Estimation results for policy periods. (a) EFA biodiversity score. (b) Farm biodiversity score. (c) Non-EFA biodiversity score. Notes: Point estimates from linear fixed effects estimation with clustered standard errors. Number of farms n = 410, time periods T = 1–12, number of observations |$N = 2,341$|⁠. The model is fully interacted with the Post variable, which is equal to 1 for observations after the new policy regime (≥2014) and 0 otherwise. |$S*Post$| gives the effect of interest for the post-reform period. Source: Authors’ illustrations using SAEDN and SFADN 2009–2020.