Monday, April 30, 2007

Alan S. Blinder has gone soft

Alan Blinder is a brilliant economist but judging from this article in the WSJ opinion journal, I'm wondering what he's been drinking. He has missed out a very large part of the outsourcing issue and because of this, he is wrong.

One of the determinants of price is productivity. If Lineland can produce 3 widgets per worker while Flatland can produce 9 widgets per worker then all things being equal, a worker in Flatland will earn 3x as much (if he earns more than 3x then it is more economical to produce in Lineland). Now because of comparitive advantage if a worker in Lineland can produce either 3 widgets or 1 blob and a worker in Flatland can produce 9 widgets and 15 blobs then even though Flatland has a considerable productivity advantage for both widgets and blobs, Flatland will tend to specialize in blobs and Lineland will specialize in widgets as this will yield the biggest advantage to both nations.
China fits into this classic mould - China produces widgets (manufactured goods) with relative efficiency while the US produces blobs (services) with relative efficiency. The end result is that the US is still a net exporter of services. That there is a huge trade deficit stems from other issues that we can discuss later (clue: a huge source of the US trade defecit is not China...).

It gets murkier when both Flatland and Lineland have specialized in the same area.
India is one of these anomalies. Blessed by a decent higher education system, and the fact that the manufacturing section was tied down by red tape, their services sector took off. This is not a sustainable model. India has hundreds of millions of poor people who don't have a hope of ever getting a high school education let alone a university education - many these people are unemployed because the traditional route up through manufacturing is blocked to them. This will be a source of major social instability in the future.

This anomaly in India is manifested in some pretty obvious ways. Witness the airport in Delhi compared to Shanghai. Shanghai is a modern, efficient airport where it is clear where to go and how to get there while Delhi is a total mess. If you're not careful in Delhi, you could easily get into some serious trouble. Other differences are also apparent - bad roads, terrible train transportation, awful telephone system, unbelievable amount of red tape - the list goes on. The services sector is booming but everything else in India seems to be struggling. Already their infrastructure deficiencies are having an impact on their growth.

Are there productivity differences between an information worker in the US vs. an information worker in India (for example)? While I am not questioning the talented young men and women in India, the answer is clear. An IT worker in India has to contend with: poor communications networks, bad roads (to go to and from work), inefficient retail networks (and logistics), less powerful machines (though the cost of a PC has been going down steadily while becoming more powerful, the cost of the most powerful machines at the time has remained high), less efficient financial markets to produce venture capital and less effective management. Were all of these factors equalized, the cost of an IT resource in India would be the same as that in the West.

It is almost surprising that we are able to outsource to India given the fact that the West has the same relative efficiencies as India. But this is not surprising when we start to break down the numbers. Where India has excelled is in the low margin, high volume sort of work. In sheer quality of research and product development, most of the innovation still happens in the West (with some notable exceptions). This does not mean that the West should be complacent. Protected industries are ripe for outsourcing as the price far exceeds the productivity of the resources. Restrictive immigration laws also works as a form of industry protectionsim by reducing the supply of workers. Old controls need to break down - unions and professional bodies need to accept that the market must determine wages not committees. And in the end, it is better to have workers come and work in your country than having the firm move away in search of workers.

Returning to Mr. Blinder, a professional in the West does not have anything to worry about if they keep their productivity advantage and if their industry has a relative productivity advantage to the rest of the economy of their country.

Olympic Security

Interesting how security is chosen for the London Olympics. Considering how London is a potential target for the myriad terrorist outfits out there, this is quite plainly stupid. Forget about quality of service and ability - anyone can arrange for security as long as they sponsor the events. Encouraging wouldn't you say?

Friday, April 20, 2007

Why I stopped worrying and learned to love string theory

I learned about superstrings when I was 12. But it was in university that I first started to really learn about what they were and the mathematics behind them. The phrase that came to mind was EPICYCLES (see above). It could be a natural bias of mine but I wouldn't characterize string theory as elegant. Aside from the fact that dimensions spring out of the theory like grasshoppers out of an open box and that the theory predicts an inordinately high number of possible universes and that just about nothing in string theory is testable, it lacks any sort of unified set of principles. It still treats space-time as a stage where strings are the actors. Fast forward to today.
String theory has become M theory and the lack of testability is beginning to worry some physicists (and the fact that all the best talent is being sucked up by this monster). Two good books address this Lee Smolin's The Trouble with Physics and Peter Woit's Not Even Wrong. And some recent developments such as the accelerating universe seem to cause problems for string theory. Of course anything can be explained by string theory, just as the positions of the planets could be explained by epicycles if you added enough epicycles.
So what are the alternatives to string theory? Well there are a few:
  • Abhay Ashtekar's and Lee Smolin's Loop Quantum Gravity which quantizes space-time and worries about unification later (makes sense as one of the biggest problems is that gravity and quantum theory are incompatible).
  • Then there is Alain Connes's noncommutative geometry which looks at a geometry linked to the noncommutativity of quantum phase space (position and momentum). Unlike ordinary numbers where 2x3=6=3x2, certain operations are noncommutative (ex. moving on the surface of Earth - if you start on the equator and move 200km north and 200km west, this will put you in a different position on earth then first travelling 200km west and then 200km north).
  • Twistor String theory, which is based on Roger Penrose's Twistor Theory, which relates real 4 dimensional Minkowsky space to 4 dimensional complex space.
I suspect that all of the theories (including M)are orbiting some common theory that they don't see yet. There are some striking similarities sometimes in the mathematics of the theories that suggest this is the case. Problem is that M Theory is getting all the limelight and funding to the detriment of the other theories and to science.

But I do like string theory and it is mathematically quite splendid.

Wednesday, April 18, 2007

Tax Pipe Dreams

Excellent article here about how the tax code can be changed.
My personal favourite is the flat tax. Several of the Baltic states have adopted a flat tax and have done really well. A flat tax is the fairest as everyone pays the same portion of their income as taxes and a 0 marginal tax rate does not penalize success or risk taking. Interestingly enough, a flat tax actually increases the tax revenues for the state because firstly there is a strong positive incentive to earn more (the Laffer curve) and there would be fewer people wanting to dodge taxes because they deem it unfair.

Gun Control

Although greatly saddened by the events in Virginia Tech, we need to sit back for a moment before we start to react and start legislating.
The US is one of the few places in the world where citizens are allowed to carry arms with little or no restriction. The problem is that most citizens don't exercise that right. This is why some punk can walk into a school and gun people down - few take their own protection seriously and even if they did, ham-fisted laws barring guns from schools and campuses prevent them from exercising that right.

Mass murderers will always find ways of killing - either by walking into a school in broad daylight and gunning people down, or by burning down the house of a sleeping family, or by hunting people down one-by-one in dark allies. The question is whether we give citizens the ability to defend themselves.

Perhaps this episode should tell us less about the legality of guns than about the structure of our communities. Remember that the warning signs for Mr. Cho's psychotic nature were abundant - but nobody cared to follow up.

Tuesday, April 17, 2007

Space Pragmatism: NASA Goes Open Source for Missions

Space Pragmatism: NASA Goes Open Source for Missions Cool idea here to leverage the open source community. Might make for some leaner more interesting code. This can only be better for science.

Irrational persistence

Michael Shermer from Skeptic Magazine is always a good read. He also has a monthly article in Scientific American. This month's article is quite interesting - about why we persist knowing that the chances of success are quite limited. Anyone who's worked for a big company knows this. The company spends $200M on a project that is going nowhere fast. Everyone knows it but no one is willing to say "Enough is enough". Anyway Mr. Shermer can explain it better than me here.

Monday, April 16, 2007

This is scary...

The German government is planning to introduce really intrusive laws that allow the government to access the computers of individual people. According to Scaeuble:

Under certain conditions it must be possible for the Federal Criminal Police Office to search computers in secret.

This is a very worrying trend. There are typically three excuses that governments use to justify this sort of nonsense: Terrorism, child pornography, and now increasingly copyright infringement. Remember the recipe of would-be authoritarians - find/invent bogeymen and cower the citizens into submission.
The article can be found here. Mr. Schaeuble has a history of suggesting draconian civil-liberties infringements. Apparently the history of his own country hasn't made an impression on him.

Camille Paglia's articles in Salon magazine

Just found a collection of Camille Paglia's writings at Salon magazine here. I've always admired her writings and her ability to see past much of the intellectual baggage that the West has been accumulating for the last 100 years (read Derrida, Foucault etc.) that has condemned impressionable college students to mediocrity.

US productivity revisited

Interesting article here about the slowdown in US productivity. Basically there are short-term impacts to productivity that are dependent on the business cycle and long term impacts that are dependent on the "fundamentals".

Short term influences happen because during an expansion period in the economy, companies may delay hiring, which means that fewer people do more work. During a slow down, companies delay letting people go, which means that more people do less work. This is just a quirk and is more indicative of whether or not the economy will slow down in the short term.

Long term influences are more important. These depend on the right regulatory framework, investment by companies and technology.
The Economist report seems to suggest that productivity growth in the non-housing sector in the US is growing at a robust 2.8%. This is "good" news. This means that outside of the housing sector, which is currently massively bloated and is slowing down, companies are doing more with less. However, on the whole, productivity growth in the US is slowing (because housing and construction are weighed so heavily).

Productivity growth, for those who don't know, is important because this is the key driver for economic growth. This means that prices go down because fewer resources go into producing more and people earn more because they are producing more (incidently, this is one of the reasons why an Indian worker typically earns less than a US worker - because of the massive differences in productivity).

Stagnant productivity means that wages can't go up without impacting inflation and prices remain constant or even increase because increased competition for the same resources will drive up prices. Stagnant productivity growth means that the economy can't grow very fast without causing inflation unless the working population is growing.

Monday, April 9, 2007

Neurological origins for the law of diminishing returns...

An interesting article on Scientific American on the neurological origins of the law of diminishing returns. I suspect that there might even be a physical reason. One can see that much of economics could have a thermodynamic origin - ie. conserveration of energy. I wonder whether there might be a physical link to diminishing returns too...

Sunday, April 8, 2007

Space Pragmatism: Venture Capital Firms Looking to Invest in New Space

Encouraging. Private space companies are probably one of the few ways that we will see some serious advancements in rocket technology. Our rockets haven't really changed that much from Goddard and Oberth.
Space Pragmatism: Venture Capital Firms Looking to Invest in New Space

DRM coming down???

After Apple, now Microsoft? Have the record companies finally come to their senses and realized that treating their customers like criminals is not the best way to sell their products. The study of online trading of music and p2p networks is a study of disruptive technologies. There will always be people who will try to shut Pandora's box but they have always failed. Oh but the damage that they can do while trying...

Will the dismantling of DRM now be accompanied by legislative changes because our enlightened leaders will see that the DMCA is no longer that relelvant? Probably not without some fights ahead but the story with COPA ( Child Online Protection Act) gives some hope.

Companies need to understand when their business model has become obsolete otherwise they risk becoming obsolete. No amount of legislation will save bad business models - at least not for long.

Saturday, April 7, 2007

The economics of climate change

I find it worrying that policy makers are rushing headlong into the climate change hype without considering the serious ramifications of what they are doing to us and future generations. Instead of spending (ie. losing) billions on reducing emissions or going for energy sources that are intrinsicallly uneconomic (read inefficient), we should be building wealth to mitigate against any future catastrophe. I know from personal experience how important wealth is to mitigate against what life throws at you - the same principle applies on a global level.
Current approaches include paying billions for wind farms, which are inherently unreliable as a continuous source of power, building solar power stations in the South of Germany (which having lived there I can attest to the fact that the sun is not something you normally see there), funding corn-based ethanol in the US even though it is still considered vastly inefficient, requires imported natural gas to produce, and artificially raises the price of corn, a staple food for many people around the world. Another approach is the use of taxation, which is gaining traction among some economists. This makes very little sense as well. Taxation of energy means that we pay more for everything. Paying more for something that is delivering the same value is equivalent to inflation. Furthermore, the money goes to the government - not exactly the pinnacle of efficiency. What will the government do with the winfall from a fuel tax? Fund more billion dollar roads to nowhere?
A more sensible approach is to build a hedge against a future crisis - and what better hedge than wealth?

Scientific Consensus

Many times before, scientific consensus has proven to be ill founded - sometimes with dangerous ramifications such as Eugenics - which made itself present in the laws of many countries. Today, we are dealing with a new scientific consensus - "climate change". While I'm not disputing the evidence surrounding climate change and I do acknowledge that it might be happening, I find suspicious anything that we have universal agreement and dissenters are treated like blaspheming heretics to be burned at the stake.

Ronald Bailey wrote an excellent article about the problem of consensus which can be found here.

Tuesday, April 3, 2007

Chinese moon rover

The Chinese are bent on moon exploration and I can already see a mini-space race in the making. Unlike the West, China has no issues about using nuclear energy in space as is demonstrated by their intention to power their moon rover with nuclear energy. Indeed if we are really serious about establishing colonies on Mars and on the Moon, and further afield, there is no choice but nuclear. The power that one can derive from solar energy decreases as the square of the distance from the sun. Mars is approx 1.5 AU from the sun (with a max. distance of approx 1.6 AU). This means that the power that we are able to derive from the sun is between 2.2 - 2.6 times weaker than on Earth. This means that either the solar panels need to be twice as efficient or the collecting area twice as big to get the same power as on Earth.
Batteries are also problematic. For example, if we powered the Galileo space probe using conventional batteries instead of Radioctive Thermal Generators (RTGs) and expected the same duration for the mission (70000 hours), the probe would have batteries weighing 47T! This does not include the fact that the batteries themselves need to be kept warm.
Robert Zubrin in his excellent book Entering Space, discusses the use of nuclear power in space. He goes one step further and recommends the use of an actual nuclear reactor.
Not only do we need RTG's, we need to move beyond them to actual space nuclear power reactors that use nuclear fission, rather than mere radioisotope decay, to generate tens or hundreds of kilowatts. The reason for this is very simple. On Earth, it has been said, knowledge is power. IN the outer solar system, power is knowledge.
Perhaps its time to get past the neo-luddites and start to look at nuclear power more rationally.

Sunday, April 1, 2007

US productivity growth slowing...

According to the Wall Street Journal, the US productivity growth is slowing from an impressive high of around 4% in 2004 to just 1.4%. The Economist newspaper highlighted this in 2006. If this is not just a glitch then it would suggest that we are unlikely to grow as fast without incurring the cost of inflation. The WSJ suggests that the lull in productivity is due to technology already achieving it's impact and that there is little productivity growth to be squeezed out from IT. While this might be contributing, I tend to disagree that this is the biggest part. There are huge sectors of the economy that still haven't embraced IT such as medicine and construction. Furthermore, there are still huge strides to be made in automation such as automating IT support and supply chain automation across all sectors of retail (not just the Walmarts).

Here's my theory: Productivity growth is slowing partly because of the increased regulatory burden in the US.

Productivity growth depends on output increasing per worker (I could be mean and suggest per hour worked but let's move on). In a tough regulatory framework, companies are forced to spend more on tasks that don't increase their ability to produce. This usually entails hiring more people, which works to reduce productivity growth.

The last few years has seen one of the most aggressive increases of regulation in the US for the last 20 years - Sarbanes Oxley. Section 404 of SarbOx requires firms to place strict internal controls on finance such as record keeping, auditing etc. This means huge spending on IT such as storage, security and software without squeezing any benefit from it. So for many large firms a part of their IT budget has been sunk into SarbOx without giving them the benefits from that expenditure.

Another less direct impact has come from the post September 11 paranoia of foreigners studying in the US. This has led to a massive drop in the import of talent that in the past has fuelled many of the bright young start-ups that have contributed to the growth in productivity.

While regulation alone probably doesn't explain the dramatic drop in productivity growth, I certainly see it contributing a fairly large component of it.

Ronald Coase the Nobel Prize winning economist first coined the word "Transaction Cost" to describe the costs associated with an economic exchange. He related transaction costs to the size of the firm. Regulators should keep this in mind when they place new shackles on business.