Corpus planning has attracted attention ever since Deborah Cameron's seminal "Verbal Hygiene" (1995). However, evaluations of corpus planning aimed at addressing linguistic discrimination have been surprisingly scarce. Because corpus planning costs energy, time and money, evaluations are important for future actions. This study discusses how an evaluation of corpus planning can be conducted by performing a detailed, critical and empirical analysis of a limited Swedish language policy action aimed at addressing linguistic discrimination. The results show that, although the policy action was successful in terms of lexical change, the change was superficial, since the desired shift in focus, from individual/group to environment, occurred to only a very limited extent in the selected texts. From this analysis, I argue that there are three significant matters to attend to when evaluating corpus planning for these situations. The first is that a meaningful corpus should be created, preferably one comprising symmetrical texts from both before and after implementation of the policy. Additionally, the research questions that are formulated should be narrow enough to be operational. Finally, quantitative and qualitative methods should be combined to ensure a broad understanding of the outcomes of the corpus planning.