On the use of hypothetical price data to estimate hedonic models in a developing country context

Given the lack of market-based property price data in developing countries, imputed rental prices have been recently used to estimate the implicit value of some home characteristics and related amenities. This study assesses the validity of using imputed rental prices to estimate hedonic models in a developing country context. Using data from Guatemala, we compare the derived implicit values of housing characteristics from hedonic models using hypothetical and actual rental prices. We estimate an endogenous switching regression model to control for potential endogeneity of the decision to rent a home. Our results indicate that hedonic models with hypothetical and true rental prices yield statistically different estimates of marginal values for some housing characteristics.