Animal welfare scientists have accumulated knowledge and developed interventions to improve livestock welfare, but these are poorly adopted in commercial practice. Animal welfare interventions are rarely tested for economic viability and this limits their uptake. This study employs Stochastic Partial Budgeting (SPB) to determine the viability of animal welfare improvements. Aggression between pigs is used as an example because there is a large literature base from which to draw interventions, and the problem has persisted for decades without resolution. Costs and benefits of three interventions to control aggression (pre-weaning socialisation, synthetic maternal pheromones and large social groups) were estimated by reviewing the academic and industry literature and by conducting a survey of sixteen pig farmers. The net effects were compared to farmers' willingness to pay (WTP) for interventions to reduce aggression as identified by recent research. Results are consistent with prior research which indicates that improving animal welfare generally comes at a cost to producers. Nevertheless, pre-weaning socialisation resulted in a neutral or positive net effect 38% of the time and should be central to campaigns promoting the control of aggression in the industry. Exposing pigs to synthetic maternal pheromones did not improve profitability but the net costs were small and within the realms of WTP for a sub-group of farmers with animal welfare goals. The net costs of converting existing buildings in order to house pigs in large social groups were beyond the realms of farmers' WTP. The approach adopted in this study, of combining SPB with WTP from the sector, should be extended to other animal welfare issues.