Land consumption and income in Ecuador: A case of an inverted environmental Kuznets curve.

• We use the spatial panel regression model to estimate the land Kuznets curve. • We examine spatial auto-correlations in the building permits in Ecuador. • A Bayesian selection method allows selecting the right spatial model and weight matrix. • Economic development has a U-shape relationship with land use. • Spatial spillover effects have a limited extent and affect only closest neighbours. The ratio of building permits to population is a key indicator to evaluate land consumption. However, few researchers focus on land consumption and its environmental spillovers, for developing countries. The aim of our study, using a Bayesian comparison approach applied to a spatial panel, is to analyse the existence of an inverted U-shaped curve relationship between land consumption and economic development, namely the environmental Kuznets curve, with data that ranges from 2007 to 2015 for 221 cantons in Ecuador. The Bayesian comparison approach allows us to identify: i) the spatial weight matrix that best fits the data, and ii) the best spatial model according to the type of spatial spillovers (local or global). These are both of extreme interest because a knowledge of the extent to which the spatial spillovers spread over space, and their functional form, supports the planning of effective land use policies. The results do not support the inverted U-shaped hypothesis of the Kuznets curve. By contrast, the curvature is convex, which means higher levels of land consumption for higher levels of wealth. Spatial spillovers spread to a limited extent, highlighting an imitation game among agents, both institutions and private agents, in the neighbour locations. Policy implications go from the strengthening of the institutional framework and local tax management, to the urban regeneration to limit real estate speculation. All these interventions should be coordinated among neighbours to avoid freeriding behaviours. [ABSTRACT FROM AUTHOR]