|Beginning with a model in which technological progress is reflected by product variety, we provide a structural approach to estimate technology spillovers allowing for spatial interdependencies. To this end, we first present a theoretical model of TFP growth by decomposing TFP into quality and variety components. We address the quality component by introducing a country?s distance to the technological frontier. Quality is assumed to be a negative function of the technological gap of country i with respect to its own technological frontier. This technological threshold is defined as the geometric means of knowledge levels in all countries. We deal with the variety component by using R&D expenditure combined with human capital stocks. In doing so, we show how a spatial Durbin model can be obtained from a theoretical model and thus better capture technology spillovers. Our TFP growth model is estimated from a sample of 107 countries for the period 2000–2011. The main focus is on the role played by technological spillovers. They impact productivity growth substantially, as do traditional factors such as R &D and human capital stock. Technological spillovers are captured by the spatial autocorrelation coefficient and the indirect impact of R &D.