Free software, illegal immigrants and the minimum wage

Luther Martin of Voltage Security looks at the tension between free/open-source software thriving in the marketplace and reconciling this with the need to make a profit.

The invisible hand at work

Milton Friedman, winner of the 1976 Nobel Prize in Economics, once described how market forces work to make lead pencils available at a price that cannot be matched by an individual consumer making their own pencils.

To do this, a person would have to find a source of wood and graphite and process these raw materials into a pencil, a job that most cannot do very efficiently.

Adding an eraser to the pencil makes the project even more difficult, potentially requiring a source of rubber. Attaching the eraser to the pencil with a metal band is even worse, requiring mining and refining of suitable ore as well as further processing of the resulting metal.

Adam Smith's 'invisible hand' of market forces, however, effortlessly coordinates the efforts of thousands of people whose individual efforts ultimately result in the inexpensive pencils that we have today.

Products involving more advanced technology provide an even more impressive example of this. You can only buy a handful of pencils for the price of an inexpensive digital watch. And although many people could probably improvise a workable pencil from wood and charcoal, they would be hard pressed to create a digital watch in a similar way.

Market forces can accomplish amazing things, but they do not always behave exactly how we want them to; the invisible hand cannot be easily or precisely controlled. Minimum-wage laws are a good example of this, and show that there may be unintended side-effects from manipulating market forces.

Minimum-wage laws reduce employment

The fundamental law of supply and demand tells us that increasing the price of something tends to decrease the demand for it. In particular, if we increase the price of labour, then we also tend to decrease the demand for it, so we should not be surprised if minimum wage laws actually end up decreasing employment.

Many studies of the American labour market have estimated that a 10 per cent increase in the minimum wage decreases employment of teenagers by between 1 and 3 per cent, so there seems to be some evidence to support this position.

Other regulations that increase the cost of labour behave similarly, reducing the level of employment as they increase costs.

The jobs that tend to disappear if minimum wages are either introduced or increased tend to be those that are priced out of the market.

If a particular job is only worth £3 per hour and employers have to pay £4 per hour for a worker, they will be effectively losing £1 per hour by employing such a worker. Such jobs will either tend to go unfilled or be filled in ways that circumvent the minimum wage laws.

In particular, many of the jobs filled by illegal immigrants are those that are worth less than the minimum wage - providing conditions under which illegal immigrants thrive is one way in which markets deal with the limitations imposed by minimum-wage laws.

The same invisible hand that directs the creation of pencils and inexpensive digital watches seems to create a solution to the problem caused by minimum-wage laws: that some jobs are not worth the minimum wage yet still need to be filled in some way.

Limitations of the enterprise software market

The requirements for traditional enterprise software are different than the requirements of home users. Enterprise users require interoperability with complex legacy systems, while home users typically only look for compatibility with a small number of relatively-simple operating environments.

Enterprise users also require around-the-clock availability of technical support, both before and after a sale, and the detailed testing process through which many business information technology purchases need to pass often require vendor support at each step of the way.

On-site visits by vendor representatives are often required as part of the sales process of enterprise software. These visits require the time of skilled technical staff as well as the expense of travel to customer sites and add to the expense of selling enterprise software.

Development of enterprise software is also expensive, even if we outsource it to places like India or China. 

The result of these additional costs is that there is probably a lower bound for the price at which traditional enterprise software can be profitably sold and a corresponding lower bound on the total cost of ownership (TCO) for it.

Thus it seems that the market for enterprise software is constrained by its own version of a minimum wage, although this constraint is not imposed by government regulations.

And just like jobs that are worth less than the minimum wage will tend to go unfilled in the face of minimum-wage laws, some enterprise software probably cannot be justified by its cost, which is bounded below by the constraints of the enterprise software market.

If the lower bound on the TCO of enterprise software is a hypothetical £60 per user per year but an application is only worth an average £40 per user per year, then using such an application will cause a net loss to its users of £20 per user per year. Such software will generally not be purchased, even if there is a need for it.

Some software may still be sold at such prices, however. Some may be sold to customers who overestimate its value; some may be sold to customers for which it actually has a higher value due to the nature of their business.

And much like labour markets may deal with minimum-wage laws by finding workers who will work for wages below the legally-allowed minimums, it seems likely that market forces will also find a way to provide software for which there is a business need, but for which the lower bound for TCO makes unfeasible to develop and market through traditional means. FOSS may be the natural solution to this problem.

There are still significant costs involved with using FOSS, to deploy and support the software. Technology may be free, but using technology is not.

It may be the case, however, that the lower TCO that is possible with FOSS provides a way to deliver products for which there is a need, but for which the costs of traditional enterprise software are too high to justify its purchase and use. This may be the invisible hand that creates the market forces that allow FOSS to thrive.

If such forces did not actually create FOSS, they have almost certainly have been factors in its success. So we may be able to understand FOSS as being a natural part of our existing capitalist system instead of being opposed to it.

November 2006