Using the home heating data, complete the following.

(a) Estimate a model that includes only the installation and operating cost variables, and check that you are able to replicate the estimates shown in exercise 16.3.

(b) Estimate a model that includes the operating and installation cost variables, as well as the full set of J−1 alternative specific constants (ASCs). Comment on the differences in the estimates relative to part (a). How does inclusion of the ASCs affect the estimate of the marginal willingness to pay for a change in installation cost? What is the likely explanation for the differences?

(c) Now consider the role of income in explaining choices. Specify a model that allows you to test whether higher income households are more or less likely to install a central (as opposed to room) heating system (i.e. more likely to select options j=1,3,5). Run your model with and without ASCs, and comment on the differences.

(d) Consider a situation in which a 15 percent of installation cost rebate is being offered when a heat pump is selected. Using the model with oc_ij, ic_ij, and the full set of ASCs, predict how the average probability of a household selecting the heat pump changes due to the rebate.

(e) Using the same model as in (d), calculate the sample average willingness to pay for the rebate program. Compare your predictions for observations 3, 14, and 21 to what you derived in exercise 16.3. Which of the two calculations do you find most plausible? Explain your reasoning.

Data from exercise 16.3

Suppose you have estimated a model that includes only the installation and operating cost variables. The results are:

Transcribed Image Text:

Vij = -0.6231 icj -0.4580oc, j = 1,...,5.