Saturday, February 19, 2005

Government and Efficiency

Government organizations are usually (but not always) plagued by politics and power, which produce inefficient and dishonest bureaucracies. Political partisanship makes it political suicide to take responsibility for a mistake. As a result, politicians take no responsibility, and the job of the lower ranks of the bureaucracy is to cover their ass when the blame comes down. The best way to avoid making mistakes is to do nothing, and to shuffle all real useful work to the bottom of the ladder. This is the zero-error principle, but it might just as well be called the zero-effectiveness principle. When a mistake is made, the worker bees take the fall, and the politically savvy middle or upper managers escape blame and continue to clog the system. The cost of political partisanship is that the government is nit-picked into paralysis. Thus, the expansion of government is the fault of all sides of the political spectrum.

Indeed, these professional bureaucrats protect themselves by building empires, and can actually turn their own incompetence to their benefit, by demanding more money to address their own failings. Governmental power attract the corruptible, who seek to turn that power and public wealth to their own ends. To counter this decay, and to attempt to prevent normal administrative errors, new departments accumulate to check existing ones (the Department of Homeland Security is an example of this.) This is how the government grows, again, usually through partisan criticism and demands for change. Ironically, much of this expansion may be the result of calls for more fiscal responsibility--bean counters on bean counters on bean counters.

The most efficient form of government is probably a benevolent dictatorship, but dictatorships never stay benevolent. The tradeoff in government is between effectiveness and damage control. Too little power and nothing gets done, too much power and the wrong thing gets done. Most democratic governments are built for gridlock, to provide checks and balance on power. You may not get the most efficient government, but you will get a less harmful one. The problem comes when a politician wants heavy handed effectiveness. This soon causes the system to grow rapidly, as the new powers attract political beaurocrats, greedy for a piece of the pie, while at the same time abuses of that power encourage the growth of institutions of restraint. It is not surprising, then, that the Bush administration has ushered in an period of unprecendented governmental expansion that will likely continue long after it has left office.

Far from solving the problem, calls for large government cuts make it worse, because these are merely simplistic, populist postures to gain votes. They act like binge and purge dieting, burning muscle and leaving fat. Political beaurocrats, who actually do no work, are politically savvy enough to escape the cuts. The ones who get cut will be those too busy doing actual work to notice the axe coming down. Indeed, the perfect political beaurocrat will be right there beside you, calling for the cutbacks, because that will score political points and make his job more secure. And the ones most likely to jump this demoralized beaurocracy will probably be the most competent, who are also the ones most likely to find other employment. The end product is survival of the fattest--who are also the ones that cost the most.

Nor are these problems unique to government; large corporations are prone to the same problems, and private industry with a pipeline to the public gravy train combines the worst of both worlds. The military-industrial complex is a good example of this. There are simply some jobs that must be done on a non-profit basis, with the public paying for and handling the books.

The cuts, once made, are rapidly felt--capitalism is, after all, a partnership between business and government. The economic system we all live under was created and is sustained by continual government intervention. It's amazing how many people don't know this and think all government should be abolished. So the cuts must be rolled back--in fact, the government must be expanded well beyond its previous size to get any useful work done.

The solutions would be less partisanship and a higher fault tolerance in the beaurocracy (rather than the zero-error principle), with emphasis on accomplishment rather than mere error-avoidance. But above all, we have to abolish the narcissistic management style, epitomized by Bush administration and encouraged by pundits everywhere, where leaders take all of the credit but none of the blame.

Monday, February 07, 2005


A heuristic is a simplified rule of thumb, an approximation of reality. We use these in programming because they are simpler than a full fledged model, easier to understand and faster in execution. I've just written a new movement system for our game, because the old heuristic was too simple to accomodate the tasks laid upon it. The two things that strike me in this kind of work are that a) it's mind numbingly hard, especially if the data is messy (in this case, it's essentially polygon soup) and b) the more accurate the model, the more likely it is to handle unforeseen cases appropriately. Getting the model right in the first place means fewer patches later to fill in the cracks. But we usually don't put the work in until the old one breaks, the bug reports come in, and we have to admit that current system just won't handle the strain, and write a new one.

The human mind uses heuristics too. We use simple pictures of the world, blunt approximations of reality, which serve us well as long as these simple notions do not have to support a heavy load. Indeed, I suspect that the picture that we all hold of the world consists of a very low resolution image with a few spots of detail. The attention to detail in these pictures vary from person to person. Prejudices result, I suspect, not just because our ideas are wrong in a kind of one-to-one correspondence with the facts, but that they lack the level of detail required to represent their subject matter accurately. Ignorance breeds prejudice by labelling parts of our world as empty spaces with the words "Here there be dragons." It is not just that the spaces are blank, but that we insist that the map is accurate. Out of pride we expend a great deal of effort in defending, not our knowledge, but our ignorance. And as I've said, it's hard work to build a more accurate model. It's easier to just insist that you're right.

The broad swaths of primary color which represent large realms of unknown details are often represented by buzz words and caricatures. Words like liberal, conservative, fascist, communist, Nazi, terrorist, fanatic, Christian, Jew, Muslim, atheist, humanist, and all the other words for social, religious, political and racial classification are typically thrown about with the blithe assumption that everyone knows what they mean. In fact, I suspect that most of the people who use these words actually have little idea of what they mean. In many cases they are employed like the caricatures used in propaganda posters, with no idea of the beliefs or histories that lay behind them. This is quite evident when an outright bigot tosses them out as an insult. It may be far less obvious when someone who accuses others of being a fascist but is himself a fascist in all but name, unaware of the consequences of his own beliefs.

One of the chief tenants of magical thinking is that if you can name a thing, you can control it. This is indeed part of the motivation behind the name calling; you're just 'that', no more, and we know how to handle 'that'. It is also the reason that the name Jehovah was not to be spoken in the Old Testament. Jehovah was not a simple word, but a pointer to an entire system of thought and experience. To bandy the word about in common usage was to presume that everyone understood what in fact very few understood. If knowledge is power, ignorance is a mortal weakness. It is dangerous to believe that you can control what you do not understand. The slave you believe cowed may turn and become your master. The enemy you believe defeated may rise again. And the very evil you most despise may appear, not just in those close to you, but in your own eyes when you look in the mirror.

The danger posed by relativistic arguments, not just in ethical arguments, but in arguments of fact, is that they despise the very effort of educating oneself, of correcting these oversimplified models of the world. The easiest way to prevent any meaninful challenge to our beliefs is to represent all evidence and argument as merely differing opinion, with no objective means of comparison. I find nothing more infuriating than to hear this argument come on the heels of a bald unqualified assertion, used only as a defense when the original assertion is challenged. The relativist speaks the unvarnished truth--until, of course, he is asked to provide proof, and then there is no truth. This legalistic maneuver attempts to strike the objection from the record while allowing the original assertion to stand. It is a maneuver that fanatics all over the world have learned to use quite well, defending absolutist doctrines under cover of relativism, triumphalism in the name of plurality, and hatred in the name of tolerance. In all cases, it is the bold refusal to correct a model of the world which has long since crumbled under the weight of evidence.

In the Old Testament, when asked his name, God answers, "I Am Who Am." Whatever other myths and stories surrounded their faith, the Jews believed first in Reality, in the truth as best as they could know it. Their dedication to learning has served them well. Whatever your god is, reality will eat it for breakfast.

And reality doesn't care a rat's ass what your opinion is.