As agricultural policy reform and its effects have become increasingly territorialised, analyses that attempt to explain or predict impacts need to be both more localised and to identify spill-over effects. Local and regional general equilibrium approaches have become increasingly popular because they can extend predictions of policy shocks obtainable from partial equilibrium sectoral models to identify the wider effects. However, agriculture is usually described as a single sector in input–output accounts, whereas policy shocks that affect constituent commodities with differential impacts will have inter-industry effects that are different to those implied by average input–output coefficients. Regionalisation of aggregated input–output tables adds further to these difficulties. The objective of this study is to develop a practical method for dealing with these problems. It describes the theoretical basis of aggregation bias and shows how it can be measured, in two contrasting case study regions in the UK and Sweden. Having established that this is a significant issue, a simple but effective procedure is demonstrated, based on additional information on variable costs, which transforms policy shocks from a direct change in agricultural output to that transmitted to the suppliers of inputs. This method provides an impact close to that which could be calculated if the general equilibrium system had indeed been disaggregated, and supports use of this approach in impact studies where insufficient time or funding are available for complete disaggregation of an agricultural sector’s regional accounts.