Non‐parametric Modelling of Spatial Price Relationships

We apply non‐parametric methods to a consideration of price transmission processes within US egg markets at the turn of the nineteenth century. Gordon (National Bureau of Economic Research Working Paper 7833, 2000) labelled this as an era of ‘Great Inventions’ which contributed to the subsequent years of significant productivity growth and noted that the development of mechanical refrigeration and transportation technologies played an important role in this growth. Our models present certain advantages relative to parametric models traditionally employed in price transmission analyses. We compare results derived from local polynomial modelling with those obtained using non‐linear threshold models. Both techniques suggest that US egg markets were interrelated at the turn of the nineteenth century. However, non‐parametric techniques often suggest a higher degree of price transmission than that implied by threshold models. Results also suggest that threshold models may have difficulties in adequately capturing price relationship dynamics, especially when these are of a highly nonlinear nature.