Different strands in social sciences have analyzed the role of specific skills on distinct political stances, however, a consensus has not been reached yet. On the one hand, sociologists stress that workers with specific skills are less replaceable so they tend to face a lower probability of job loss. On the other hand, political scientists argue that specific skills are less transferable than general skills and therefore, they are associated with worse consequences in case of job loss. For this reason, there is not an agreement about the effect of specific skills on redistribution preferences. In the current paper I urge for the need to distinguish between firm- and occupation-specific skills. Following labor economics literature, I stress that occupation-specific skills (OSS) are more relevant than firm-specific skills (FSS) in shaping labor market prospects and that a new theory is required. The main difference outlined is that, contrary to FSS, occupational context -concretely occupational unemployment rates (OUR)- moderates the effects of OSS. With high OUR workers with high levels of OSS have low transferability of skills and at the same time their repleaceability is high. On the contrary, when OUR is low, OSS is associated with high transferability of skills and low repleaceability. I provide a simple and tractable model that grasps different relations between OSS and redistribution preferences. To show the relevance of this distinction I use the European Social Survey to study redistribution preferences with the insights of this theoretical framework. Moreover, I discuss different OSS measures.