The Wake County database contains a variable lakedist, which measures the distance (in hundreds of meters) from the property to the nearest lake.
Consider using this variable to proxy the value of access to lake-based amenities by including it in a first stage hedonic regression.

(a) Examine specifications that add lakedist and the natural log of lakedist to the baseline specification. Interpret the coefficient estimates and report the sample mean prediction for marginal willingness to pay for access to a lake. Does the functional form for distance matter?

(b) This specification includes no variables that account for other spatial aspects, such as school quality, access to employment centers, or access to other landscape amenities. How might this affect the estimates of the implicit price of lake access?

(c) Using the level distance variable, examine regressions that first add a control for commuting time (median_commute), and then spatial fixed effects (submarket dummy variables zone1,…,zone18). How does this affect your coefficient and marginal implicit price estimates?

(d) We might hypothesize that lake proximity is a highly local driver of home prices, in which case we would no longer expect to see an economically significant effect on property values beyond a certain distance or cutoff point. With this as motivation, Phaneuf et al. (2008) use a transformation of distance given by:

where maxd is a hypothesized cutoff distance. Note that 0≤lakedist≤1, and that if lakedisti>maxd the variable codes to zero. Since tr_dist is a decreasing function of lakedist we expect a positive sign when it is included in the regression. Examine a regression that includes this transformed variable for maxd=10 (i.e. 1 km). What is the average implicit price of distance for homes that are less than 1 km away from a lake? Note that you do not want to calculate the marginal implicit price of tr_dist per se, but instead want to consider the marginal effect of lakedist operating through the non-linear transformation.

Transcribed Image Text:

tr_dist, = max 1/2 4). of. lakedist, maxd