Let us now briefly reiterate some of the key features of the of the model of internet behavior constructed in this essay.
The first aspect of the model which was introduced was the demand for internet usage. Internet usage was argued to influenced by five main factors. These factors were the price of usage, the price of substitutes and compliments, changes in the consumers income, consumer preferences and lastly, the network effect. In this essay the direct network effect was focussed on. It was argued that changes in network size would, up to a certain network size, produce a positive marginal benefit. Past this certain size, the benefit would become negative. This enabled a relationship to be established between demand and network size, such that where the increase in network size produced a positive marginal benefit, demand would increase, and where it produced a negative marginal benefit, demand would decrease.
The concept of the congestion externality was then introduced. The size of the congestion externality is related to the application mix at the time the congestion occurs and the characteristics of those applications in use, in particualr, their ADU. It was argued that if the application mix was dominated by applications with large ADU's and poor tolerance of delay, that this would produce a larger congestion externality than where the mix was dominated by applications with small ADU's and better tolerance of delay.
The congestion externality was then related back to the marginal of benefit of network size. It was argued that because reduced congestion would raise the marginal benefit of usage at that network size, the effect of lowering congestion would be to increase demand at congestion affected network sizes.
This is significant since it suggests that attempting to reduce congestion by expanding capacity is unlikely to be successful. By expanding capacity and temporarily lowering congestion, the marginal benefit of usage is increased and demand increases. This increase in demand may even completely cancel out the reduction in congestion initially bought about by the capacity expansion.
By contrast, it was argued that the introduction of a congestion toll may be more effective in reducing congestion. However the determination and implementation of an optimal congestion toll is problematic. Indeed this is made even more so if one allows for the possibility of positive externalities being associated with internet usage. The final suggestion of this essay was that evaluation of these two alternatives requires further research into a number of key issues and assumptions made during this essay.
Thus as the internet continues to grow in size and in its capabilities, the role of economics is likely to continue. Developments such as e-commerce, cybermoney, real time video and content-push technologies will all have an impact on the way people conduct their affairs. All of these are likely to increase the amount of traffic on the internet and have an impact on network size. A model for determining the most efficient means of catering for this increased traffic should thus be considered as an important task for economists wishing to make a contribution to this field.
Last update 12-Jun-98. To visit my personal web site click here.