2008-11-01

15,000 pages

It was mentioned in an article that the deregulation of the CDS mechanism was in a rider written by Phil Gram some years back. This rider was placed on the budget, which that year was about 15,000 pages long.

Hold it right there.

Never mind the CDS or Phil Gram.

15,000 pages?

Now imagine if you were some big business. Something like Bechtel, Exxon, IBM, or GE. And you ran it by producing a budget each year. That budget runs to thousands of pages, each page written carefully to specify exactly how much money is to be spent where by whom on what for how long. And the board gets this sent to them for one long meeting, during which they add riders and occasionally subtract things, then they vote it up or down, as a whole. And then you pretty much have to live with just that until next year, except for really big things like war and rescuing the banking system.

How long would such a company be in business? Unthinkable. Even the most hardened bureaucrat in the business world would know that such a process is ridiculous. And yet, in the last 40 years, our government has grown the size of the budget document by about 200 times what it was in the late 1960s. It has created a monster in which billions of dollars may be hidden in plain sight, a rigid centralized system that would have been the envy of Stalin.

It simply cannot work. It is not even within an order of magnitude of size and complexity of any process that can work. Even if every representative we send to Washington is both a saint and a genius, they cannot succeed with such a process.

It needs to be cut down and decentralized. The force that inflated the 1960's budgets, the desire for more open government, can be met by putting the budget and accounts of individual agencies directly on the web. The central government should just balance the big-picture budget: so much for transport, so much for health, so much for education. Maybe with some guidelines for how much may be spent redeeming bonds, how much on current costs, and how much on investments in people and infrastructure. Then the agencies can internally decide how to split that into various projects. Congress can act as a board of directors, like committees, and the plan can be visible in blogs, published budgets projected and actual, as well as reports on results.

Sure, there will still be a lot of paper. But the big picture will be clearer and the agency budgets will of necessity be more practical. Less places to stick the pork and more abiltity to match budgets to priorities.

If you are electing a representative who thinks the status quo makes sense, ask them to explain why. And get them to work to fix it. This is no way to run a country. Democracy does not need to be implemented stupidly or by accidental overgrowth of paper.

2008-10-04

Brilliant commentary on the financial crisis

Not where you would first think to look, but this show is intelligent, lucid, and not superficial.

http://www.thisamericanlife.org/Radio_Episode.aspx?episode=365

Give it a listen.

2008-10-01

Bailing out a sieve

Never mind. It was more complicated than I thought.

There was somewhere around 1 to 2 $trillion of house value in the bubble (out of perhaps $10 Tn total, not including commercial property), and tied to that was a disproportionate amount of uncertainty since all the securitized mortgages mixed them together and then the CDS chains of insurance created multipliers. Now in theory each mortgage sums to one when you count up all the pieces and match buyers to sellers along the chain.

The problem is, they don't match up. Pieces of the chain fail or are not synchronized because of lack of liquidity and so the contractual values get claimed at several points in the chain and not satisfied. Before the chains can work properly, they are in pieces and you have 10s of $Tn of these failures dangling in commerce-space. And it is not helped by all the players independently trying to claim what they can and refuse what they can. If someone claiming on you goes bankrupt then their ability to press the claim is deferred or even vanishes, so there are profits to be had by delaying, and meanwhile the same logic is applied to your claims. Sauve qui peut.

Cool. Financial wizardry intended to reduce risk instead, because the implementation is faulty, has ended up greatly magnifying the destruction.

In the end, money ends up in very different places than where it starts, somewhat randomly.

Learning from Experience

Over the years I've liked some of the things McCain did. But I'm not going to vote for him.

You see, I believe in accountability, and in the record of experience. In this case accountability for a record of poor judgement. Support for going into Iraq when it was clearly a trumped up case with no solid evidence, certainly nothing to support killing people (I can say that: I wrote that diagnosis to my friends the day after seeing the atrociously embarassing charade at the UN Security Council). A complete botch of what started as a good idea on immigration policy. Support for the un-Patriot Act and other Bush shenanigans. And choosing a totally unprepared Sarah Palin as a person to have a significant chance of being president within 4 years - something that would surely ruin any chance of reshaping the USA and rebuilding its strength. We cannot have a complicated, powerful nation like the USA run by a person of mediocrity.

No, it just does not add up to good judgement. A record of experience yes, but the experience is that McCain just is not good enough. And one reason why accountability is a good principle is that it removes people proven (or most likely) to have bad judgement. There is plenty of experience that people with bad judgement in the past will continue to make many mistakes. That should be one reason we are alert to a candidate's record. This is not a matter of forgive and forget, this is a case of what you saw is what you would be getting.

I'm not exactly sure Obama is ideal: they never are. Still, he is clearly an exceptional person quite capable of running a brilliant campaign and so far has acquitted himself well in debates. A reasonable chance of being a good chief executive. While holding McCain accountable for his record, the summary of his experience is he is not good enough. We need to elect only the best. The lesson of the past 8 years makes the cost of incompetence starkly clear.

One Megaword

Government has a tendency to decline in quality. Stopping this is difficult. Revolutions don't usually work, the odds are heavy on getting a new crew of criminals. Constitutions can be circumvented, ignored, or even forgotten, as the US Congress illustrates. So a good elected government drones on, churns out bad law and pork barrel, and over time the junk accumulates. A two century old democracy has a lot of stupidity codified and ever decreasing will to deal with it. How could this be remedied?

An indirect way is to limit the work product of government. Consider if the total of legislation could not exceed a million words - and moreover, words found in a ten year old dictionary just to keep them simple and to the point. If you want to write something new in, you have to cut something old out to make room. And give the courts the ability to cancel sections as being incomprehensible, so you can't get too cryptic. Hey, maybe just appoint a jury in a randomly selected jurisdiction and have them read one new law and pass judgement on whether it is clear (maybe answer a set of multiple choice questions about it). Assign each law (including laws alterred by deletions) to a different district. Make sure we can all understand the rules.

After all, nearly all of us remember less than a million words. If some issue is not important enough to rate a few sentences in the most important million words, then firstly why is congress concerning itself with such pointless detail, and secondly why should citizens need to worry about it? A million words, that is about 2,000 words per congress person or 10,000 words per senator. And still a million words each citizen might be touched by.

Oh and that includes budget items. And you can't cancel a section within 4 years except retroactively (which means if it was law anyone in conflict with it is pardoned and compensated, and if it was revenue the government gives it back with interest).

Maybe that would get a government to sit up straight and pay attention to the essentials.

2008-06-15

Controlled substances

But of course, they are not. Illegal drugs kill people, finance gangsters and terrorists, and bring down governments. But "controlled"? Not even close.

Why? A simple economic analysis shows the fundamentals. Fierce inelastic demand, diverse sources world wide, and limited trade routes. Hey presto, the trade routes are amazingly profitable and impossible to close down. That same economic analysis shows the key to control: usurp the trade routes and become the supplier.

That's right, set up official suppliers. Regulated legalization.

It has happened before, with prohibition and alcohol. It worked.

If addiction is going to be an engine of power, why leave it to the scum of the world to profit by it? Why have our soldiers killed over drug profits in Afghanistan, or drug financed arms used to subjugate a country like Burma, or drug corruption tearing at our neighbors in Mexico or Columbia, or our police and civilians killed over drugs in our cities?

It is not as if the "war on drugs" has any success to recommend it. Time to get real.

Addiction is a medical problem. Becoming a regulated supply does not mean encouraging addiction. You can regulate it in ways that help addicts to get control of themselves or even eventually recover. Require supervision by a doctor. Monitor the amount used. Require education - hey, you can't buy that junk until you pass the test where at least you anwer 20 questions about how it can kill you.

Ah, you say, but if it is regulated then that means the addicts will still go to the gangsters to get it. True, to some extent. Some folks will need more than the regulations allow, more than a doctor is willing to sign up to watch. They'll trade with friends who fake addiction, and they'll trade with criminals with alternate supplies. Yes. And some of those addicts are going to die using stuff they bought from "official" suppliers. But since we accept that hundreds of thousands of people already die from alcohol and tobacco every year (many many times more than are likely to die from "hard" drugs) it looks like this is part of freedom we can get used to.

It will not be worse than the situation today. It might be a little more visible, but if done right you balance the legality of the supply with the access to and education about treatment, and the number of addicts is no worse. Certainly the number of addicts reduced to ruin and crime to support their habit should be much less.

The essential thing is that the regulated channels must sell cheap. You pass the test, you got the signature from your doctor, you take your quarterly physical and listen to the lectures - whatever - but at the end of the day when you put down your money on the counter it costs peanuts. The major drugs - heroin, cocaine, marijuana - they are after all agricultural products. Cigarette prices with taxes thrown in are a good model of the costs to the addict.

Gangsters just can't compete. There is no margin in it. You can't finance terrorism or buy major armaments or corrupt governments on tobacco money. You can't even buy bling to live the city life as a dealer. Government regulated supply should explicitly, unmistakeably aim to sell low. Instead of a war on drugs, this is using monopoly economics to price the gangsters out of business. If the regulated supply is cheap, even the trickle of backstreet sales to folks who for some reason do not want to join the regulated system will have prices kept down to some ratio of the legal price. Done right, that means there just is not enough money to support criminals.

Wouldn't it be nice to see the Taliban's control over Afghan poppy farmers wither away? And Columbian and Mexican cartels collapse? And our cities have their biggest source of crime cleaned up without a shot needing to be fired or an extra jail built?

It makes a lot more sense than anything we are doing today.

2008-06-14

Embargoed lucre

Spread freedom and democracy - that is the US line of the moment. If you put cynicism on one side, it does have a nice ring to it and would actually be welcome in many places. But doing it at the point of the gun is clearly fake and other beligerant policies have little credibility. Is there any other approach that could make a difference?

One lever that is rarely used is finance. It is not always true, but there are many places where a corrupt tyrant has ruled for a while, padded some foreign accounts, and escaped to retire in luxury and usually a new head of state keen to repeat and improve upon the avarice of the predecessor.

Often what is happening is perfectly obvious, documented, and beyond dispute. However since we have the fiction of sovereignty we find our hands tied while a dictator is in power, then when they disappear to some (equally sovereign) refuge the victimized nation finds itself obliged to pay the debts lest it be a stain on its credit rating.

So, what would happen if that credit link was repudiated, in advance? Suppose the USA and Europe, together or separately, were to take issue with regimes of obvious corruption and oppression (consider North Korea and Burma as current examples) and simply say: any debts incurred by those regimes (including all its agents) shall be considered to be the personal debts of the dictator, without legal recourse to any successor regime.

This is in effect an economic embargo not needing military force. At that point, anyone extending credit (which is pretty much anyone trading goods or armaments) has to consider that they have no recourse when the regime ends. It is not as simple as getting cash payment on the spot, since that cash is generally financed by external loans (written of course to have nothing to do with arms or whatever is in trade, but mortgaged against the assets of the country) which would at that point suddenly be very bad business deals.

For good measure make any transfers from that government to any external account (named or numbered) subject to repudiation by a successor regime ("We would like our money back, please"). This makes it more complex to plan for that retirement haven.

Of course this idea has some holes. The financial centers of the USA and Europe would be displaced for shady dealing by other finance centers with less scruples. Still, it would not be long before they realize that the twin problems of picking up some of the worst credit risks along with being tarred with some of the worst political liabilities is not worth it. It only works today because everyone has allowed it to. It is based on international laws that are rooted in centuries ago when it was the norm for countries to be ruled by thugs who affected manners (known as "royalty") and such things need uprooting from time to time.

As soon as some major players change their rules the old system will seem untenable and even ridiculous - as it has been for a long, long time. There clearly is no ethical or practical reason for the current system where a nation is held responsible for the debts racked up by thugs who are supported by the arms and corruption which were financed by those debts. We could change this, we certainly have no need to keep such rules.

2008-06-12

City traffic

City traffic is a problem around the world. It is undeniably attractive that individuals should have transport that gives personal freedom, but the current system of cars does not work well and cannot scale to serve everybody. Can we fix this?

Electric vehicles hold a lot of promise. Of course they are low pollution and potentially efficient, but one big advantage seldom discussed is they are much more precise. Regulating an IC engine is messy and slow. You get at best a tenth of a second lag on an accelerator plus the transmission gears and torque converter add imprecision. An electric motor, by contrast, can be controlled to the millisecond and is potentially closely coupled to a wheel. This means that electric cars in principle can be driven in close formation convoys and in efficient sequencing down streets and across intersections without stopping. Of course, you would need computer control since this is beyond human capabilities, but most drivers would probably be happy to put the daily commute on autopilot if it would be faster and predictable.

The payoff for a city is that much less of its valuable space needs to be invested in roads, and the existing rights of way could carry increasing traffic without the expensive and near impossible politics of building new roads or widening existing.

Of course, electric vehicles are tough to build and battery technology could take years or decades to become practical. But there is an alternative way to go, combined with batteries or hybrid drive: electrify the roads. Why not provide electric power through the roadway to participating cars? It would cost, but it would upgrade the capital value of the city and probably pay for itself in a few years so it is an interesting path to explore.

You don't need to put rails in the road. Keep your rubber tires and asphalt surface, but experiment with embedding conductive materials in the formulation of both. If the front and back tires connect as a circuit (a configuration that supports bikes as well as cars) then you can probably get a couple of kilowatts per wheel pair which is enough to sustain urban speeds.

How to avoid electrocuting pedestrians, and how to make the front and back tires get opposite parts of the circuit? Well, you make the system smart. The car would need to signal the road through the wheels (or possibly through separate sensors) so computers controlling the road know the vehicle, the speed, and where each tire is. A pedestrian does not signal the road nor come close enough to a moving car to be included in the active circuit, so to someone walking the road is just a safe paved surface.

The vehicle could then be lightweight since it does not need onboard battery for long range, with current lithium batteries a 50 lb set would allow a half ton car easy range in the suburbs to and from the core electrified network plus boost for hills and acceleration or across gaps in the road supplies. Since there is no heavy IC engine or transmission, with a powerful electric motor under 100 lb, making a 2 person vehicle with luggage space with a loaded weight in the half ton range seems quite reasonable.

Of course the roads would be expensive to build, you would probably need to prefabricate the surface elements and lay them out. However, many cities have property values of $1M per 100 ft or often much higher, so it does not seem at all impossible that something on the order of $100k of equipment might be embedded in the road to maximize traffic flow while minimizing vehicle costs, dangers, and pollution.

Our cities today are pretty stale in terms of transportation. We could transform them and not necessarily by forcing everyone into shared, inflexible public transport systems. We are by no means at the end of the road for personal automobiles.

2008-06-07

Think week

This week's goal, one essay per day. The underlying theme will be to take a big problem and suggest a lunatic solution. Of course, there will be some risk that the craziness actually makes sense. Which of course, since the sane world is often senseless, should be expected.

2008-03-01

Opening Salvo

After 5 decades, everything seems a little crazy. Still, some of it seems to have a kick and may be worth sharing. In general I'm not fond of rambling, and have no intent to blog a diary. Just the occasional idea, boiled for urgency and condensed to essentials.