A Short Introduction to Software Economics

What follows is a part of my master thesis that I wrote on the conditions of the competition in IT industry in Turkey. Before delving into the details of the topic I wanted to give a short account of software economics. Most of the time people understand the internal economical analysis of a software project when the topic is considered. That’s quite reasonable but there is another aspect of software economics, which is about its macroeconomic conditions in a market and for example how those conditions affect the behavior of a software house that strives to exist in a very unstable market such as the one in Turkey. So below is a short account of software economics as an instance of a wider area of study, the economics of knowledge.

I’ll provide Turkish translation of this very shortly.

According to famous Austrian economist Carl Menger, knowledge is embodied in capital goods. (Baetjer, 1998) For example cars carry the knowledge and skills of their designers and engineers in their bodies and features. Software is an extreme case of this approach. Because software is abstract in nature and reality of software is not inherently embedded in space. (Brooks, 1986) In software it is much easier to differentiate its knowledge aspect from its physical aspect. Software is embodied intellectual capital. (Baetjer, 1998)

Software products are inherently different than most of other engineering products. First of all software products are abstract and intangible in nature as stated before.  Software is invisible. (Brooks, 1986) Software products are not manufactured in the sense of other engineering manufacturing such as cars or electrical appliances; they are developed. Most of the cost of a software product is spent for developing and later maintaining it not for manufacturing it. That means once a software product is developed it costs almost nothing to manufacture it by writing it on a CD/DVD or another media or making it downloadable over the Internet.

There are fundamental differences between the economics of physical things and the economics of information. (Evans & Thomas, 1999) (More specifically economics of software may mean two different things: Economics of a software project and economics of a software market. In this study the second one is meant by the term.) Economic goods can be categorized according to whether they are rival or non-rival and whether they are excludable or non-excludable in consumption. Rival goods are those that can be consumed by only one consumer at a time. Most of the physical, durable or nondurable, goods, apples or hammers for example are rival goods. But non-rival goods can be consumed by more than one consumer at a time. For example radio waves carrying information are non-rival in that they can be absorbed by many antennas from air at the same time. Being excludable means non-paying customers can be easily excluded from having a good such as apples and hammers again. For non-excludable goods it is very expensive or sometimes totally impossible to exclude others from having benefit from it. Conventional economic goods are both rivalrous and excludable. That means they can be traded in competitive markets and are privately provided. (Romer, Endogenous Technological Change, Oct., 1990)  But ideas and technology are different:

Conventional economic goods are rival and excludable and are privately provided. Public goods are neither rival nor excludable and cannot be privately provided. The interesting feature of technological advance is that it is alleged to have the characteristics of a public good but is privately provided by firms that do research and development. (Romer, Capital, Labor, and Productivity, 1990)

Technology and more specifically information technology is nonrival. Once a piece of software is created it can be used anywhere and any number of times because ideally it is independent of any physical object. When we think of a piece of software that must have a piece of hardware to run on, then we can realize that calling software a perfectly nonrival product is an idealization. So most of information technology products including software mush have a complementary good. But for example human capital is absolutely rivalrous because it is totally tied to human body. So goods including technological ones have different levels of nonrivalry.

Nonrivalry has two implications that are very important in terms of comprehending the nature of the economics of software: One is the fact that nonrival goods can be accumulated without bound on a per capita basis. For example the benefit of a technological innovation can go much beyond what a worker can produce in her lifetime.  The second one is that making nonrival goods absolutely private is almost impossible due to spillovers. Once a technological achievement is achieved by a firm, that firm would normally try to protect it via patents and other legal ways so that only can it get a competitive advantage out of that achievement. But that’s not easy, for example the period of time to prevent such an invention from being used by other economic agents does not last longer than that of most physical goods. So nonrival goods have incomplete excludability which is  a function of both technology and legal system. (Romer, Endogenous Technological Change, Oct., 1990)

A software product is an intellectual property of the person or institution that developes it. But it is still too hard to completely protect a piece of software’s behavior or properties or any technique or approach embodied within it from others to copy, imitate it in a variaty of ways; because software is abstract.

When a piece of software is sold while the new owner has it, the seller does not cease owning it, itcan sell the same software again and again with almost no extra cost if that software is of type packaged software. This is called zero-marginal cost of software production. But developing a new software may cost the same as the first one.

The third and most fundamental premise is that instructions for working with raw materials are inherently different from other economic goods. Once the cost of creating a new set of instructions has been incurred, the instructions can be used over and over again at no additional cost. Developing new and better instructions is equivalent to incurring a fixed cost. This property is taken to be the defining characteristic of technology. (Romer, Endogenous Technological Change, Oct., 1990)

But as explained before since the enterprise software products need a lengthly customization process for the customer the principle of zero-marginal cost does not equally  apply. This is totally due to the nature of the customized enterprise software products.

Software never decays but a software product may become unuseable. When this happens it is discarded and new and possibly better one is installed for use. Software never wears out but it frequently changes through new releases, bug fixes, etc. to keep up with new business and technological requirements.

Cost is also another topic that needs to be handled differently in the world of software because software has its own peculiarities. For example classical dichotomy of fixed cost vs. variable cost seems does not seem to be clarifying for software. Most of the cost for software is spent not for developing it but maintaining it. On the other hand different types of software have different behavior in terms of these two types of costs. While packaged software has almost zero-marginal cost customized software may have more marginal cost that its fixed cost. That’s why some software engineering costs model include different types of costs. (Wang, 2007)

Traditional and technological economical goods behave differently when one of the inputs in their production is increased. For traditional economic goods law of diminishing return holds true: In the production of those goods if one of the inputs is kept increased in amount while other inputss are fixed eventually marginal product of that input will go down.

Mechanisms of increasing returns exist alongside those of diminishing returns in all industries. But roughly speaking, diminishing returns hold sway in the traditional part of the economy—the processing industries. Increasing returns reign in the newer part—the knowledge-based industries. Modern economies have therefore become divided into two interrelated, intertwined parts—two worlds of business—corresponding to the two types of returns. (Arthur, 1996)

Information has perfectly increasing returns: spend the money to learn something once, and that knowledge can be reused at zero additional cost forever; double the number of uses and the cost per use halves. (Evans & Thomas, 1999)

Software products are believed to have increasing returns. This is mainly due to following reasons:

  • Development costs of software products are very high. Because main resources for a software products is know-how. Know-how of specific business processes and needs, know-how of software development processes, techniques and tools, and etc. Software development projects requires large amount of R&D activities. Ideally and generally once a piece of software is built it can be reused and sold again and again at a fraction of the development cost.
  • A software product attracts more and more customer once it reaches certain level of acceptance among the customers. So once a software product becomes popular even the speed of its increasing return increases. This is called network-effec or network externality. Internet triggers the network-effect and helps a lot good and bad words about a product permeate every single Internet-citizen’s mind. So the more it gains prevalence, the more likely it will emerge as a standard. (Arthur, 1996)
  • Once individuals or enterprises invest in a software product they orient themselves around that product. They get trained on that product, they change their customs accordingly, they even integrate that product into their IT infrastructure i.e. they all incorporate that product in their daily  and professional activities. After all that they follow the updates and new versions of the same product at some extra cost. So more money a customer spends on a software product more it goes into it. That’s called vendor lock-in that makes switching costs for the owner very high. Thus when a product gains a significant amount of market share its producer can set higher upgrade prices, for example. These all create effects that increase the return of a software product.


  • Arthur, W. B. (1996, July-Aug). Increasing Returns and the New World of Business. Harvard Business Review .
  • Baetjer, H. J. (1998). Software as Capital. IEEE Computer Society Press.
  • Brooks, F. (1986). No Silver Bullet. IEEE Computer .
  • Brynjolfsson, E. a. (Dec. 1996). Network Externalities in Microcomputer Software: An Econometric Analysis of the Spreadsheet Market. Management Science, Vol. 42, No. 12 , 1627-1647.
  • Evans, P., & Thomas, W. (1999). Blown to Bits. Harvard Business School Press.
  • Liebowitz, S. J. (Spring 1994). Network Externality: An Uncommon Tragedy. The Journal of Economic Perspectives, Vol. 8, No. 2 , 133-150.
  • Low, L. (1992). Economics of Information Technology and the Media. River Edge, NJ, USA: World Scientific Publishing Company, Incorporated.
  • McBreen, P. (2001). Software craftsmanship. Addison-Wesley Professional.
  • Romer, P. M. (1990). Capital, Labor, and Productivity. Brookings Papers on Economic Activity , Microeconomics, Vol. 1990, 337-367.
  • Romer, P. M. (Oct., 1990). Endogenous Technological Change. The Journal of Political Economy , 98 (5), 71-102.
  • Wang, Y. (2007). Software Engineering Foundations. Auerbach Pub.

Bu yazı toplam 1046 defa görüntülenmiştir.