The lighting market model predicts market share as an aggregate of many individual purchase decisions using two analytic components: an econometric logit model that considers cost factors influencing each decision, and a technology diffusion curve that considers time dependent market factors influencing each decision. Additionally, LED penetration is calibrated through a reverse engineering process comparing past LED market share values predicted by the model to actual historical values. Figure 2.2 summarizes this approach. This approach of using a logit model and a technology diffusion model in concert is well tested and has been previously used in many forecast models