How Much is a Computer Worth?

Does the cost of a computer teach us anything about the value of labor?


Let’s talk about a specific computer to make the discussion simpler. The example will be a standard computer, produced by the millions, which sits on desks everywhere. You plug it into the wall, connect a display, keyboard and mouse, and you can compute with it. You load in a program to process data, put in the data, and you get some output.

Now let’s talk about two of these. One sits somewhere where it processes data from some major corporation, or bank, or government institution, and the output is used for far-reaching decisions, affecting large amounts of money, many people, or both. Another sits somewhere where it processes data from a small business, maybe a little restaurant or a retail outlet for mass-produced clothing. The computers are identical. You could swap one for another and not know the difference except for the serial numbers. Is the first computer worth tremendously more than the second one?

No. They cost the same amount. They are made of parts which each individually cost the same. You can get another one to do exactly the same processing of data at some computer store or on the internet. It is just surprising to some that a computer which processes incredibly important data affecting incredibly important decisions costs the same as the one which processes mundane data affecting almost no one. Why is this? Why doesn’t the use of a computer change its value, and allow the computer store or the online site to charge hundreds of times more for the one which is going to be used for the important job. Why don’t they ask beforehand what the use is and charge accordingly? Obviously charging by the value of the use would make more money for such computer distributors.

The reason is that fungible objects have prices set by the manufacturing cost, not the value of the use. Competition among manufacturers of computers sets the range of prices that can be charged, and more powerful components change the price, not more powerful uses of the resulting computer. A bigger hard drive or a faster CPU can get a bit more money for a computer, but only by a factor of a few, not by thousands.

If there was only one computer in the world, it might be worth some immense sum, but when there are computers in every nook and cranny, the price of a computer is pretty much fixed. One distributor might make a few percent more profit by having a better return policy, or having faster shipping, or something else, but this amounts to some minor percentage, not a factor of hundreds or thousands.

Identical computer, near identical price.

What about the wiring of the computer? If it is connected to a printer in the retail clothing store is it less valuable than if it is connected to a printer in an important office in a major bank? No, plugging it somewhere doesn’t affect the value? How about if it is loaded with an expensive software program rather than an ordinary, run-of-the mill software suite? Nope. Computer still costs the same. What if it is attended by important people instead of unimportant people? No change. Computer still costs the same. There is simply no getting around the fact that fungible objects have a price set by manufacturing costs, plus distribution costs, rather than by their intended use.

Let’s consider the software in more detail. If one software program is written by a good programmer, from instructions he or she has received, and another is written by a good programmer, also from instructions, but different ones, and program length and complexity is about the same as the first one, is it worth more? Software programs aren’t identical, but the cost of programming it is fairly fixed by the length and complexity of it. If you think of a data input – data output type of program, there might be 20 items to consider for each of two programs, and 20 conditions that have to be examined, and 4 outputs get computed. The cost of the program is measured by the cost of the programming, and would be about the same. If there were side conditions, such as speed of computation, there might be some percentage saved by having a more efficient program, but this doesn’t have much play in simpler situations, such as the one we are considering. The cost of both the computer and the program it runs is governed by the cost of construction. Period. If one program is used in the bank’s high official’s office and the other is used in the retail store owner’s office, they are still both worth about the same money. We do assume both are tested programs produced by experienced programmers, rather than some spaghetti code put together by a neophyte, just as we formerly assumed the computers are not made in someone’s garage out of random components.

How about if they are networked; does that change the price? Suppose the first computer in the example is connected to another, identical computer, in another giant bank and both of them are connected to a third, identical computer, in a third giant bank, and so on. Has the computer now become worth a thousand times more that it did for the retail store owner? No.

What if the government of the land where these two examples lived made a law saying that computers for high-importance tasks needed a special license, which required extensive fees and long delays and a very restricted quota. Every such computer would have to be registered, and grave penalties imposed for violating the regulations. Only a few would be available each year. Suddenly, the high-importance computer is worth thousands of times what the low-importance computer is. By throttling the supply, the demand price goes up. A giant bank with a high-importance computing job can easily afford to pay these fees, and high-importance computer brokers would be there to collect them. There would be extensive opportunities for corruption and crime. Thus, it now becomes clear how to make a computer worth thousands of times what an identical one would be worth. The same mechanism would work with the software.

Let’s compare this electronics example with a biological one. People can process data. They have brains which are composed of layers of neurons, but the number of neurons doesn’t vary by very much, and the speed of operation doesn’t vary by much, and the organization of the brain is pretty much the same. One measure of intelligence is IQ, and it is a bell-curve distribution. There are lots of people around with IQ’s of 130 or most any other number short of the extrema. In many jobs, intellectual ones, people function like computers, in that data comes in and decisions come out. Why are they paid so differently?

Education and experience provide the rules by which decisions can be made. The goal of education is to replace experience as much as possible, but unlike computer algorithms, no complete set of instructions is available for decision-making, and therefore it is somewhat random in outcome. People simply do not have the solid rules needed for a good computer software program, so there is nothing available except relying to some degree on chance. Why are different individuals paid so differently, and especially why when there are no good rules for them to follow, nothing derived from repeated exercises in similar situations, and nothing from theories of management and personal relations? What we do have is a natural human tendency to form hierarchies where individuals are stacked over one another, and guilds where various procedures are instituted so that supply of certain training is limited. While these both have some benefits, they obscure the fact that people in our modern era are largely fungible for employment, and if guilds and hierarchies did not occupy such dominant roles in our culture, the value of an individual would be much more closely related to the cost of raising a person plus the cost of educating him or her. In other words, like computers and software programs, people would have a value related to their inherent costs. Like government throttling of supply of computers to high-value uses, these two factors can greatly raise the cost of an individual to an enterprise.

People do not even have the variation in price imposed on computers by the technological progress in the field of semiconductors and other related areas; people haven’t changed much at all over the last few millennia. So, the intrinsic value of people is much more likely to be within a range of a few times average salary, on an annual basis, or lifetime earnings, on a lifetime basis.

This has to do with something in economics called ‘just deserts’, which means that by and large, people get something proportional to the value of their contribution. ‘Just deserts’, if applied to computers, would entail the same government bureaucracy discussed above, but even larger if it was not simply restricted to only computers in very high-value operations. Perhaps the theory of ‘just deserts’ does not make any sense at all, but is simply a simplistic explanation, unrelated to economic reality, of the observation of the huge disparity in the worth of different individuals caused by the non-economic actions of guilds and hierarchies.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s