Connect-The-Dots: Identification of Heterogeneous Marginal Willingness to Pay Functions under Time-Varying Preferences
Research output: Contribution to conference › Paper › Research
1.33 MB, PDF document
Models based on residential location choice have become commonplace in the non-market valuation literature. Rosen (1974) provides a utility-theoretic basis for hedonic models to be used to measure the welfare consequences of changes in local public goods and amenities. However, his proposed two-stage estimation procedure embodies a number of difficult econometric problems that have become the focus of research for decades. My paper builds upon the "inversion" approach suggested by Bajari and Benkard (2005) and the buyer-panel extension of that work proposed by Bishop and Timmins (2018). The latter paper shows how data on repeat purchases can be used to flexibly recover preferences with rich individual heterogeneity, but the method is unable to deal well with time-varying individual attributes that might prompt residential location changes. I expand that approach to deal with any number of time-varying individual attributes including income, family structure and other drivers of housing choice. I apply that method to detailed longitudinal data from the Danish census, and use the estimates to value non-marginal changes in violent crime rates. I demonstrate a significant and policy-relevant bias from failing to properly account for the endogeneity problems in Rosen (1974).
|Publication status||Submitted - 2020|
- Faculty of Social Sciences - Non-Market Valuation, Marginal Willingness to Pay, Hedonic Analysis, Violent Crime
Number of downloads are based on statistics from Google Scholar and www.ku.dk
No data available