116 Ethnic sorting in football transfers to other clubs seem to have little to no effect on their average ingroup share or outgroup fractionalization within their clubs. We can think of several reasons for why this is the case. Firstly, the degree in which clubs truly compete with each other through their ethnic composition is most likely quite limited. As mentioned before, most clubs offer roughly the same thing to members with respectively Dutch or migrant backgrounds, and due to selective entry and dropout, this is unlikely to change drastically over time. This means that for most members transferring to another club isn’t an interesting option. Secondly, even if clubs within the same area differ strongly in their ethnic composition, composition induced transfers will only occur if members weren’t effectively sorted upon first entry. In other words, members need to join diverse clubs with low ingroup sizes while they have a substantially better option. This is in itself quite unlikely (McPherson et al., 2001; Wiertz, 2016). Thirdly, under these circumstances, transfers for other reasons may substantially water down or even counteract niche formation. Indeed, if ethnic sorting is strong upon first entry, changing to another club because of a relocation, differences in competitive level, differences in playing schedules or another reason, may very well lead to lower ingroup shares or higher degrees of outgroup fractionalization. When taken together, these reasons make it possible that ethnic sorting through changing clubs is so limited that it comes lost within the rest of the traffic between clubs. An important drawback of this study is that the locations of the clubs remain unknown. If these data were available, it would become possible to find out more about how clubs in each other’s vicinity compare to one another in terms of ingroup share and outgroup fractionalization. This would not only give us a much better idea of howmuch room there is for ethnic sorting through transfers between clubs to begin with, but it would also allow for more complex quantitative research designs in which these transfers can be statistically modelled. Another issue which needs consideration is the fact that amateur football clubs do not form a closed system of group membership. They compete with various other formal and informal groups for members’ time and resources. This can be especially relevant for the membership of ethnic minorities who potentially have more interesting options outside of amateur football, which motivate them to drop out of their clubs. In sum, this study has shown that even though selective tie-formation and tie-dissolution are both driving forces behind the social segregation of civic life (McPherson et al. 2001), they do not necessarily work in tandem. In amateur