Siblings are a potentially important source of political socialization. Influence is common, especially among younger siblings and those close in age, who tend to interact most frequently. This suggests that the positions of an individual's next-older sibling will hold particular sway. In policy questions with a gender gap, then, those whose immediately older sibling is a sister will be more likely to absorb the typically female preference; those born after a brother, the male preference. Evidence from the United States shows that this pattern holds for general left-right orientation as well as for the preferred balance between public and private sectors. Just as American women are more likely to lean left and to see government intervention positively, so are Americans whose next-older sibling is female.Beliefs and preferences serve as the foundation for models of political economy; 1 they are similarly central to the study of everything from elections and legislative voting to international relations. 2 Understanding of the sources and formation of beliefs is consequently critical for political science. Many sources shape political preferences: direct experience (of, for example, recession), elite framings and the mass media all give information conducing to particular viewpoints. Of particular importance, though, are social interactions: not for nothing is the process of preference formation usually called 'socialization'. Parents, teachers, peers, fellow members of clubs and organizations -each provide cues about proper behaviour and the workings of society. Accordingly, all have received attention from scholars looking to trace the origins and evolution of individuals' political beliefs.Curiously, however, siblings have received relatively little attention as a source of socialization, despite their conspicuous place in the lives and perceptions of young people. In part, this is because it is hard to distinguish the independent effect of individuals' sisters and brothers, since the same parents and, often, community influences that affect the individuals themselves also shape the siblings. Moreover, in most (but not all) households, all members share racial and socio-economic characteristics, further reducing the scope for identifying and isolating siblings' specific influence.Yet siblings do vary. Most importantly, they differ in sex -and do so in quasi-random ways that expedite empirical analysis. Psychologists and sociologists have extensively considered this variation in siblings' sex, finding widespread effects on behavioural and educational outcomes. One of the more sizeable and consistent of these effects reflects the