Abstract
This paper implements the “spatial first differences” (SFD) method to a Ricardian approach to analyze spatial differences in climate and evaluate their impact on farmland values in New Zealand. We use property valuation data from 1993 to 2018 across different agricultural land-uses. Ricardian analyses harness cross-sectional variation in climate and farmland values to estimate the effect of climate on agriculture. However, Ricardian studies are vulnerable to omitted variables varying slowly or not varying over time. The idea of the SFD approach is to apply spatial analogous of the “first-differences” estimator for longitudinal analyses to control for omitted variables that vary differently than climate variables among neighboring units. The results suggest that a warmer or drier climate is associated with higher farmland values in New Zealand. These findings may be mediated by productivity differences, costly physical improvements, and climate amenity values. Besides contributing novel insights by applying the Ricardian approach to New Zealand data and overcoming challenges of unobserved heterogeneity with the SFD method, we conclude that, for a better understanding of the welfare implications of climate change, it is crucial to differentiate between the effects of costly improvements and amenity values, apart from the pure climate productivity effect.