Excess Intertia and Cross-Model Validation

Demonstration of Excess Inertia

Note: in order to run the simulation referred to in this slide, go here, to the Java Applet version. You will be directed to download the latest version of the Java plug-in.

In the graphic to the left, we have constructed a demonstration of the excess inertia concept from the previous slides. Initially, the two firms are connected, which could signify that they are joined in the same compatibility network, or that they adhere to the same industry standard. Furthermore, their network externality parameter, β, is at the same low value, 2.0. This signifies that the standard is inferior, in other words that belonging to this network isn't very beneficial for the two firms.

Press "Go" once to see the firms converge to the nearest equilibrium under these conditions. In the equilibrium, the outputs of the two firms are equal, thus two lines in the graph below join together.

Press "Go" once more to see what happens when Firm 1 leaves the network, and switches to a superior standard. The agents to the left disconnect, and Firm 1's β goes up to 2.1. The green line in the graph below represents Firm 1's output. When the system settles into the closest equilibrium, Firm 1 has a higher output and profits than Firm 0 owing to a better standard, yet still lower than that in the pre-switch situation.

To observe what happens if both firms switch to the superior standard, press "Go" one more time. Agent 0 upgrades to β = 2.1, the two agents join in a network, and converge to a new equilibrium. Note that Agent 1's output is now higher than in the two previous situations.

Clearly, both firms would benefit if they were to switch to the superior standard. If we were to assume that the firms did not trust each other to switch, they would remain on the lower standard, as predicted by Farrell and Saloner.

Notice how our method mirrors that of Farrell and Saloner. We must simply assume that firms do not trust each other, but the payoffs under different player decisions emerge from the Katz and Shapiro model. Farrell and Saloner exogenously assign payoffs, but the lack of trust between firms emerges as a result of the model parameters. The next step would be to construct an encompassing model in which both payoffs and uncertainty were endogenous.

Previous Slide                                                           Next Slide