This study uses aggregated municipality data, for the years 2001 to 2009, to explore whether direct payments to farmers affect agricultural employment in Swedish municipalities. The decoupling reform in 2005 implied that Sweden had to introduce a grassland subsidy, one that had unexpected redistributive consequences as it suddenly increased CAP payments to grassland regions. In some municipalities, the payments more than doubled. Thus, since this particular policy change is truly exogenous, the reform is used to identify a subsidy effect. With a difference-in-difference model, a permanent increase in agricultural employment can be attributed to the grassland subsidy. Each additional job in agriculture costs about SEK 223,000 (about € 25,000) in subsidies, which is low. However, the subsidy effect is primarily saving jobs in agriculture, i.e. the grassland subsidy may be slowing down the process of structural change in grassland regions.