this from a policy perspective:

1. The cost to achieve ever-longer life continues to rise.
2. It goes on forever.

When integrating under the curve yields an infinite value, the only way that society can afford to pay for your medical care is if you produce at least as much value as you consume in health care.  (I could add in food, clothing, and shelter, but they're all mouse nuts compared to the health care.)

So we're now finally in a position to posit the Radically Moderate Laws of Health Care Costs:

Law #1:  For each individual, the average cost of extending life by some increment is greater than or equal to the cost of extending life by the previous increment.

 and

Law #2:  The only way you can live forever is if you pay for it yourself.

 I think that covers it.

Let's now think about cost controls for Medicare.  As long as we maintain the current policy of paying for everything, forever, you may be able to flatten the curve, but you can't avoid Law #1.  (In practice, if you flatten the curve too much, you wind up with shortages, which ultimately reduces life expectancy.  So you can make Medicare costs sustainable, but only if you're willing to shorten their lives.)  So, at some point, Medicare has to be capped.  Not cost controlled--capped.  We're going to decide that we'll only pay so much as a society to extend the life of each of our citizens.  After that, Law #2 kicks in.

But of course there's a good side to Law #2:  As (rich) people demand--and pay for--better and better technology, the life extension cost curve flattens more and more, making longer and longer lives available even to those depending on the government for health care.  In other words, they get more bang for their capped buck.

So there you have it:  the topsy-turvy world of health care policy, where living longer is a public policy disaster, and the best way to extend life for future generations is to limit it for prior generations.  But even so, this so far outstrips the results of cost controls that there's really no alternative.

Is There a Way Forward for the Nuclear Industry?

I write this on day #5 of the post-tsunami Fukushima Daiichi nuclear crisis. As of right now, here's where we seem to be:

Unit #1 has had a loss-of-coolant accident (LOCA) that has presumably resulted in partial core damage. Tokyo Electric Power (TEPCO) has had to do repeated emergency ventings of steam, and Cs-137 and I-131 have been detected in the environment, both around the plant and further downwind, indicating that the cladding around the fuel rods has been degraded enough to have the steam reacting directly with fission products. There has been a hydrogen explosion in the building that surrounds the actual containment, but the containment--and more importantly the reactor pressure vessel--remain intact. TEPCO, however, has been unable to establish normal recirculation of coolant to the reactor and has resorted to pumping seawater directly into the reactor in an effort to cool it, making the reactor itself a write-off. However, because there is no circulation, TEPCO is periodically having to release radioactive steam to get the pressure low enough to pump in more water.

Unit #3 has almost exactly the same story as unit #1: Same LOCA, same partial core damage, same hydrogen explosion in the outer building, same seawater feed-and-bleed, although everything happened about a day later than in unit #1.

Unit #2 has also had the same LOCA, but it appears that the condensing torus below the reactor has partially failed, going from 3 atmospheres of hydrostatic pressure to only about 1. During the actual failure, very high levels of radiation (tens of milliSieverts an hour) were detected outside the buildings, but this has since declined. This seems (to me) to be consistent with unfiltered reactor stream going straight into the environment without being first filtered through the torus. I'm a bit hazy on exactly when the torus is used for stream venting, but I'd kind guess that the idea here is to quench as much of the steam as possible before it goes into the environment. If there's less water in the torus, we can expect the vented stream to be more radioactive in the future, but not as radioactive as when the radioactive stuff already in the torus spilled out, presumably mostly in liquid water.

Unit #4, which was shut down for maintenance, has had a fire in or near the cooling ponds for spent fuel. If the spent fuel became uncovered and caught fire, then radioactive particulates will have been wafted into the air during the fire, producing a fairly high-dose radiation cloud. The fire is now out. I'm assuming that it's not a huge deal to keep the cooling ponds at all 5 reactors covered up, but the fact that the fire occurred either means that this is harder than I'd expect or somebody is suffering from task overload. It'd be hard to blame them if it's the latter.

Up until today, the highest reading I'd heard for environmental radiation was about one milliSievert (mSv) per hour, but during the torus incident, levels as high as 8 mSv/hr were observed, falling back to about a third of that (let's called it 3 mSv/hr). For comparison, levels of 250-1000 mSv per day are considered enough to produce mild radiation sickness, so it's entirely possible that workers have spent enough time on-site to be moderately poisoned.

Radiation has been detected offshore by both the Seventh Fleet and on land in Tokyo proper, where levels are reported (on the cable news networks) as being approximately 10 times normal background, which would be something like 14 microSieverts (uSv) / day. For comparison, a chest X-ray imparts about 40 uSv. This is obviously not healthy but it's far from disastrous.

Readers of this blog (all 2 of you) will know that I'm about as pro-nuke as you can be. In terms of comparative risk, nuclear power is by far the safest form of energy ever invented. As a global warming non-denier/non-alarmist, I think that, if you're really serious about reducing CO2 emissions, nuclear power is the only option that will scale up in any reasonable time frame. And as a follower of geopolitics, I'm convinced that, in a world where world energy demands are going to skyrocket from increasing affluence in south and east Asia, adoption of nuclear power could be one of the only hopes for maintaining the fragile Pax Americana.

But that may all be out the window now.

I am still confident that the crisis in Fukushima will ultimately be brought to a less-than-catastrophic conclusion. It's gonna be worse than 3 Mile Island, but way, way, way, way better than Chernobyl. In many ways, it will be the best existence proof we have that, even when pretty much everything that can go wrong does go wrong in the middle of a 1000-year black swan event, the impact to the environment is quite minor.

But the amount of hysteria that has been whipped up by the media on this story simply is not going to subside without a long-term impact on public opinion. These people have been absolutely despicable in their eagerness to translate a pretty serious situation into a catastrophe in the interests of selling a few more eyeballs to their advertisers. I have a hard time faulting these guys for wanting to make money--that's what they're in business for, after all--but one would hope that a little bit of action in the public interest would be appropriate. Apparently not. I'd like to say that I'll consume less mainstream news, but mainstream news is still the predominant source for information, all bloggy pretense aside. These people suck, but they suck in a mission-critical kinda way.

So what now?

First, let's be clear that something went seriously wrong with the engineering. One unit having a LOCA during a giant tsunami is one thing; 3 having almost identical failures is evidence of a systemic design flaw. I've seen very little coverage of this point yet, and it is going to be the single most important lesson learned out of this whole mess.

I can think of only two failure modes that might be responsible for this. The first, and less likely, is that the heat exchangers for all units were fouled by the tsunami. These are boiling water reactors. In a BWR, the coolant water actually boils, creating (slightly radioactive) steam, which is expanded through turbines to generate electricity. The steam is then condensed (cooled) in a heat exchanger, so that the reactor pump can pump relatively cool liquid water back into the reactor. (You want to use the same water over and over if you can, because it's mildly radioactive with N-17 and tritium.) This basic recirculation loop forms the cooling cycle for the reactor.

The heat exchanger needs cool water, presumably from the ocean, to cool the purified steam/water in the reactor itself. If, perhaps, the heat exchanger is damaged because it's getting nothing but tsunami muck, then it can't re-condense the stream to liquid water. No liquid water means that the recirculated pump won't work. Voila! LOCA.

But I'd think that providing fail-safe mechanisms to get cool water would be easy to provide. Perhaps the earthquake and tsunami wiped out redundant piping for the heat exchangers? Or maybe the inlets got buried in debris? I'm guessing--no clue.

The second, more likely--and more alarming--possibility is hinted at in an article from the New York Times, on Sunday:

Christopher D. Wilson, a reactor operator and later a manager at Exelon’s Oyster Creek plant, near Toms River, N.J., said, “normally you would just re-establish electricity supply, from the on-site diesel generator or a portable one.” Portable generators have been brought into Fukushima, he said.

Fukushima was designed by General Electric, as Oyster Creek was around the same time, and the two plants are similar. The problem, he said, was that the hookup is done through electric switching equipment that is in a basement room flooded by the tsunami, he said. “Even though you have generators on site, you have to get the water out of the basement,” he said.

This would be a pretty serious design flaw. And it's also a plausible one: If I were designing a redundant system, I'd consider what I had in front of me: A seawall designed to accommodate a 7 meter tsunami, a diesel backup system, and a battery backup. And I'd assume that those systems would allow me enough time to fly in a portable generator if everything went pear-shaped. But I could see myself overlooking the fact that the electrical connections I start out with are the only ones I'll ever get to use, because the switching area is flooded out.

So, at the very least, there's going to be a significant "lessons learned" phase to any rehabilitation of the nuclear industry. It will be incredibly painful, because the industry has marketed itself as having thought of everything. That won't fly, now. Instead, they're going to have to re-brand themselves a bit more humbly. That's a bit of a tight-rope act. Nuclear power needs to be perceived as being absolutely bullet-proof.

The other piece of fallout, if you'll pardon the expression, from the unit #4 fire is that the waste problem is going to go from being something that reasonable people could pooh-pooh to an absolute showstopper. I predict that we'll have legislation outlawing the use of temporary cooling ponds for anything other than waste that's simply too hot to move, and I'm also willing to bet that ponds will be mandated to be retrofitted with hermetic, fire-proof containments. That's a big expense, and it's also yet another licensing hoop to jump through. At the end of the day, I think the argument that a centralized storage facility isn't really necessary just went out the window. At the same time, I'll bet that this accident significantly weakens the case for closed-cycle reprocessing, which of course is really the proper solution to the waste disposal problem.

Ultimately, the one thing that may save the nuke industry is the fact that containment held: "Look everything went wrong, and we destroyed the plant, but the amount of radiation released was minimal." Making that case will require a lot of careful marketing and much more education on when radiation is and isn't dangerous. The public is scared to death of this stuff, and the media is more than happy to prey on their fears in the interest of ratings. This is going to be a tough slog.

UPDATE 1:19 PM CDT: The radiation dose level for a while was higher than I understood (search for the "15 March, 2011, 11:25 UTC" update):

At 00:00 UTC on 15 March a dose rate of 11.9 millisieverts (mSv) per hour was observed. Six hours later, at 06:00 UTC on 15 March a dose rate of 0.6 millisieverts (mSv) per hour was observed.

These observations indicate that the level of radioactivity has been decreasing at the site.

As reported earlier, a 400 millisieverts (mSv) per hour radiation dose observed at Fukushima Daiichi occurred between Units 3 and 4. This is a high dose-level value, but it is a local value at a single location and at a certain point in time. The IAEA continues to confirm the evolution and value of this dose rate. It should be noted that because of this detected value, non-indispensible staff was evacuated from the plant, in line with the Emergency Response Plan, and that the population around the plant is already evacuated.

Update 3/20/11: Here's a handy-dandy radiation dosage chart.

Fun With Public School Statistics

Since it seems to be a day where I'm actually blogging, I've been meaning to post links to some excellent analyses of public education by Dave Burge at Iowahawk. Part 1 dealt with how Wisconsin and Texas public schools actually stack up against each other when one controls for race/ethnicity, and part 2 answers some criticisms that were raised against part 1.

It's way too easy to look at bulk statistics and draw erroneous conclusions, even when you know better. I'd glad that somebody had enough on the ball to do the arithmetic.

(I'd blog on what I thought about the Wisconsin union business if I weren't floating in an Ambien-induced bubble of no better than semi-reality, but I am. Chronic insomnia is a bitch. Maybe later... after I've slept... Oh, wait... I have... I think...)

The Japan Earthquake and the Fukushima Daiichi Nuclear Plant

From World Nuclear News:

Later, Tokyo Electric Power Company (Tepco) reported that emergency diesel generators started as expected at the Fukushima Daiichi nuclear power plant, but then stopped after one hour, leaving units 1, 2 and 3 with no power for important cooling functions. This led the company to notify the government of an 'emergency' situation, which allows local authorities to take additional precautionary measures. An evacuation has been ordered of over 1000 people living within three kilometres, while engineers worked to restore power.

Almost nine hours later, an announcement from the Ministry of Economy, Trade and Industry said that three of four mobile power supplies had arrived at Fukushima Daiichi and cables were being set up to supply emergency power. Other power modules were in transit by air.

However, pressure inside the containment of Unit 1 at Fukushima Daichi had been steadily increasing over the time that emergency core cooling systems have not been active. Tepco reported at 2am that pressure had increased beyond reference levels but was within engineered limits.

The company then announced a decision to reduce the pressure within containment "for those units that cannot confirm certain level of water injection" by the safety systems. "We will endeavor to restore the units and continue monitoring the environment of the site periphery."

This has to be the origin of Clinton's ridiculous statement about the Air Force flying water into Japan; I could understand the Japanese wanting to borrow a mobile power cell or two, but the water thing is idiotic and of course is credulously being repeated in the media.

The anti-nuke crowd is distorting this just as quickly as they can get their hands unwrung. As far as I can tell, Fukushima Daiichi had a double failure, went to battery backup, and is taking steps to bring in alternate power, which is of course what the battery backup is for in the first place. This is clearly a serious accident, but everything seems to be working as designed, as far as I can tell.

UPDATE 9:48 CST: Guess I haven't called this one very well so far:

Meanwhile at adjacent Fukushima Daini, where four reactors have been shut down safely since the earthquake hit, Tepco has notified government of another emergency status.

Unit 1's reactor core isolation cooling system had been operating normally, and this was later supplemented by a separate make-up water condensate system. However, the latter was lost at 5.32am local time when its suppression chamber reached 100ºC. This led Tepco to notify government of another technical emergency situation.

Tepco has announced it will soon begin controlled releases to ease pressure in the containments of units 1, 2 3 and 4 at Fukushima Daini.

A three kilometre evacuation is in progress, with residents in a zone out to ten kilometres given notice of potential expansion.

In addition, the Daiichi site is leaking radiation from something other than its controlled release. No way to tell if this is resulting from core damage or some minor problem.

Now, Let's Play "Get the Metric Defined Right"

Per the previous post, I'm still lost in parameter space trying to get the metric right on marginal utility for end-of-life health care.

First, it varies depending on how old you are. If you're ninety-five, a small amount of money is unlikely to prolong your life much. If you're twenty-three, a small amount of money may give you decades of extra life.

Second, as you move further out on the age axis, there is some point at which an infinite amount of money doesn't by you any more life. We might refer to this as the "your time is going to come" principle. So, on average, if you were to create a graph with the abscissa as "current age" and the ordinate as "cost per extra expected year of life", that graph has a vertical asymptote at some fairly advanced age.

Third, the shape of this graph has varied over time. Most important for our discussion, in days of yore, the graph bounced around at some fairly low value, rose a little bit as you got older, and then suddenly shot up sharply along the vertical axis as you reached your allotted three score and ten (or whatever it really was). Today, however, two things are happening:

1. The vertical asymptote is moving further out.
2. The rate of approach to that asymptote is shallower. In other words, you can actually extend your natural lifespan with some degree of medical intervention

Now, from that it's pretty easy to deduce that for any age 'a' and instantaneous medical expenditures as a function of your age:

infinity
sum(Medical expenditure(t))
t=a

is going to get bigger and bigger, the shallower and shallower that medical expenditure curve gets. But that's not the worst part.

The worst part (from a Medicare solvency standpoint--it's friggin' wonderful from the individual's standpoint) is that that vertical asymptote may be getting ready actually to tilt to the right. That pretty much means that if you want to live forever, a dollar value can be attached to what that will cost. It's likely a pretty high value, but it might no longer be infinite.

I wish somebody could produce this graph, and then animate it over time. I'll bet it would focus our conversation on health care considerably.

This Stuff Always Sounds So Much Better When Someone Other Than I Says It

From a Mike Stopa op-ed piece in the Boston Globe:

The death panel notion persists, however, because it denotes, in a pithy way, the economic realities of scarcity inherent in nationalizing a rapidly developing, high-technology industry on which people’s lives depend in a rather immediate way. G.K. Chesterton once wrote that vulgar notions (and jokes) invariably contain a “subtle and spiritual idea.’’ The subtle and spiritual idea behind “death panels’’ is that life-prolonging medical technology is an expensive, limited commodity and if the market doesn’t determine who gets it, someone else will.

But, as I've harped upon before, it's worse than that. The marginal cost for an extra year of life for the elderly is declining, not increasing. That's actually bad news from a cost-containment standpoint, though. Back when an infinite amount of money got you no extra life, it was pretty simple. Now that mere hundreds of thousands of dollars can buy you a couple of extra years, it's a slam-dunk to want the extra time--as long as somebody else (like Medicare) is paying for it.

The amount of money spent per Medicare recipient has to be capped somehow, or the system will simply collapse. You can do that by denying certain treatments (the "death panel" approach), or you can do it by capping lifetime or per-year benefits.

Guess which one will promote more medical innovation?

Back to Stopa:

To the extent that ObamaCare ultimately succeeds in imposing uniformity on basic health care, it will likely lead to the creation of secondary markets for providing insurance against various health eventualities and access to “heroic’’ procedures to extend life. Water runs downhill and it’s a good thing that it does. First, we need to have people buy the expensive medicines and experimental technologies. Europe has discovered this as its regulated system of medicine has driven its pharmaceutical industry farther and farther behind that of the United States. Capping costs kills innovation.

But, in addition, Palin is right. Death panels are an inevitable consequence of socialized medicine. The law of scarcity demands them.

A mature discussion of health care must recognize basic economics so that we can think ahead on how to satisfy the demands of those who are not satisfied with base-level care.

Maturity--what a concept.

Brief update: I just noticed that I've allowed the words "Palin is right" to enter my blog, even if they were in a quotation. I'm not sure what to do about this.

Can the Financial System Be Made to Operate in the Public Interest?

Tyler Cowan thinks probably not:

A key lesson to take from all of this is that simply railing against income inequality doesn’t get us very far. We have to find a way to prevent or limit major banks from repeatedly going short on volatility at social expense. No one has figured out how to do that yet.

It remains to be seen whether the new financial regulation bill signed into law this past summer will help. The bill does have positive features. First, it forces banks to put up more of their own capital, and thus shareholders will have more skin in the game, inducing them to curtail their risky investments. Second, it also limits the trading activities of banks, although to a currently undetermined extent (many key decisions were kicked into the hands of future regulators). Third, the new “resolution authority” allows financial regulators to impose selective losses, for instance, to punish bondholders if they wish.

We’ll see if these reforms constrain excess risk-taking in the long run. There are reasons for skepticism. Most of all, the required capital cushions simply aren’t that high, so a big enough bet against unexpected outcomes still will yield more financial upside than downside. Furthermore, high capital reserve requirements insulate bank managers from the pressures of both shareholders and bondholders. That could encourage risk-taking and make the underlying problem worse. Autonomous managers often push for risk-taking rather than constrain it.

What about controlling bank risk-taking directly with tight government oversight? That is not practical. There are more ways for banks to take risks than even knowledgeable regulators can possibly control; it just isn’t that easy to oversee a balance sheet with hundreds of billions of dollars on it, especially when short-term positions are wound down before quarterly inspections. It’s also not clear how well regulators can identify risky assets. Some of the worst excesses of the financial crisis were grounded in mortgage-backed assets—a very traditional function of banks—not exotic derivatives trading strategies. Virtually any asset position can be used to bet long odds, one way or another. It is naive to think that underpaid, undertrained regulators can keep up with financial traders, especially when the latter stand to earn billions by circumventing the intent of regulations while remaining within the letter of the law.

It’s a familiar story, repeated many times in the past. If one recalls the Basel I capital agreements for banks, the view was that we would make banks safer by inducing them to hold a lot of AAA-rated mortgage-backed assets. How well did that work out? So, with no disrespect to the regulators or the sponsors of the recent bill, it is hardly clear that enhanced regulation will solve the basic problem.

For the time being, we need to accept the possibility that the financial sector has learned how to game the American (and UK-based) system of state capitalism. It’s no longer obvious that the system is stable at a macro level, and extreme income inequality at the top has been one result of that imbalance. Income inequality is a symptom, however, rather than a cause of the real problem. The root cause of income inequality, viewed in the most general terms, is extreme human ingenuity, albeit of a perverse kind. That is why it is so hard to control.

Another root cause of growing inequality is that the modern world, by so limiting our downside risk, makes extreme risk-taking all too comfortable and easy. More risk-taking will mean more inequality, sooner or later, because winners always emerge from risk-taking. Yet bankers who take bad risks (provided those risks are legal) simply do not end up with bad outcomes in any absolute sense. They still have millions in the bank, lots of human capital and plenty of social status. We’re not going to bring back torture, trial by ordeal or debtors’ prisons, nor should we. Yet the threat of impoverishment and disgrace no longer looms the way it once did, so we no longer can constrain excess financial risk-taking. It’s too soft and cushy a world.

That’s an underappreciated way to think about our modern, wealthy economy: Smart people have greater reach than ever before, and nothing really can go so wrong for them. As a broad-based portrait of the new world, that sounds pretty good, and usually it is. Just keep in mind that every now and then those smart people will be making—collectively—some pretty big mistakes.

How about a world with no bailouts? Why don’t we simply eliminate the safety net for clueless or unlucky risk-takers so that losses equal gains overall? That’s a good idea in principle, but it is hard to put into practice. Once a financial crisis arrives, politicians will seek to limit the damage, and that means they will bail out major financial institutions. Had we not passed TARP and related policies, the United States probably would have faced unemployment rates of 25 percent of higher, as in the Great Depression. The political consequences would not have been pretty. Bank bailouts may sound quite interventionist, and indeed they are, but in relative terms they probably were the most libertarian policy we had on tap. It meant big one-time expenses, but, for the most part, it kept government out of the real economy (the General Motors bailout aside).

The other important idea from this piece is that Cowan equates rising income inequality solely from the abuses of the financial sector. He argues that, if you remove this, inequality has changed very little for 98% of the US population.

He also cites "threshold earners", people who work only to make a specific amount of money and then devote the rest of their time to pursuits that improve their quality of life, as another factor that's making inequality appear worse than it actually is.

If Cowan is right, we're likely to be subject to financial crisis after financial crisis, until we finally learn how to prevent people from gaming the system. The problem is that the people gaming the system are smarter than the people trying to engineer a solution, so they have a built-in advantage that's likely to last decades until the legal and regulatory frameworks have walled-off all the low-hanging fruit.

The big problem here is that the technology is still accelerating the information flows upon which the financial beast feeds, which means that the crises are likely to be sharper, less anticipated, and closer together.

An important piece, very clear, and ultimately pretty depressing.