["Australian Journal of Agricultural and Resource Economics, EarlyView. ", "\nABSTRACT\nThe purpose of this paper is to discuss the micro‐ and macro‐outcomes for the standard neoclassical carbon tax (whose burden falls on the producers) versus the implication of a budget neutral, performance‐based EU Common Agricultural Policy (CAP) focusing on non‐CO2 emission (N2O and CH4, measured in CO2eq) reduction, especially in the livestock sector. The subsidy is budget neutral as financed via the current basic payment of the first pillar of the CAP. The policies are applied to the single farm level model FarmDyn to take into account heterogeneity between farms and regions and the macro model MAGNET to take into account market feedback and possible leakage effects. Farm level mitigation potentials and costs from FarmDyn are used to calibrate the EU27 national milk production Marginal Abatement Costs Curves (MACCs) in MAGNET. On the other hand, scenario‐specific changes in prices of agricultural inputs and output from MAGNET are used as input in FarmDyn, improving the consistency of the modelling framework. The results of the simulations conducted in this framework show clear conclusions. On the one hand, imposing a standard carbon tax on non‐CO2 emissions in the agricultural sector is the most effective method to achieve a greenhouse gas (GHG) emission reduction as the carbon tax can lead to significant reduction of total agricultural GHG emissions (MAGNET). Nevertheless, this result is connected to several trade‐offs, such as farm income losses in the short term (FarmDyn) and losses in agricultural production and employment (different by country in the EU27) in the somewhat longer term (MAGNET). On the other hand, subsidising the emission abatement costs allows for mitigating almost completely the negative trade‐offs while still allowing a substantial reduction of agricultural emissions. However, subsidies easily result in overcompensation and increased prices of inputs and technologies that mitigate emissions.\n"]